the systematic improvement of advice given by public ... · the systematic improvement of advice...

The Systematic Improvement of Advice Given By

Public Sector Call Centres

Thesis submitted to:

QUEENSLAND UNIVERSITY OF TECHNOLOGY

Faculty of Information Technology

Research Centre for Information Technology Innovation

Submitted in fulfilment of the requirements for the degree of

MASTER OF INFORMATION TECHNOLOGY

Neville Lindsay Schefe

Dip.T. (Science/Mathematics)

B.App.Sc. (Comp)

Grad.Dip. (Software Quality)

Supervisor: Greg Timbrell

June 2006

i

Statement of Original Authorship

The work contained in this thesis has not been previously submitted for a degree or

diploma at any other higher education institution. To the best of my knowledge and

belief, the thesis contains no material previously published or written by another

person except where due reference is made.

Neville Lindsay Schefe (Student Number: n0135127)

Signed: _______________________

Date: _______________________

ii

Keywords:

Knowledge Management; Call Centre; Information Systems; Problem Types; Quality

of Service

Abstract:

The persistent demand for increased accountability and value for money in the public

sector from both the public and governments raises the issue of quality of service in

advice-giving by governmental agencies. The goal of this study is to develop a model

to validate frameworks able to contribute to improved advice-giving through the

application of knowledge management principles.

Zack’s (2001) Four Knowledge Problem Model, Brogowicz, Delene, and Lyth’s

(1990) Synthesised Service Quality Model, and Markus’s (2001) Theory of

Knowledge Reuse are used to examine knowledge strategies in advice-giving

through the application of a case study methodology. Two Queensland public-sector

call centres are investigated.

This study confirms that although the studied call centres operate under differing

business drivers, agents have developed strategies generally consistent with those

suggested by Zack (2001) to deal with uncertain, complex, and ambiguous problem

types. No equivocal problems were encountered in the study. The solution of the

former problem pair of uncertainty and complexity relies on knowledge that is

codified and stored in databases, while the latter equivocality and ambiguity, seeks

out experts who apply both technical and functional knowledge to the problem

resolution. Roles performed by call-centre agents predominantly align with those

described by Markus (2001), with the opportunity to enhance performance through

contribution by shared-work producers to knowledge repositories. The problem-

solving strategies employed by agents and the technical capabilities of the call

centres combine to deliver a level of service quality which, although meeting client

expectations, has been able to be improved through the application of knowledge

strategies targeting efficient problem resolution through knowledge reuse.

iii

Table of Contents

Statement of Original Authorship............................................................................................... i

Keywords:.................................................................................................................................. ii

Abstract:..................................................................................................................................... ii

Table of Contents...................................................................................................................... iii

List of Figures............................................................................................................................ v

List of Tables ............................................................................................................................ vi

List of Tables ............................................................................................................................ vi

Acknowledgments ................................................................................................................... vii

Introduction.................................................................................................................................. 1

1.1. Context of the Study..................................................................................................... 1

1.2. Background to the Research ......................................................................................... 3

1.3. Research Aims.............................................................................................................. 5

1.4. Limitations of the Study ............................................................................................... 6

1.5. Outline of Thesis Structure........................................................................................... 6

Context of the Study and Review of Literature ...................................................................... 10

2.1. Introduction ................................................................................................................ 10

2.2. Academic Literature ................................................................................................... 10

2.3. Summary..................................................................................................................... 32

Research Method ....................................................................................................................... 34

3.1. Introduction ................................................................................................................ 34

3.2. Objectives of the Research ......................................................................................... 34

3.3. Approach .................................................................................................................... 35

3.4. Customer Expectations............................................................................................... 43

3.5. Problem Resolution Strategies.................................................................................... 44

3.6. The Knowledge Management Strategy ...................................................................... 47

3.7. Quality and the Call-Centre Response........................................................................ 49

3.8. Pilot work ................................................................................................................... 50

3.9. Implementation........................................................................................................... 51

3.10. Summary..................................................................................................................... 58

Call-Centre Work, Organisation, and Processes .................................................................... 59

4.1. Introduction ................................................................................................................ 59

4.2. A Broad View of Call Centres.................................................................................... 59

4.3. Evolution of the Study: CC-A .................................................................................... 60

4.4. Evolution of the Study: CC-B .................................................................................... 63

4.5. Operational aspects of CC-A...................................................................................... 66

4.6. Operational aspects of CC-B ...................................................................................... 70

4.7. Summary..................................................................................................................... 74

A Framework for Analysis ........................................................................................................ 76

5.1. Introduction ................................................................................................................ 76

iv

5.2. Zack’s (2001) Four Problem Model............................................................................ 76

5.3. Summary..................................................................................................................... 84

Patterns from Observations, Interviews, and Responses........................................................ 85

6.1. Executive Management Interviews............................................................................. 86

6.2. Management and Team Leader Interviews................................................................. 86

6.3. Responses to Calls ...................................................................................................... 95

6.4. Service Quality.......................................................................................................... 101

6.5. Summary................................................................................................................... 104

A Pragmatic Approach to Call-Centre Strategies................................................................. 106

7.1. Service Value............................................................................................................ 107

7.2. Summary................................................................................................................... 119

Conclusion................................................................................................................................. 121

8.1. Introduction............................................................................................................... 121

8.2. Research Question .................................................................................................... 122

8.3. Implications for Theory ............................................................................................ 124

8.4. Implications for Practice ........................................................................................... 125

8.5. Further Research Opportunities ................................................................................ 126

8.6. Summary................................................................................................................... 127

References ................................................................................................................................. 128

Appendices ................................................................................................................................ 134

Appendix 1 - General Information Processes in CC-B.......................................................... 134

Appendix 2 - Average Advice Line Wait Time – CC-A........................................................ 135

v

List of Figures

Figure 1 - Organizational Knowledge Evolution Cycle (Rus et al., 2001, p. 17) ........................ 17

Figure 2 - Capability-Creating Activities: Shared Problem Solving (Leonard, 1995, p. 60)....... 20

Figure 3 - Relationship among service-quality concepts (Dedeke, 2003, p. 280) ....................... 27

Figure 4 - Rules and cycles in resolving an equivocal problem (Weick, 1969, p. 77) ................ 31

Figure 5 - Convergence of Multiple Sources of Evidence (Yin, 1994, p. 93) ............................. 37

Figure 6 - The Structured Case Research (Carroll et al., 1998, p. 65) ......................................... 39

Figure 7 -Query-Response Cycle (Timbrell et al., 2005b, p. 546)............................................... 40

Figure 8 - Tiered Query-Response Cycles (Timbrell et al., 2005b, p. 550)................................. 41

Figure 9 - Micro-performance Approach to Trust (Van de Walle & Bouckaert, 2003, p. 894) .. 44

Figure 10 - Generic Taxonomy of IT Helpdesk Calls (Pentland, 1991, p. 102) .......................... 46

Figure 11 - Capability-Creating Activities: Problem Solving (Leonard, 1995, p. 60)................. 50

Figure 12 - Kinds of Responses to Calls to IT Helpdesk (Pentland, 1991, p. 115) ..................... 54

Figure 13 - Age Demographics of staff employed by the agency hosting CC-A ........................ 62

Figure 14 -Query-Response Cycle (Timbrell et al., 2005b, p. 548)............................................. 65

Figure 15 - Operational Context of the Call-Centres Studied...................................................... 66

Figure 16 - Call logging system – CC-B...................................................................................... 71

Figure 17 – Agent Performance Graph (Interview 2b - Full transcript, June 16, 2004) ............. 73

Figure 18 - Research components................................................................................................ 85

Figure 19 - Complex Problem Devolution................................................................................... 98

Figure 20 - Problem Devolution Hierarchy ............................................................................... 101

Figure 21 - Graphical Representation of Frequency of Quality Criteria.................................... 103

Figure 22 - Resolution Rate - Quality Matrix ............................................................................ 106

vi

List of Tables

Table 1 - Knowledge States (Zack, 2001, table 2) ....................................................................... 32

Table 2 - Summary of Zack’s Four Knowledge Problem Model (Zack, 2001, figure 1)............. 45

Table 3 - Three Cycles of the Structured Case Method as Applied to This Study....................... 51

Table 4 -Coding Scheme for Software Support Process (Pentland & Rueter, 1994, p. 494) ....... 53

Table 5 - Proposed Coding Scheme for Problem Type................................................................ 55

Table 6 - Proposed Coding Scheme for Resolution Strategy ....................................................... 55

Table 7 - Proposed Coding Scheme for Quality Characteristics .................................................. 56

Table 8 - Data definition for data capture tool ............................................................................. 57

Table 9 - General information processes in CC-B ....................................................................... 64

Table 10 - Summary of the Strategies Employed by Both Call Centres ...................................... 74

Table 11 - CC-A, CC-B, and the Q-R Cycle................................................................................ 75

Table 12 - Summary of Knowledge Strategies Used in the Call Centres by Problem Type........ 80

Table 13 - Frequency of Repository References in Interviews .................................................... 87

Table 14 - Frequency of Response Strategies .............................................................................. 94

Table 15 - Frequency of Quality Criteria ................................................................................... 102

Table 16 - Analysis of critical incidents (De Ruyter, Wetzels & Van Birgelen, 1999, p. 1139)109

vii

Acknowledgments

The genesis of the problem being addressed in this study lies with frustration at being

asked to “work smarter, not harder”, to “be open and accountable” and to “do more

with less”. Does this mean I am not working smart, that I am unethical and that I can

be more frugal in the application of resources to my work? On reflection, I decided

that to get different outcomes, I needed to do things differently and started on a

journey that has probably only taken the first few small steps to where I hope to end

up.

The inspiration to try different things in the apparent monolith of government came

from my Executive Director, Jan Archer, who had the job of forming a new team in a

previously untried model of corporate service delivery. Her support and mentoring

destroyed the myths I had long held as being absolute truths in operation of

government. The support of my current supervisors and all the staff responsible for

the two call centres studied has been exceptionally tolerant and obliging over the past

three years. I am extremely grateful for the access provided and the honesty with

which issues have been addressed. The interpersonal exchanges in the call centres

provided me with a valuable insight into the informal activities that contribute to

their operations. Also, their interest in the actual work I was doing and the positive

approach they took reinforced my feeling that well-managed staff in an environment

that provided both personal and organisational challenges and responsibilities will

continue to develop to meet those challenges.

The validation of my work is based in the coding and analysis of actual calls made to

the call centres. I wish to acknowledge the support of Kate Bowman who provided

the assistance, as timelines started to shorten, in the moderated coding of the calls.

She did this accurately and completed the work to meet a tight schedule.

The academic staff members at QUT have cajoled, encouraged, critiqued and

contributed in a way that kept me on track from the very first day I commenced this

programme. To combine a research effort with the demands of full time work means

decisions on priorities and milestones need to be carefully attended to. The work put

in by Greg Timbrell to ensure papers prepared for conferences and publication met

both timelines and an academic standard is greatly appreciated. I know I gave him

some late nights in the editing/review phase. The need for sound methodology was

viii

provided by Professor Guy Gable. He assisted tremendously in the formulation of the

original research question while Professor Alan Underwood was instrumental in

continuing this through to maturity.

Finally, as any researcher knows, the support of family is essential for the successful

completion of such a project. To Joan, who has now caught the bug and spends many

weekends and late nights in the study, Ben and Sam, I appreciate the sacrifices the

family has made to accommodate my excesses. I derive inspiration though from the

fact that my whole family is currently studying and achieving tremendously at

university level. Thank you and best of luck with your own endeavours.

Introduction

1

Chapter 1

Introduction

This dissertation investigates the ability to measure and improve the quality of advice

provided by public servants. The simple question of “How do public service

organisations learn from daily routines and translate this knowledge into better

service performance?” is addressed through a case study of two Queensland

Government Departments. Milner (2000) identifies desired global reforms in the

public sector, and the need for the creation of competence through knowledge

management strategies. Such competencies are also discussed by Pentland (1991) as

they apply to a dynamic, knowledge-intensive task environment. The goal of this

thesis is to validate existing theoretical models for customer interaction and problem

resolution within the Queensland Public Service.

1.1. Context of the Study

Faced with a decline of trust in public and private institutions (Gualtieri, 2004;

Mahon, 2003), chief executives are attempting to rebuild faith through quality of

service and response to client and citizen service expectations. Rather than adhere to

technocratic utopian strategies that look to technology as a saviour, contemporary

leading-edge agencies must now align their business requirements and the

organisational relevance of implementing a service-oriented architecture (Davenport,

Eccles, & Prusak, 1992; Kreizman, Kotsikopoulos, & Shah, 2005). Such alignment is

reflected in the trend to identify business problems correctly, and formulate their

related strategic solutions. Hence technology, such as call-centre automation, is now

seen as a driver for performance-based innovation. Clients also continue to demand

an increasingly higher quality of service, with each cycle of improvement feeding the

expectation of further improvement.

Public-sector Chief Executive Officers (CEOs) have recently added a new dimension

to performance reporting, emphasising integrated service delivery. Their strategic

documents reference e-commerce, electronic transactions for government services,

one-stop shops, and 24 x 7 access – all of which is currently referred to as “virtual

government”. Public expectation acknowledges the importance of merit and equity,

yet demands efficiency and effectiveness (Milner, 2000). How then is a CEO to

Introduction

2

deliver a public service, governed by political outcomes rather than economic ones,

that meets the expectations of an increasingly demanding client group?

This research seeks to investigate the application of a model that is capable of

informing public-service knowledge strategy and its associated technology

investment process. So as to limit the number of uncontrolled variables in the study,

and to take advantage of access to data, the thesis will focus on the analysis of

strategies employed in public-sector call centres. An analysis of problem types

encountered, and their associated mapping to an appropriate strategy based in

knowledge management principles, is used to identify areas of strength and

weakness. This method enables outcomes of such strategies to be predicted and

measured, thus engendering confidence that investment in such strategy will provide

commensurate return.

The research problem has its genesis in the increasing demands for efficiency in the

public sector, with public expectation of increased performance levels of public

officials. Information technology has been promoted as the panacea to many ills in

both commercial and public enterprises; however, the sharpened focus on

accountability of IT systems following Year 2000 projects has caused management

boards to take a more critical look at IT and other investments which purport to

deliver efficiencies.

Concurrently, the knowledge management agenda has gained increasing prominence.

It is widely accepted that the current decade will see growth and maturity in

knowledge industries where customers will continue to expect high quality, value-

for-money services executed in a timely and convenient manner. The public sector

will not be exempt from this trend. Knowledge management strategies and their

promise of capability improvement will therefore be called on to deliver further

public-sector efficiency and effectiveness. Successful knowledge management

programs sustaining themselves through demonstrable business results are

exemplified in the following statements:

• Ford Motor Company's Manufacturing Best Practices Program improves the

efficiency of manufacturing processes and the quality of products.

• Rolls-Royce's Knowledge Acquisition and Modelling Process improves

project management processes.

• BP's knowledge management program started with the exploration and

production process, but now includes project management processes.

Introduction

3

• Shell Oil's Global Learning and Development involves continuous learning

around the exploration and production process.

• CSC's Communities of Practice started with a bid-and-proposal process and

now extends across the company into all lines of business.

• J.D. Edwards' Knowledge Garden started by focusing on the sales support

process and expanded to include resellers and customer service.

• The Knowledge and Learning Practice at the World Bank Institute focused on

improving development projects funded through the World Bank. (Caldwell, 2004, p. 2)

Notwithstanding, Return On Investment guarantees are being sought prior to

approval of business cases, owing to links with IT and the scepticism of Chief

Executive Officers for fad management innovations. Call centres are an option in the

drive for cost-effective delivery of services.

1.2. Background to the Research

This research has been limited to the call centres of each of the organisations being

studied, since they provide an instance of service provision able to be documented

and analysed. Initial literature searches provided little contemporary rigorous

research on the contribution of call centres to business objectives, instead primarily

focussing on quantitative performance measures such as time-in-queue, abandonment

rate, and talk time. This study is unique in that it links business goals, knowledge

strategy, and quality criteria as they apply to call centres in the Queensland public

sector. Direction has been taken from Yin (1994) to ensure the criteria for quality

research design are met in this study. The call centres provide a consistent and high

volume of rich data, with instances able to be tracked from initial contact to

resolution of queries on up to 200 calls per day per operator. This data has been

available for real-time capture through “double jacking”1 conversations, and

historically through calls recorded for other organisation-specific purposes. This data

is supplemented by interviews, observations, and published materials, which have

been used both to scope the research environment and to validate research findings.

Interviews with senior departmental managers have determined the business drivers

for each call centre, and their integral position to the broader objectives of quality

service provision via multi-channelled delivery mechanisms.

1 “Double jacking” is the process whereby the researcher is able to listen in to the conversation between the caller and the agent without influencing the interactions. Supervisors use this facility as a component of individual agent performance analysis.

Introduction

4

One strategy embraced by the public sector to enhance services is e-government: the

application of communication and information technologies to the organisation and

operation of government (Teicher, Hughes, & Dow, 2002). An element of the

e-government strategy is the utilisation of call centres and their associated

technology. Side by side with Internet-based services, the call centre remains a

central plank in the Government’s service delivery platform for both public and

internal inquiries. This “joined-up government” or “citizen-centric government”

concept has been studied by a number of researchers (Ling, 2002; Milner, 2000).

This concept expresses a fundamental change in paradigm from a traditional

benevolent or patriarchal model to one of active engagement by the broader

community. Call centres provide the opportunity for the efficient and equitable

delivery of quality advisory and support functions to government, which is in-line

with the tax-paying public’s expectation that government will deliver efficient and

effective public-sector operations (Davis & Weller, 2001).

Government agencies are now expected to develop performance indicators for the

services they provide, which is a way of measuring the progress toward government-

declared objectives. Managing for Outcomes (MFO) (2003), a Queensland

Government initiative commenced in 1997, addresses integrated planning, budgeting,

and performance management, and is intended to achieve quality, client-responsive

services, value for money, and improved resource allocation. The emphasis on

measures of efficiency and effectiveness is reinforced in the 2002 MFO Performance

Management Framework document:

“Through monitoring output performance, the efficiency and effectiveness of

service delivery, including the extent, quality and benefit of services in

relation to cost, can be determined.” (Performance Management Framework,

2002, p. 1)

Difficulties have been encountered in the measurement of service outputs and

outcomes, since the effectiveness of certain critical Government functions (e.g.,

compliance functions) has little to do with service delivery, or discrete

measurable outputs (Teicher et al., 2002). More meaningful measurement will

be founded on the decomposition, closer examination, and greater

understanding of these services, including those provided by call centres. In the

quest to improve government service quality, efficiency, and effectiveness, an

Introduction

5

analyst must ask: “How do workforce and technical strategies combine to assist

the enterprise to measure and improve on the quality of advice provided to

external customers and internal clients?” Several authors have provided

tantalising components of the puzzle. These include Zack (2001), who offers a

model of problem types and strategies to solve them; Markus (2001), who

investigates repositories and organisational memory systems in order to develop

a theory of knowledge reuse; and Timbrell Nelson, and Jewels (2003), who

devised a methodology to test the knowledge reuse theories as they are applied

in government agencies. In order to develop a defensible response to the

previous question, the principles and concepts underpinning the work of these

authors have been used to investigate two Queensland Government call centres.

1.3. Research Aims

This research project aims to provide evidence that the quality of advice given by

call-centre staff can be improved through the judicious application of a model both

able to predict the outcomes of investment in a knowledge management strategy, and

to measure the impact of such a strategy. The research question is:

“How well does Zack’s (2001) framework represent problem types, and

hence impact on strategies utilised in response to queries encountered in

a public-sector call-centre environment?”

More specifically, the study aims to:

(1) Develop a model that defines the interrelationships between customer queries

and call-centre knowledge management strategies in the provision of

subsequent responses;

(2) Suggest knowledge management strategies to improve appropriate knowledge

re-use within call centres;

(3) Focus discussion and promote constructive interaction for developing a

sophisticated understanding of the call-centre environment, specifically

within the public sector;

(4) Explore the characteristics and implications of systems-based initiatives and

knowledge management interventions that may impact upon a call centre’s

ability to realise service quality benefits;

Introduction

6

(5) Inform academic research directions in call-centre management; and

(6) Provide practical advice to call-centre management.

The examination of two Queensland Government call centres through a knowledge

management lens has the potential to provide guidance in these pursuits. The models

and descriptions in this thesis represent a better understanding of the knowledge-

intensive environment characteristic of a call centre.

1.4. Limitations of the Study

a) The study was limited to two Queensland public-sector call centres to reduce

the variables impacting on the findings. This has, however, reduced the

capacity to confidently extrapolate or generalise to other advice-giving

environments and to service delivery activities outside the Queensland public

sector.

b) Problems of equivocality have not been encountered in the study, so no

claims are able to be made with respect to their place in the framework. This

also limits the ability to claim a highly coherent alignment of Zack’s (2001)

taxonomy with behaviours in the public-sector call centres.

c) The call–encoding methodology has been based on work done by Pentland

(1991) in IT helpdesk environments, where he develops a grammar to analyse

problem-resolution “moves”. In this study, recorded calls have been double-

coded by two different people with a high degree of correlation. Live calls

were not recorded, and hence have been unable to be coded a second time.

This leads to an inherent weakness due to the inability to verify the reliability

of these coded, but not recorded, calls.

1.5. Outline of Thesis Structure

This thesis has seven chapters. The first chapter articulates the processes used to

cover the research question in a scholarly manner. Chapter Two further delves into

the motivation for this study and reviews the relevant literature to provide the

foundation for the further enhancement of aforementioned theories. The chapter

looks at the literature available to assist decision makers in the pre-assessment and

post-implementation evaluation of knowledge management strategies aimed at

improving the capability to provide advice through the burgeoning incidence of call

Introduction

7

centres. The found literature points to a model which incorporates the needs and

quality perceptions of clients (callers), enables analysis of problems raised by callers

via a taxonomy developed by Zack (2001), focuses on the benefits of knowledge

reuse as a strategy influenced by Markus (2001), and then uses determinants of

quality published by Brogowicz et al. (1990) to assess the quality of advice (and

hence improvements based on the selected strategies). Such a model is underpinned

by the assumption that knowledge is the currency of successful organisations, and

that their leaders are able to appreciate the underlying principles and objectives of

knowledge management.

A case study methodology has been chosen in order to analyse the activities of the

call centres. The rationale for this is argued in Chapter Three. Yin’s (1994) case

study method underpins the robustness of the developed models. Although a

substantial amount of quantitative data is available and used, the multi-method

approach suggested by Gable (1994) is not employed. This is due to the fact that the

data is used to develop the model in a qualitative way with greater reliance placed on

patterns, observations, and rich context rather than a detailed statistical examination.

The stepwise process involved in research question definition in the case

methodology, selection of cases and data gathering/analysis techniques, data

collection, analysis, and report preparation has been enhanced by the application of

the iterative methodology suggested by Carroll, Dawson and Swatman (1998). Each

cycle better informs the conceptual framework underpinning the study and provides a

better starting point for a further investigation of the available data.

Chapter Four contains an overview of the organisational structure, tasks performed,

and the related business drivers for the call centres being studied. The comparison

and contrasts able to be derived indicate differing philosophies and strategies used,

with one call centre effectively being divided into teams of teams in order to develop

expertise in specialist areas of operation, while the other develops generalists and

enhances their ability to locate tier-2 expertise, as defined by access to specialist

knowledge primarily of a scientific nature. It also highlights variation in the use of

technology to achieve organisational business goals. Interactive Voice Response

(IVR) systems, a dedicated single source database of business information, and

systems to collect call metadata each exist in one call centre but not the other. This

notwithstanding the performance indicators remain predominantly those reflected in

Introduction

8

the literature to measure generic outputs such as time-in-queue, time taken to

complete, and abandoned calls (Anton & Gustin, 2000). An assumption underpinning

call-centre operation is that referral to tier-2 expertise is a more expensive activity

than reliance on a local call-centre resolution. Self-sufficiency of call-centre

operations (or the degree to which calls are resolved in the call centre) is a goal of

each department. Knowledge strategies used to support this are consistent with the

nature of calls being made. The policies of the agencies do not always reinforce the

goals stated by management during interviews, however, particularly when conflicts

between efficiency and effectiveness arise.

Since multiple sources of data have been considered for this study, specific

frameworks described in Chapter Five are used to gather and analyse data sources

such as interview reports, brochures, templates, recorded calls, and real-time calls for

each of the organisations. The Four Problem Type Model (Zack, 2001) is used to

determine the relevant resolution strategies. Characteristics concerning quality are

extracted from the work of Brogowicz et al. (1990). Response strategies, based on

the work of Markus (2001) and Zack (2001) are used to assess the resolution of

queries through knowledge reuse. The Query-Response Cycle (G Timbrell &

Shepperd, 2002) provides an appropriate framework to link problem, resolution, and

quality in the call-centre environment. Direct observation is able to be employed as

an analytical tool because of the regimented and static nature of these call centres

(with each having single traditional work sites which were easily accessible).

Although the call centres are compared and contrasted, the sum total of the recorded

and live calls is used to produce a more comprehensive coverage of the problem

types described by Zack (2001).

Chapter Six is used to illustrate the patterns and responses obtained from the data

collected in the call-centres. Interviews with management and team leaders, direct

observations of work practices, and coded calls are analysed using NVivo qualitative

research analysis software to distil trends and learnings from a range of perspectives.

A pragmatic approach to call-centre strategies is discussed in Chapter Seven. The

alignment of organisational goals with the role of the call centre and knowledge re-

use strategies employed contributes to the success of the call centre. The dilemma of

who defines the quality of advice (caller or expert) is resolved through the

introduction of the concept of service value. This uses the service quality paradigm

Introduction

9

of Brogowicz et al. (1990) to incorporate the issues of customer satisfaction and

service quality determinants of personal quality, technical competence, and the

functional service offering of the agency.

The final chapter, Chapter Eight, addresses the implications of the thesis’s findings

for government call centres. It outlines the future extension of this research to a

more general organisational environment. Concluding remarks provide advice to

call-centre management which arise from this study.

Context of the Study and Review of Literature

10

Chapter 2


2.1. Introduction

In this chapter, a range of academic literature is investigated in a structured attempt

to provide a robust foundation for the qualitative data gathering employed in the case

studies. Alignment of issues with research interests is achieved through an

understanding of the contemporary public-sector context and the predominant

organisational drivers in Section 2.2. Since call centres have been identified as the

suitable response to efficiency drivers (Milner, 2000), Section 2.3.1 focuses on call-

centre performance. The measurement of advice using quality principles is

investigated in Section 2.3.2, and revisited in Sections 2.3.7 and 2.3.8, where

improvement strategies are introduced based on knowledge management principles.

Since the underpinning assumption is that better problem-solving capability will

drive improved performance through knowledge reuse, knowledge management,

generation, capture, and reuse are the foci of Sections 2.3.3 through 2.3.6. The

strategy employed in this analysis is to identify the significant concepts involved in

the improvement of advice provided by public-sector call centres, and to develop a

framework to converge the work of thought leaders.

2.2. Academic Literature

2.2.1 Call Centres and Performance

Call centres have become the major access point for customer interaction with an

organisation. Many organisations are transforming traditional backroom call centres

through rejuvenated technology to become the primary front-line interface of the

enterprise (Anton, 2000). The major business driver for call-centre expansion,

particularly within the public sector, is cost. In 1999, Riggs and Thyfault estimated

the cost of handling a phone call to be $US3 compared to $US25 for responding to a

written letter, the traditional communication channel for government bureaucracies.

Tsoukas and Vladimirou (2001) confirm this in their study of a call centre in Greece

where speed-to-answer contributes to both the lowering of operational cost and the

delivery of high quality service. They identified the importance of accumulated


11

experience and knowledge of other agents to the organisation, and highlighted the

social relations at work compared to those absent in technical systems. Compelling

reasons for improved call-centre service cited in literature include loyalty, attracting

new business, and cost cutting, as well as demand from customers (Wilde, 2000).

This accumulated experience and knowledge requirement of call centres is also

predicted by Gartner researchers (Close, 2003). Here, knowledge management

strategy is proposed as a significant determinant of service performance in a call

centre. The predominance of codification strategies (scripting, issues databases, and

document repositories) employed in call centres to improve efficiency satisfies

criteria determined by Hansen, Nohria, & Tierney (1999): that to be competitive,

organisations need to become expert in either codification or personalisation. A

knowledge strategy that includes “The informal memory system (both individual and

collective) which has gradually built over time, consisting of the individual stocks of

experience held by each operator, and by stories shared in their community”

(Tsoukas & Vladimirou, 2001, p. 985) will provide greater capability for agents to

give improved quality advice to customers. The degree to which this heuristic

contributes knowledge to call-centre performance is significant, but is not able to be

quantified by the traditional statistical reporting.

Humphrey (1989) refers to the need to be able to measure activities in order to

develop and improve them. An awareness by executives of the impact of

performance management systems on multi-functional teams in the area of service

delivery is minimal (Meyer, 2004). Significant investment is being made in

technologies within call centres which lead to issues of Return On Investment, with

such reporting requiring some form of benchmark to justify the benefit of the outlay

through achievement of outcomes. This is particularly the case in the public sector,

which is under continuous pressure to improve performance in a climate of

decreasing funds. While responsiveness is important, it is not a sufficient condition

to achieve customer satisfaction. The measures predominantly published on call-

centre performance relate to time-in-queue (on hold), time in conversation, calls per

hour, and other similar quantitative measures (Anton & Gustin, 2000). Duder and

Rosenwein (2001), for example, suggest a formula to calculate the number of call-

centre agents to achieve zero abandonment rates. Their argument is that by reducing

time-in-queue, abandonment will fall and the quality of service will improve. The


12

US Internal Revenue Service employed a similar operations research approach where

the main objective was to allocate resources in such a way as to minimise the number

of customers who receive a busy signal or who must hold for a long time when they

call for information (Harris, Hoffman & Saunders, 1987). In addition, QANTAS

employed a model utilising queuing and integer linear programming to reduce

waiting time and evaluate the relationships of staff size to waiting time and service

time (Gaballa & Pearce, 1979). Yet, in all call-centre interactions, at the point the

caller makes initial contact, the primary purpose of its existence (i.e., taking of

queries and offering of advice in response) has not been fulfilled. The perceptions

being formed by clients in their early interactions with the call centre will impact on

their impression of the organisation. This serves as an indicator that attention needs

to be paid to the responsiveness attribute of service quality.

2.2.2 Measuring the Quality of Advice

Further measures, particularly with respect to customer satisfaction, are obtained

from exit surveys of call-centre clients. Many factors affect the perception a caller

may have of the quality of advice received, and hence impact on how exit

questionnaires are scored. In some instances such advice may not be palatable to the

caller, but the public servant’s creed of “without fear or favour” requires that such

advice be given to as high a standard as possible. Jungermann (1999) suggests that

there is an implied competence in a receiver to take advice. Hence, a caller who

requires a pragmatic response but receives a political justification would at best be

confused by such a reply. Without the benefit of a face-to-face meeting, which

allows a broader range of interpersonal communications cues, call-centre agents have

to rely on much more subtle changes in intonation or use of qualified paraphrasing.

Harvey et al. (2000) suggest people are better at assessing the quality of advice given

than applying that advice. Many a parent laments the waywardness of their children,

who in turn have no doubt as to the quality of advice from the parent; rather the

inclination is to express their individualism in a way which directly challenges this

advice. The role of a public-sector call-centre agent is to provide quality advice.

Whether the caller acts on this advice is another matter.

The conflict arising from existing literature is that the reliability of tools such as exit

surveys needs to be tested against a known benchmark. While these traditional

measures have served the management of efficiency, it has been at the expense of the


13

development of effectiveness measures. In the call-centre environment, effectiveness

can be improved by measures relating to quality of advice. The ability to quantify the

quality of advice given to callers is a necessary precursor to the analysis of calls.

Given that a large volume of calls is to be analysed in this study, this process needs

to be efficient and repeatable. The analysis also demands a different perspective

such as employing a knowledge view of call-centre activities that will progress the

goal of applying appropriate effectiveness measures.

Call-centre operational staff have relatively junior positions and provide advice

based on both formal and informal knowledge acquisition processes (Kjellerup,

2004). Most common of the formal processes are training (which is provided to

varying degrees), and use of information databases, which more often than not

contain static information. Zack (2001) indicates this is quite an adequate response

to queries on quantifiable data where there is an unambiguous response easily found

by the advisor (e.g., “By what date do I have to have submitted a tax return?”), but

not so effective in qualitative areas which may be complex, uncertain, ambiguous, or

equivocal (e.g., “Should I sue a trader who has provided me with defective goods?”).

Goh (2002) suggests that a well-developed knowledge management strategy is

critical to business and therefore public-sector success. However, prior to investing

in such a strategy the question, “How do various knowledge management strategies

affect the quality of advice given by public-sector call-centre staff?” needs to be

answered. By refining models proposed by contemporary knowledge management

practitioners and others with expertise in the measurement of quality of advice, a

framework will be developed to provide confidence in realising organisational

benefit from investment in knowledge reuse. Knowledge reuse is a specific area of

interest in knowledge management which Markus (2001) suggests provides

successful outcomes when allied to information needs.

2.2.3 Defining Knowledge Management

In their critique of knowledge management, Baker and Badamshina (2002) suggest

that the domain’s definitions tend to be so broad and vague as to have little meaning,

and in other cases too narrow in focussing on mechanistic and sequential process

directed at knowledge creation, capture, sharing, and (re)use. Since Polanyi (1967)

argued the two aspects of “knowing what” and “knowing how” were present in any

instance of a person’s knowledge, cognitive psychologists have developed a view of


14

knowledge creation that involves a continuous interaction between implicit

knowledge (acquired without a conscious attempt to do so) and explicit knowledge

(as occurs when conscious learning strategies are applied) (Nonaka, 1994; von

Krogh, Nonaka & Aben, 2001; Nonaka & Takeuchi, 1995). Knowledge can also be

viewed as a process of simultaneously knowing and acting and focuses on the

application of expertise (Brown & Duguid, 2000). Another view defines knowledge

as an object that can be viewed as capable of being stored and manipulated (Markus,

2001). Finally, knowledge can be viewed as a capability with the potential for

influencing future action (Carlsson, 2001).

Given the variety of interpretations of the concept of knowledge and its application,

it is expected that a similar variety of explanations exist for knowledge management.

Alavi and Leidner (2001) identify differing implications for knowledge management

based on the perspective held on knowledge. Knowledge viewed as an object implies

knowledge management should focus on building and managing knowledge stocks

and using information systems to store, gather and transfer it to users. Knowledge as

a process requires knowledge management to focus on knowledge flow - the

processes of creation, sharing, and distribution of knowledge. Knowledge as a

capability means knowledge management strategies build on core competencies, and

create intellectual capital. The predominant position taken in this study is that

knowledge is a capability since it actually addresses the link between knowledge,

knowledge management and organisational performance.

The Standards Australia AS 5037(Int)-2003: Knowledge Management Standard

(2003) states that :

“[knowledge management] Is a multi-disciplined approach to achieving

organisational objectives by making best use of knowledge. It involves the

design, review and implementation of both social and technological processes

to improve the application of knowledge, in the collective interest of

stakeholders.” (2003, p. 1)

This definition is consistent with the previously discussed material. The extension of

these concepts and the subject of many texts is the development of the tools and

processes that allow an organisation to systematically create, capture, share, and

leverage this knowledge. Each opinion leader has a comment which challenges or

supports the view that knowledge is able to be managed, and puts into perspective


15

the role of knowledge in the performance of an organisation. A consistent theme is

the human aspect – that a person needs to take stimuli from a range of sources and

process them in such a way as to improve or add value to what was previously

known, or to translate this into action. The better an organisation exploits this

capability, the better it is able to meet its business objectives (Choo, 1996).

In their seminal work in knowledge management, Davenport and Prusak (1998)

define “knowledge” in terms of the work of people’s minds trying to incorporate

experiences, values, insight and context to a new experience.

“Knowledge is a fluid mix of framed experience, values, contextual

information, and expert insight that provides a framework for

evaluating and incorporating new experiences and information. It

originates and is applied in the minds of “knowers”. In organisations, it

often becomes embedded not only in documents or repositories but

also in organisational routines, processes, practices and norms.”

(Davenport & Prusak, 1998, p. 5)

Davenport and Prusak (1998) then go on to show how knowledge can be used to

improve organisational performance. By annotating features of a project carried out

by British Petroleum, they list the underpinning principles of knowledge

management as including concepts of sharing, trust, technology as an enabler,

rewards, measurement, and creativity. This pragmatic approach avoids the

philosophical debates which arise from taking an abstract view of knowledge (and

subsequent investigation of the meaning of truth) and identifies the role of

management to ensure knowledge assets are used to produce organisational benefits.

Hence, the knowledge that is peculiar to the organisational context is to be exploited

by (knowledge) managers to improve performance, action, and outcomes (Pentland,

1991). Thompson and Walsham (2004) analyse Blacker’s (1995) five images of

knowledge, and conclude that most organisational change programs seldom address

these contextual components simultaneously. The components are defined as

embrained – or latent mental potential; embodied – or historically developed filters

and routines; encultured – or convergent expectations about intentions of others;

embedded – or organisational alignments; and encoded – or explicit symbolic forms.

The degree of simultaneous consideration of all these factors by knowledge

managers provides an organisational knowledge profile that is unique and distinctive


16

to a firm and provides its relative competitive advantage (Tsoukas & Vladimirou,

2001). The flow of knowledge within a firm is a function of the encultured

contextual component and is determined by a knowledge market (Davenport &

Prusak, 1998), which, like the well-studied commodity markets in economics,

determines the value of knowledge. The management of this market determines the

behaviours relating to knowledge strategies.

Competitive advantage of organisations in the twenty-first century will be obtained

through creating and sharing knowledge (Senge, 1997). The ability to integrate

theory, core staff capabilities, and institutional learning through best practice will

ensure sustainability of the organisation. Leonard (1995) provides a knowledge-

based model of core capability improvement for an organisation. She identifies

capability gap (1995, p. 138) as that gap between what the company knows and what

it needs to know in order to be productive and competitive. She refers to employee

knowledge and skill, technical systems, managerial systems, and values and norms of

an organisation as dimensions of core technological capabilities that provide strategic

advantage. Deficits may be met by in-house development of knowledge, brought in

via consultants/technology partners or other strategies such as takeover/strategic

partnering with competent firms. The concept of stolen ideas is also considered as a

legitimate process for capability enhancement. Not only do the ideas have to be

replicated, but improved functionality or creative application is required in order to

exploit the innovation that results from enhanced organisational capability. Sony

Corporation provides an example of the way it successfully gained market share

from other games machine manufacturers by adding games controllers that vibrate.2

This is the corollary to obtaining competitive advantage through the inimitability of

products or services. Opportunities for the public sector to exploit intellectual

property of others are restricted, since any behaviour seen as being remotely

dishonest causes considerable ethical and political concerns. Given the nature of

public sector organisations, options for bridging capability gaps are primarily limited

to internal capability improvement. They are focussed on three components of a

knowledge lifecycle identified by Wiig (1999): namely, that of originate/create

2 The Rumble Pack case (http://news.bbc.co.uk/2/hi/business/4387045.stm) refers to a plagiarism row between Sony and Immersion, a small, California-based developer of digital touch technologies, over game controllers that vibrate.


17

knowledge or generation; capture/acquire knowledge; and utilisation, which

combines the remaining phases of transform/organise knowledge, deploy/access

knowledge, and apply knowledge. Figure 1 illustrates the relationships and

organisational knowledge transitions among these phases (Rus, Lindvall, & Sinha,

2001).

Figure 1 - Organizational Knowledge Evolution Cycle (Rus et al., 2001, p. 17)

2.2.4 Knowledge Generation

Knowledge generation is taken as the starting point in the improvement of

organisational capability (Davenport & Prusak, 1998). In keeping with the

knowledge market metaphor, the availability of capable staff is the significant

limiting resource in the public sector. Public policy development requires

considerable research and innovative thinking, particularly in response to the

challenge of winning back the trust of citizens (Davis & Weller, 2001). Selecting

existing staff for the purpose of innovation, providing appropriate challenges, and

recognising and rewarding them for their work are widely utilised strategies in

development of individual capabilities (Peters & Austin, 1985). Support for internal

staff selection exists in one of the presuppositions underpinning Neuro-Linguistic

Programming (NLP)-oriented training that suggests people are ultimately able to

draw on their own capabilities to address any situation, provided they are aware of

their capability and know how to use it (Bradbury, 2000). Leonard (1995) also

addresses this knowledge acquisition, and introduces the concept of prototyping and

experimentation. She presents a research and development model consistent with

halla

This figure is not available online. Please consult the hardcopy thesis available from the QUT Library


18

earlier themes, indicating the need for knowledge generation in order to maintain a

sustainable organisation. The tying of experimentation to corporate resilience is

tendered by Leonard (1995) as an alternative to long-term strategic planning that

may be incapable of coping with uncertain (future) business environments. Such

experiences produce actionable knowledge (from both success which motivates to

further develop and failure which provides an opportunity to examine alternate

strategies), which in turn contributes to the core capability and hence sustainability of

the organisation.

Two apparent dilemmas are raised by the literature in regard to knowledge

generation. The first is that if there is too much comfort among staff, there is no

incentive to innovate. One strategy able to be employed here is to intentionally

“…try to evoke a sense of crisis among organisational members by proposing

challenging goals” (Nonaka & Takeuchi, 1995, p. 28) The introduction of creative

tension through challenges, either perceived or actual, or by formation of teams

composed of staff with different skills and values can create the “creative chaos”

referred to by Nonaka and Takeuchi (1995). The other apparent quandary is that

success is often the greatest enemy of innovation. This is a subset of the initial

dilemma discussed, with the point being that success often creates comfort or a sense

that past practices will continue to be successful into the future. This is a lesson IBM

has learned, although for some time grave concern was held for the company that had

defined technology in the workplace for much of the century. During the 1980s

IBM's mainframes suffered from competition from more flexible UNIX servers. The

company responded by channelling massive resources into attempting to stabilise its

market position with its traditional products and services. It sustained multi-billion

dollar losses. Under new leadership, rather than promoting any particular product or

content, IBM focused on end-user services such as assisting companies to transform

themselves into secure Internet-based trading communities ("Blue is the colour.,"

1998).

2.2.5 Knowledge Capture

Knowledge capture is the second key systematic process identified as a critical

organisational capability. Differing from the generation of knowledge, technology is

seen to provide a tool to codify knowledge in such a way that it becomes useful in

other circumstances (Davenport & Prusak, 1998). Dixon (2000) introduces the


19

concept of translation or modification where the same issue is likely to be repeated,

but in different localities and under differing circumstances. Hansen et al. (1999)

assert that a firm’s knowledge strategy should reflect its competitive strategy; it

should have an 80-20 mix of either codification or personalisation methods,

specialising in only one of the areas, but without ignoring the other. Codification

strategies, which rely heavily on investment in IT and reward for people contributing

to the document databases, are more aligned with maintenance of high revenues

through reuse of the knowledge asset. High profit goals driven by a strategy of

personalisation requires the development of person-to-person networks and reward

for direct interaction with others. Simply capturing the information does not,

however, make it useful. Thompson and Walsham (2004) support an integrated

strategy, but extend the model to include embrained, embodied, encultured,

embedded and encoded contextual components. Bryar (2001) raises issues of

information overload, but suggests that the combination of a well-defined taxonomy

and associated technology at least ensure information can be relocated if correctly

classified and tagged for metadata searching. Attempting to excel in both

personalisation and codification strategies increases risk of failure, while ignoring

either strategy has a similar impact (Hansen et al., 1999).

2.2.6 Knowledge Reuse

The final aspect identified in the literature refers to the sharing of knowledge and its

socialisation: the process whereby tacit knowledge is able to be transferred between

individuals and organisational groups through shared experience, space, and time.

This important value-creating process allows the organisation to obtain competitive

advantage through better decisions or quicker time to market (Handzic &

Chaimungkalanont, 2004). Knowledge sharing is referred to by various terms

including “knowledge transfer” (Davenport & Prusak, 1998; Dixon, 2000; Goh,

2002; G. Timbrell, Andrews, & Gable, 2001), and “knowledge reuse” (Kucza, 2001;

Markus, 2001). Davenport and Prusak (1998) suggest organisational value is created

by this knowledge transfer when the two actions of transmission (sending or

presenting knowledge to a potential recipient) and absorption (by that person or

group) improves its ability to perform new actions, innovate, or change behaviours.

Value is obtained through the efficient reuse of knowledge in the execution of

repetitive tasks or solving similar problems (Majchrzak, Neece, & Cooper, 2001).


20

Markus (2001) identifies four knowledge reuse situations based on the purpose of the

knowledge and the role of the seeker and refers to them as shared-work producers,

shared-work practitioners, enterprise-seeking novices, and secondary data miners.

This is validated by Timbrell et al. (2003) with the identification of a further group

identified as “primary data miners”. Given that call-centre agents predominantly act

as individuals, Markus’s (2001) model provides a framework for focussing further

study on knowledge sharing behaviours by government call-centre operators, in an

attempt to improve contextual problem solving both at the individual and team

levels. Figure 2 illustrates the role of shared problem solving as a core capability,

necessary to address the increasing complexity from globalisation and the

proliferation of formal education (Leonard, 1995). Provided an appropriate

repository exists, benefits accruing from knowledge reuse in a call centre are

reflected in:

• Reduction in search time through the application of informed search

strategies;

• Consistency of responses through alignment of past scenarios with current

queries; and

• Continuous improvement in quality of response through opportunity to reflect

on knowledge sources, resolution strategies and robustness of advice

(Leonard, 1995, p. 60).

Figure 2 - Capability-Creating Activities: Shared Problem Solving (Leonard, 1995, p. 60)

PRESENT

INTERNAL

FUTURE

Shared Problem Solving

Improving Knowledge

Experimenting

Implementing and Integrating

EXTERNAL

Core

Capabilities


21

Shared problem solving is not the recognised method of operation of Queensland

Government call centres where a traditional management paradigm requires

individual agents to take responsibility for their client callers. However, in order to

improve performance through knowledge reuse, the contribution by individual agents

impacts on the overall problem solving performance of the call centre. All of the

reuse situations defined by Markus (2001) and Timbrell, Nelson, and Jewels (2003)

are applicable to this study of knowledge reuse in call centres. The following

analysis suggests that shared-work producers are able to improve efficiency through

codification; shared-work practitioners are significant players in current knowledge

reuse; expert-seeking novices (both callers and less experienced agents) need to

develop a range of knowledge sources and search strategies; while primary and

secondary data miners are more likely to perform data mining activities in an attempt

to identify strategic activities to improve efficiency and effectiveness of call-centre

operations. These differing knowledge transfer strategies are addressed by role which

for some individuals may change depending on the type of task allocated at the time.

Shared-work producers are categorised as a homogeneous team producing

knowledge for their own later use. They also have a need to know the “what,”

“how,” and “why” of previous decisions. Although primarily gathered for personal

use, benefits would accrue to all agents if the embedded knowledge were accessible

to all call-centre staff, since they frequently keep good (personal) records as a by-

product of their work, but often omit the rich context surrounding the incident, which

is a necessary component for extracting learnings for future projects. Anton (2000)

refers to the proportion of customer intelligence available to call-centre staff as

strongly favouring static information.3 Information is collected by call-centre agents

in their informal (personal) working notes, by filling in forms (both paper- and

electronic-based), and in formal reports resulting from research required to answer

the caller’s query. Utilisation of this range of data collection methods is not

universally adhered to. In terms of efficiency, performance measures reduce the

motivation to either codify or share such information in a formal way, thus reducing

3 Modern technology is more able to deliver identification details of a caller to the desktop, but is expensive to implement and maintain. At the lowest level, intelligence provides the caller’s phone number. If linked to customer management systems (CMS), further information such as name, address, preferred products, previous call history, etc. are able to be accessed.


22

the benefit from knowledge reuse and the consequential increased service quality

opportunity (Anton & Gustin, 2000).

Shared-work practitioners do similar work, but may be employed in different

environments. They acquire new knowledge from external sources and have a need

to know how and why a particular procedure works. Several roles within the call

centre reflect this reuse situation, since the establishment of call centres addresses the

modern organisation’s need to balance flexibility and stability with increased

openness to customers’ needs and the protection of core practices and routines

(Arzbacher, Holtgrewe & Kerst, 2002). The static information databases, pamphlets,

and other cues used by agents are developed by experts from outside the call centre,

and may not have the call-centre agents as their target audience.

Part of the expertise of the shared-work practitioner is the ability to locate material

relevant to the query in a timely way. For this reason, initial training for call-centre

staff focuses on the ability to locate specialist knowledge, either through codification

processes (use of Yellow Pages-type directories) or personalisation processes

(attendance at functional work group meetings, for example) (Hansen et al., 1999). A

number of call centres have a tiered structure where reference to tier-2 staff will

facilitate the application of their specialist knowledge to the query. These shared-

work practitioners are able to improve the efficiency of the call centre by sharing

their learnings (in the categories of close knowledge distance4 and most useful) with

other specialists, as well as other agents who may be able to exploit it at a later time.

Alternatively, functional specialists within the call centre may have calls directed to

them by integrated voice response (IVR) technology. In all cases, the shared-work

practitioners take calls referred by call-centre staff and carry out more detailed

activities, which often relates to problem resolution. It is likely that as performance

pressures impose on call-centre staff, more referrals are made to these practitioners.

Business strategies for the use of call centres vary from firm to firm, and will range

from clearing house for routine transactional processes (e.g., renewal of licenses and

4 Knowledge distance is a term used by Markus (2001) and Timbrell et al. (2003) when referring to the gap between what a person knows and what they need to know in order to perform some task or solve some problem. In this instance, if a person makes notes about a particular event, they know the context in which the information was collected and hence will know how to apply it in other similar situations.


23

permits), to complex problem-solving roles such as the use of an IT helpdesk. An

appropriately selected knowledge management initiative tied to the firm’s business

strategy will improve the ability of front-line staff to manage a wider variety calls,

thus leading to a more efficient call-centre operation (Zack, 1999).

Expert-seeking novices seek answers to questions not previously encountered by

them, and acquire knowledge through advice from other experts. Often operating

without a capability to identify the relevance of contextual information, expert-

seeking novices may have difficulty defining the problem at hand, and may rely on

knowledge designed for novice audiences, or on access to specialists. This role

category equally applies to callers and novices within call centres. In technical areas,

manuals are seldom written for the novice, which leads users with a problem (callers)

to seek assistance from a service provider. The primary reason for the existence of

many “helpdesk”-based call centres is to provide advice and assistance. In his study

of IT helpdesk software support actions, Pentland (1991) distinguishes between a

problem and a question. He suggests questions may well be better answered by

reference to available information sources such as program documentation. This is a

problem defined by Zack (2001) as one of uncertainty. In many cases, the caller

probably has easy access to such information, but will call either due to lack of

confidence in locating the required information, or inability to interpret it in a

meaningful way. Problems relating to software support, which Zack (1999) defines

as complex, ambiguous, or equivocal, are the domain of the helpdesk, where experts

can apply diagnostic processes in order to resolve the issue causing concern. The

knowledge distance in these instances is much larger than when discussions relate to

operating instructions (since the call-centre agents are technical experts and the

callers primarily applications users), and is complicated by the reliability of the

information supplied by the caller, and the understanding of the resolution

instructions. In their study of a Greek call centre, Tsoukas and Vladimirou (2001)

also discovered the difficulty of communication between customers and operators

due to the lack of sophistication of the caller.

Consideration must also be made for the expert-seeking novices who exist within the

staff of the call centre. Due to the turnover in staff in call centres, there is an

abundance of these novices who have a knowledge of the local context, but are

unlikely to know which aspects of this context are important (Tuten & Neidermeyer,


24

2004). Knowledge of the operations is gained on the job through degrees of training,

but more likely through informal networks of perceived experts, with that decision

being based as much on personality as actual expertise. Strategies based on advanced

technology are often cited as an effective option for call centres, and are reflected by

comments such as:

“One major component of the Government Call-Centre project is to put in

place the necessary hardware, software, and networking infrastructure to

support the various call centres…The customer service officer is able to

access back-office host computer systems and databases.” (Milner, 2000, p.

109)

Milner (2000) goes on to predict that call centres and the supporting infrastructure

will be at the forefront of public sector strategies within the next ten years, so as to

deliver client-focussed modes of operation. The benefits of this investment are not

well identified in research literature on call-centre performance, however, although

Anton (2000) predicts that the call centre of the future will integrate the Internet with

calls to allow the caller to see the agent servicing them. This technology will lead to

reduced talk times and time to answer, as well as service quality improvements in

accuracy of information (both from caller and agent).

Secondary data miners attempt to develop new knowledge though analysis of

records. This study, which accesses, analyses, and reflects on documents created in

the call centres among other things, aims to develop new knowledge able to be

applied in more general advice-giving situations. It thus satisfies the secondary data

miner typology. In traditional call-centre environments, analysis has generally

referred to the efficiency measures of average speed of answer, time-in-queue,

resolution averages, total calls, and a range of other quantitative outputs which are

generally easily extracted from call-centre management software. Timbrell et al.

(2003) refer to call-centre managers as primary data miners because of the close

“proximity to knowledge context” and that they reuse this knowledge to discover

processes to improve performance or remove activities that may be in conflict with

business goals. Feinberg, Kim, Hokama, de Ruyter & Keen (2000) suggest that

interest in these measures may simply be an artefact of the technology since they

become important because they can be measured. Their research shows that fast

access to the agent and successful resolution of the caller’s problem are the essential


25

elements of an effective call-centre operation. Milner (2000) extends this concept by

suggesting that analysis of call-centre operations needs to provide senior

management with awareness and knowledge of client needs and the capability of the

organisation to deliver on these. This suggests a degree of sophistication above that

of mere qualitative analysis of traditional sets of output metrics, and may well

provide the competitive advantage achieved through call-centre agents becoming

“truly knowledge workers” (Anton & Gustin, 2000). The ability to reuse knowledge

in a call centre lies in its capacity to connect “want-to-knowers” efficiently with

“knowers” (or at least access to documents and a context to make sense of the

information contained within them). Contemporary call-centre managers are

responding to the changing business environment5 through their desire to develop a

capability in their agents to recall reasons for decisions (or the “when, what, how,

and why”) – a classification referred to as rationale knowledge by Markus (2001, p.

62). It is reasonable to assume significant benefits will accrue to call centres able to

put such knowledge management practices in place.

2.2.7 Defining Quality of Advice

The nature of a call centre is that it provides relative anonymity to the caller.6 This

may influence the quality of advice given since environmental factors affect the

quality of advice given (Rogers, Hassell, Noyce & Harris, 1998). Rogers et al. (1998)

found that lower socio-economic groups received a lower standard of advice in

community pharmacies than those of higher status groups. In a call-centre

environment, unless a caller provides details to identify characteristics which may

introduce bias, a caller’s identity should not influence the advisor. However,

telecommunications innovations are able to deliver significant profile information to

a call-centre desktop which may include caller identification (from the

telecommunications carrier) and other profile information from internal systems

(e.g., if the caller telephone number is linked to company transaction databases)

(Anton, 2000). The agent’s advice can then be tailored to fit the caller’s profile,

5 Shorter product life cycle, global competition, and superior service quality are some of the challenges faced by organisations in their quest to sustain a competitive position in the marketplace (Zairi, 2002). 6 A caller may withhold identifying information and block technical triggers (such as call identification), but may be required to provide unique qualifiers that allow identification (such as Medicare number), if material benefits are being negotiated.


26

which is advantageous, but may also lead to degradation in quality if socio-economic

bias is introduced.

A determination needs to be made as to who defines the quality of the advice given.

Callers who receive advice contrary to what they see as acceptable are likely to seek

a second opinion elsewhere. This is not at issue in this study, since the public sector

organisations studied are effectively monopolistic. Its significance is rather that a

reliable measure of quality is available and that as strategies are put in place to

improve the advice, impacts are able to be reported. A simple resolution to this

situation is that an expert is used to rate the advice being given, which is a trivial task

in terms of monitoring, but a complex one in terms of resource allocation and impact

on employees being monitored. Jungermann (1999) provides a model for advice-

giving. He asserts that advisors attempt to match options to the clients, rather than

provide a detailed analysis of issues being considered. This may be in order to satisfy

the person who is requesting the information, or it may result from the advisor

prejudging a decision about the competence of the caller to understand the range and

complexity of issues being considered. Jungermann (1999) also analyses the take-up

of advice based on the opinion-related and personality-related attributes of both the

giver and receiver, and how they relate to degrees of expertise, credibility, and

confidence. Harvey et al. (2000) provide some hope for using the receiver of advice

as a reliable source of information able to contribute to quality measures, in that they

conclude that a receiver of information is better at assessing the quality of advice

than using it. If it were possible to obtain post-advice information independent of the

need of the receiver to use that advice, such data may provide a guide to its quality.

Brogowicz et al. (1990) provide a service quality model synthesised from two

sources referred to as “The Nordic School” and ”The North American School”. The

model is multi-dimensional and separates technical quality (staff, systems, etc.) from

function quality (policies and procedural issues) to provide a service gap instance.

The smaller this gap, the better the perceived quality of service. The determinants,

able to be empirically defined, include “tangibles, reliability, responsiveness,

assurance and empathy” (Brogowicz et al., 1990, p. 33). Courtesy is another term

used to cover these interpersonal interactions (Tsoukas & Vladimirou, 2001).

Pragmatic thinking towards exceeding customer expectations in marketing generally

supports the strong positive relationship between customer and service provider


27

(Estelami & Maeyer, 2002). Brogowicz et al. (1990) suggest competitive advantage

can be obtained through a strategy aimed at decreasing customer expectation, rather

than increasing quality of service. This is an interesting approach and may have been

used for a long time in the public sector, given the degree of esteem with which it is

held. The aim of this study is, however, to identify a range of strategies able to be

applied in an advice-giving environment with a high degree of confidence that they

will deliver performance improvement. Unfortunately, the actual levels of quality as

determined by accuracy, timeliness, and appropriateness tend to become enmeshed in

the perceptions of the callers – whether owing to issues of courtesy, empathy, or

expectation of a beneficial result.7 The concept of service value is investigated by

Dedeke (2003). Figure 3 shows the relationship between service quality, service

value, and client satisfaction. Dedeke (2003) establishes that customers may tolerate

poor service in certain circumstances, and be dissatisfied with excellent quality in

others. In areas of complaints management, these apparently irrational outcomes are

highly probable.

Figure 3 - Relationship among service-quality concepts (Dedeke, 2003, p. 280)

7 An aggrieved caller may call looking for an outcome which may not be legally or morally appropriate. In this case, no matter what the advice, the caller will not assess it as being of high quality.

Service Value Behaviours

Service

Quality Satisfaction

Perceived

Costs

Attitudes

Attachments

− Performance quality of outcomes

− Performance quality of processes

− Performance quality of process system

− Sacrifices

− Direct costs

− Inconveniences


28

2.2.8 Knowledge Strategies to Improve Advice-giving

The drive for efficiency in call centres, while delivering better service quality, has

concurrently caused the knowledge required to do work such as claims processing to

be embedded in systems. This requires prescribed patterns of interaction, but

minimises relationships both within and beyond the organisation (Cross, Liedtka &

Weiss, 2005). The focus of such systems changes from distinct roles (as

characterised by staff external to the call centre) to one of process flows (role

anonymity). Routine response networks, which rely on scripted and easily located

knowledge sources, are structured to reduce effort in resolution of repeat service

calls. They are also structured to locate experts for second-tier referrals quickly to

address well-structured problem types having minor degrees of abstraction. Zack

(2001) refers to these problem types as uncertain (i.e., “having insufficient factual

information about the goal, situation or task, and some lack of confidence in the

consequent inferences, estimates, or predictions required” (Zack, 2001, table 1)).

However, since they are unable to instigate required cognitive and affective

processes, such routine response networks become ineffective in problems which are

complex (i.e., with “too many situational elements and relationships to coordinate or

consider simultaneously” (Zack, 2001, table 1)); ambiguous (i.e., showing

“inadequate knowledge about, no explanation for, or understanding of a goal,

situation, or task” (Zack, 2001, table 1)); or equivocal (i.e., providing “multiple

interpretations of a goal, situation, or task” (Zack, 2001, table 1)), since by definition

it is not possible to define a single process flow to address these problem types

(Jonassen, 2000). Cross et al. (2005) conclude that management needs to develop

strategic views on social networking methodologies for the purposes of knowledge

reuse/referral services, which are best suited to their business goals. A modular

response network, consisting of experienced practitioners, provides the capacity to

solve complex problems through re-sequencing and reuse of knowledge acquired in

previous (similar) situations. The defining characteristic of modular responses is their

ability to disaggregate complex problems into smaller ones that are both familiar and

solvable by the experienced network members. Resolution of unstructured, ill-

defined problems (ambiguous and equivocal) is enhanced through the development

of customised response networks that link experts capable of developing innovative

solutions.


29

Core capability to enable innovation is also discussed by Leonard (1995). She

extends her concept of capability (or capability gaps) to the problem-solving arena

and concludes that shared problem solving is a critical competency of a successful

organisation. She warns that core rigidities established in prior successful problem-

solving situations can cause dysfunctional behaviours that hinder innovative

thinking. Leonard (1995) thus recommends the introduction of creative abrasion,

controlled introduction of tension into the workplace, to expand the available

knowledge resource. The management problem that arises from her work is to

balance the benefits accrued from the application of solutions from highly skilled

staff with blockages that come from emotional attachment to their mindsets and

problem-solving biases.

Leonard (1995) addresses the primary purpose of problem solving for innovation; a

broader analysis of problem is required to address all the activities of organisations,

since a knowledge-based view of a firm recognises the multi-dimensional nature of

knowledge. Zack (2001) contends that “Different literatures have addressed only

portions of this knowledge-problem space” (Zack, 2001, p. 17). He suggests

knowledge is an organisation’s most enduring and strategic resource, and presents a

taxonomy of problem types (complexity, uncertainty, equivocality, and ambiguity) to

integrate work done by other researchers. These problem types are explored further

below.

Complexity

An organisation’s response to complex problems, as those with “many interrelated

variables, solutions and methods” (Zack, 2001, p. 18), is determined by its capability,

and includes actions to:

• Develop richer knowledge in order to allow issues to be dealt with as a

familiar single problem; and

• Reduce complexity through decomposition to simplified (solvable/familiar)

components.

Uncertainty

Problems of uncertainty, characterised by the “lack of information or factual

knowledge about current and future states” (Zack, 2001, table 1), are addressed

through:


30

• Acquisition of additional knowledge; or

• Improvement in knowledge and ability to predict, infer, or estimate.

Ambiguity

Ambiguous problems represent “an inability to interpret or make sense of” (Zack,

2001, table 1) a particular situation. Tsoukas and Vladimirou (2001) suggest

ambiguous problems occur when callers are unclear of what they want and therefore

often provide redundant information. Skilled operators guide callers to articulate

their problems clearly, which enables an appropriate response by the agent. Zack

(2001) provides the following strategies to solve ambiguous problems:

• Reframing the problem to a more meaningful situation;

• Iterative cycles of incremental points of clarification; and

• Interactive face-to-face conversations.

Equivocality

Problems which appear to have “multiple meanings for or interpretations” (Zack,

2001, table 1) of the particular situation are referred to by Zack (2001) as problems

of equivocality. Coordinated action is required to resolve the range of interpretations

in a problem of equivocality by:

• Cycles of interpretation;

• Interactive discussion; and

• Negotiation to converge on one meaning.

Weick (1969) states that there is a requirement for an organisation to disrupt order to

make it possible to create order: to take information in equivocal form and transform

it into one that is unequivocal. He suggests an inverse relationship exists between the

degree of equivocality in a situation and the number of “rules” available to resolve

the problem. Since this process is cyclical, Weick (1969) finds an inverse

relationship exists between the number of rules available and the number of cycles

required to simplify unequivocal problems, as is represented in Figure 4. This

implies that a highly equivocal problem will have few rules (either formal or

heuristic) available, and hence would require many incremental cycles to define an

agreed meaning of the situation accurately.


31

Figure 4 - Rules and cycles in resolving an equivocal problem (Weick, 1969, p. 77)

Zack (2001) further develops a matrix, displayed in Table 1, which relates the

knowledge types (declarative, procedural, causal, and relational) to a range of

knowledge states determined by problem type. This presents both a taxonomy for

codification of requests to the call centre, and guidance in the strategic management

of knowledge needed to satisfy such requests. The matrix of 16 cells implies a

targeted knowledge management strategy for each cell, with the impact of such a

strategy able to be assessed in terms of its effectiveness in improving capability to

resolve such problem types.

Amount of equivocality in input

Number of rules used to assemble the process

Number of interlocked cycles selected for application to input

Amount of uncertainty removed from input

Inverse relationship Inverse relationship

Inverse relationship Direct relationship


32

Knowledge State

Knowledge Type

Ambiguous Equivocal Uncertain Complex

Declarative (know about)

No schema for classifying or describing something

Multiple possible classifications; or vague boundaries between classes

Classified or described within some level of confidence

Classified with confidence, but within a schema having many intricately related attributes

Procedural

(know how)

No script for describing a procedure or event sequence (e.g. tacit skill)

Multiple possible procedures or sequences

One procedure defined within some level of confidence of appropriateness or predictability of sequence

One procedure or sequence defined with confidence but having many intricately related steps

Causal (know why)

No explanation for why something occurs or occurred

Multiple possible explanations

Once explanation proposed within some level of confidence of appropriateness or predictability of cause/effect

One explanation defined with confidence, but having may intricately related causes and/or effects

Relational (know with)

No understanding of relationships among elements

Multiple possible relationships among elements

One set of relationships defined within some level of confidence or predictability

One set of relationships defined with confidence, but having many intricately related elements.

Table 1 - Knowledge States (Zack, 2001, table 2)

2.3. Summary

Initial research indicates a gap in the literature for the application and evaluation of

knowledge-based principles to call-centre situations. This research effort to address

the question “How well does Zack’s (2001) framework represent problem types, and

hence impact on strategies utilised in response to queries encountered in a public-

sector call-centre environment?” will make a contribution to the literature. Markus

(2001) and Timbrell et al. (2003) describe the characteristics of actors and their roles

in knowledge generation, capture, and reuse; Brogowicz (1990) provides a

framework for investigating service quality; and Zack (2001) presents a taxonomy of

problem types likely to be encountered. The existence of general principles and

practices for client interaction, problem resolution, and knowledge management

implies the existence of a conceptual matrix which informs response strategies to

customer queries that address these generic categories to produce optimal outcomes

for both caller and the service provider. Hence, the desired outcome from this


33

research project to improve quality of call-centre advice resulting from the

application of targeted knowledge strategies will include a range of strategies

targeting problem-types specific to the organisation.

Research Method

34

Chapter 3

Research Method

3.1. Introduction

Chapter Two explored a range of contemporary theoretical models that provide

guidance in the application of principles of knowledge management in the

improvement of advice-giving in a public-sector environment. This chapter focuses

on the selection of a research methodology and an associated protocol to ensure that

a logical and defensible approach is applied. The aims of the research and adopted

approach are initially articulated. The characteristics of customer needs and

classification of their problems is then investigated in order to inform the data-

gathering component of the protocol along with activities to identify responses of

call-centre agents. Finally, the practicalities of encoded data gathering are discussed.

3.2. Objectives of the Research

This research project aims to provide evidence that quality of advice given by call-

centre staff can be improved through judicious application of a model both able to

predict the outcomes of investment in a knowledge management strategy, and to

measure the impact of such a strategy. The research question is:




More specifically, the study aims to:

(1) Develop a model that defines the interrelationships between customer queries

and call-centre knowledge management strategies to provide subsequent

responses;

(2) Suggest knowledge management strategies to improve appropriate knowledge

reuse within call centres;

(3) Focus discussion and promote constructive interaction for developing a

sophisticated understanding of the call-centre environment, and within the

public sector in particular;

Research Method

35

(4) Explore characteristics of, and implications for, systems-based initiatives and

knowledge management interventions that would impact upon a call centre’s

ability to realise service quality benefits;

(5) Inform academic research directions in call-centre management; and

(6) Provide practical advice to call-centre management.

3.3. Approach

The study has primarily relied on qualitative analysis. By focussing on two

organisations, participant observation and interviews have provided the opportunity

to capture current organisational behaviour. To develop the richness of the study

further, archival data have been randomly selected to support the real-time

observations. The criteria for selecting a suitable research technique for this study

have included:

• An exploration of what is actually happening compared to what executives,

management, and agents suggest is happening; and

• Aligning these events with the theoretical frameworks identified as probable

determinants of improved advice-giving in public-sector call centres.

Access to many years of archival data from one of the organisations has provided a

temptation to develop a strong analytical tool based on statistical methods. The

combination of the complete data set (qualitative observations plus the statistical

analysis of archived data) would deliver benefits beyond that of a pure case

methodology (Gable, 1994). However, the study is attempting to identify causal

relationships (such as: better problem-solving capability leads to better advice),

which are not well served through descriptive statistics (Kaplan & Duchon, 1988).

Hence, the archival data are used to validate activities identified through observation

and interview by following a pure case study method. Yin (1994) recommends the

use of a case study methodology in a situation that allows a more complete

examination of factors which may produce a specific phenomenon. It gives support

to questions of why events happened, and how they have impacted on a

contemporary social organisation. Yin (1994) suggests this approach is particularly

applicable if boundaries between context and phenomenon are not clearly evident.

The major mode of analysis recommended by Yin is pattern matching. The analysis

Research Method

36

requires the investigation of patterns as they occur in the case study, and comparison

of findings with the theoretical model being investigated. A significant correlation of

predicted patterns with empirical examples implies a high level of internal validity.8

This test for quality of research design is satisfied in the study through the

investigation and verification of Zack’s (2001) problem types, Timbrell, Koller et

al.’s (2005) Query-Response (QR) Cycle, and Brogowicz et al.’s (1990) quality

criteria.

Another test which establishes the quality of empirical research is construct validity.9

Yin (1994) recommends the use of multiple sources of evidence and identification of

a chain of evidence to achieve this goal. He identifies the strengths and weaknesses

of each of the following six sources of evidence, and represents the convergence of

these in Figure 5. These strengths and weaknesses are summarised as:

1) Documentation as a source of evidence is stable, unobtrusive, and exact, and

has broad coverage, but suffers from irretrievability, bias from selectivity and

author access may be a problem.

2) Archival records are similar to documentation, and although they tend to be

more precise and quantitative, they are also less accessible due to privacy

issues.

3) Interviews can be targeted and insightful, but may reflect bias and

inaccuracies due to recall and reflexivity.

4) Direct observations have strengths in their reality and ability to capture

context, but are apt to be expensive, time consuming and reflexive.

5) Participant observation is similar to direct observation and can provide

insight into behaviours and motives but again may suffer from bias.

6) Artefacts provide insight into cultural features and technical operations but

suffer from selectivity and availability (Yin, 1994).

8 Internal validity establishes a causal relationship whereby certain conditions are shown to lead to other conditions (as opposed to spurious relationships) (Yin, 1994, p. 33).

9 Construct validity establishes correct operational measures for concepts being studied (Yin, 1994, p. 33).

Research Method

37

FACT

Archival records

Documents

Observations (direct &

participant) Structured interviews

and surveys

Open-ended interviews

Focused interviews

Figure 5 - Convergence of Multiple Sources of Evidence (Yin, 1994, p. 93)

Open-ended interviews with senior management of the two call centres studied have

been used to obtain information on the strategic roles of the call centre, business

drivers, and performance expectations. Interviews structured on the knowledge roles

of participants, as defined by Markus (2001), investigate the operational aspects of

the call centres from the perspective of team leaders. Observations (including direct

coding of calls) have been carried out over an extended period, with supplemented

documents provided to support claims in relation to operational procedures.

Use of replication logic in multiple case studies provides external validity.10 This

methodology has been applied to two public-sector call centres. This provides

limited confirmation of the replication logic, and also provides the opportunity to

expand the scope of future studies to commercial and non-call-centre environments.

No generalisations will be claimed as a result of this study, apart from the

verification of the models under investigation in public-sector call centres.

Reliability,11 the final quality test identified by Yin (1994), is obtained through the

definition of a protocol and the utilisation of a database to collect data in a pre-

10 External validity establishes the domain to which a study’s findings can be generalised (Yin, 1994, p 33).

11 Reliability demonstrates that the operations of a study – such as the data collection procedures can be repeated, with the same results (Yin, 1994, p 33).

Research Method

38

determined format. The protocol is able to be followed in multiple case studies

because the application of theoretical frameworks defined in academic literature

supports the protocols used in this study. A forms-driven database has been used to

codify the calls. A pilot set of calls has been used to train both the researcher and the

research assistant in the codification of calls. The recorded data used in the research

has been coded twice, once by the researcher and again by the assistant, and later

reconciled to ensure consistency. Variations have been discussed and resolved

through revisiting the underpinning framework developed by Zack (2001).

High quality analysis in empirical research is accomplished through addressing the

following four principles that should attract the researcher’s attention:

• Show that the analysis relied on all the relevant evidence;

• Include all major rival interpretations in the analysis;

• Address the most significant aspect of the case study;

• Use the researcher's prior, expert knowledge to further the analysis. (Yin,

1994, pp. 123-124).

These principles are used in Chapter Six to converge the sources of evidence to give

a clear and lucid account of the knowledge strategies and their effectiveness in

public-sector call centres.

Carroll et al. (1998) describe a process to undertake case study research. This

methodology is iterative in that the research cycle (as shown in Figure 6) is able to be

repeated as many times as is necessary to validate the issues under investigation. The

cycle commences with the development of a conceptual framework, and is followed

by development of a research design (modified as needed in successive cycles), data

collection, analysis, and reflection. Outcomes from each cycle inform the conceptual

framework and may lead to the development of theory relating to the research topic.

The study activities have traversed the loop three times. The first cycle was used to

establish that the overall objectives of the research were likely to be achieved within

the initial scope of the project. The meetings with senior management in both

organisations indicated a deficiency in both academic literature and pragmatic

models to inform the performance measurement and capability improvement of

public-sector service provision. Access to work environments was approved subject

Research Method

39

to caveats which protect the confidentiality of government business and the privacy

of individuals concerned.12 This cycle was completed with a need to identify a solid

protocol to form the basis of future data collection. This protocol was developed and

tested in the second cycle. This process involved structured interviews (based on the

work of Markus (2001)), data collection,13 categorisation (based on the work of Zack

(2001), and Brogowicz et al. (1990)), and codification (based on the work of

Pentland (1991)).

Conceptual Framework

Theory & knowledge Research themes

Literature

Insights

Theoretical foundations

Analyse Collect

Data

Plan Reflect

Figure 6 - The Structured Case Research (Carroll et al., 1998, p. 65)

The final cycle required the researcher to revisit the objectives of the research

project. It further entailed a final investigation of the contemporary literature to

determine if any recent work had been developed. The conceptual framework,

referred to as the Query-Response Cycle (Figure 7), had developed further, and was

utilised to refine the focus of the literature searches and reviews. It provided a

knowledge-based view of call-centre activities, where callers consign their query to

the call centre in order to have their problem resolved (Timbrell et al., 2005b). The

12 Ethical issues were also covered by the ethical research approval process within the Queensland

University of Technology (QUT).

13 A forms-driven Microsoft Access database was developed to assist in the collection, categorisation, and codification of data.

Research Method

40

cyclical nature of this model implies a continuous pay-off assessment is performed

by the caller, who may choose to be content with the response received (resolved),

re-query when the knowledge distance has been measurably reduced by a previous

cycle and there appears to be a benefit from continuing (re-query), or drop out of the

cycle without resolution if the response is not converging to a solution

(abandonment).

Figure 7 -Query-Response Cycle (Timbrell et al., 2005b, p. 546)

Within the call centre, a strategy of reference to second and third tiers simply

commences a new Query-Response (Q-R) Cycle with the directing of the caller’s

problem to more specialised (expert) agents. As reported by Timbrell et al. (2005b),

this escalation represents additional (opportunity) cost to the organisation through

employment of labour of specialists, and the diversion of these specialists from other

activities. Queries may be consigned several times. Each consignment starts another

Query-Response Cycle (Figure 8) until the resolution is passed back, either directly

or via the tiers, to the original querist. The elements on the left-hand side of the

Query-Response Cycle model, i.e., query reformulation, search strategies, use of

knowledge sources, creation of the response set, and the final offering of one or more

Research Method

41

responses, will vary according to the maturity of the knowledge infrastructure. The

actions of the respondent will vary according to the perception of service quality of

the querist (Brogowicz et al., 1990). The type of knowledge problem entering the

cycle has an effect on the knowledge processing and subsequent consignments to

more expensive knowledge resources such as expert-group tiers. Variations to this

model include referral of an action back to the caller (such as returning to the

problem source to gain more information), or with the agent taking the problem off-

line to carry out research free of the restriction of the telephone-bound

communication. These variations are specific to the search strategies and knowledge

resources represented in the left section of the model.

Figure 8 - Tiered Query-Response Cycles (Timbrell et al., 2005b, p. 550)

A simple model of the service process was also devised by Pentland (1991).

Although not cyclical, he cautions that, “While the steps in the process can be

thought of as though they were linear, this is not always the case” (Pentland, 1991, p.

92). The process for software support is defined as

“Opening ⇒ Classifying ⇒ Assigning ⇒ Diagnosing ⇒ Responding ⇒ Closing.”

Research Method

42

Opening (which equates to the query in Timbrell et al.’s (2005) model) is the process

of getting basic information about the caller, identifying the problem, and arriving at

the mutual agreement that the call centre will be able to assist with the problem.

Classification (Timbrell et al.’s (2005) query reformulation) is the next step towards

assembling a response. Pentland (1991) notes that:

“Even when an entitled caller states explicitly what he or she wants,

further discussion may reveal that something else is actually the problem,

or that there are multiple problems, or that there is no problem at all. The

classification of a call remains problematic until the call is actually

resolved, because one can never be sure that a problem has been correctly

identified until it is resolved.” (Pentland, 1991, p. 95)

This aligns with the problem types of Uncertainty, Ambiguity and Complexity in

Zack’s (2001) framework.

Assigning the call is the process of redirecting the problem to a specialist – a process

which management is attempting to minimise in the call centres being analysed in

this study. Diagnosis creates a more certain and agreed interpretation of the problem.

This may not be easily established, particularly if the knowledge distance between

the caller and agent is significant. Responding should involve an acceptable outcome

for the caller, but may cause the caller to consider that no solution will be found. In

this instance, the agent needs to act to preserve the relationship with the caller. A call

is closed only if the caller agrees that it is.

The objective of this research is to identify strategies and practices embedded in the

left sector of the Q-R Cycle (Timbrell et al., 2005b) which are able to improve the

capability of public-sector agencies. This enabling capability14 for the agency may

then provide competitive advantage if it is inimitable (i.e., a core capability), as

identified by Leonard (1995). Requests for information from a call centre in this

environment are varied and generally not predictable, even though they often relate

to political activity at that time. Factors to be addressed through application of the Q-

R Cycle include the nature of the problem (through a taxonomy provided by Zack

14 Public-sector agencies focus on effectiveness and efficiency principles in the delivery of government services. Hence, the preferred practice would be to assist other public-sector agencies to gain the same capability (hence it should be imitable) in order to enhance the service quality of the whole government further – not just of the agency.

Research Method

43

(2001)), determinants of quality of advice (Brogowicz et al., 1990), and the ability to

match and measure the impact of a particular knowledge management strategy,

particularly knowledge reuse (Markus, 2001; Timbrell et al., 2003).

3.4. Customer Expectations

There is a continuos and growing expectation that decisions and advice from public

servants is informed and rational. The Queensland legislation Freedom of

Information Act 1992 and Judicial Review Act 1991 embeds these principles. The

concept of informed citizens is significant in the current political climate: people

have an expectation that the systems of government are transparent, efficient, and

deliver a public benefit. The consequence of the information society is the increasing

options for choice. An expectation has been created that the public sector will

provide accurate and complete responses to requests for information, advice on

actions relating to government regulation, and simple personal benefit from services

provided by the public purse.

The problem for the public sector is the translation of such expectation into

successful strategy. How is a call-centre staffer, quite junior in status, able to respond

confidently to the range of requests directed from the public at large, which often has

an over-inflated expectation that this individual can provide the answer? A range of

knowledge strategies identified in the literature are able to be applied in a public-

sector call centre to solve this problem. Recruitment of intelligent staff, adoption of

structured approaches to training, use of both formal and informal networks of

“experts,” targeted systems development, involvement in discussions around the

coffee machine, and formal after-action reviews are several recommended

approaches (Choo, 1996; Cross et al., 2005; Davenport & Prusak, 1998; Leonard,

1995; Wiig, 1999). These strategies will be successful provided organisational

creativity, operational effectiveness, and quality of services are all improved through

competent use of knowledge (Todd & Southon, 2001).

Figure 9 illustrates that effective and efficient performance leads to satisfied citizens,

in turn inducing them to trust government (Van de Walle & Bouckaert, 2003). A

subsequent benefit is the establishment of trust between caller and agent, which

Jungermann (1999) sees as crucial in a service environment. This case study

approach will investigate the interrelationships between effectiveness and service

Research Method

44

quality on the basis that client trust is an outcome determined by these former two

criteria.

Figure 9 - Micro-performance Approach to Trust (Van de Walle & Bouckaert, 2003, p. 894)

3.5. Problem Resolution Strategies

In order to improve operational effectiveness, Zack’s (2001) taxonomy simplifies

what appears to be an impossible array of problem types into a manageable,

unambiguous set of knowledge states. The problem types of complexity, uncertainty,

equivocality, and ambiguity matched with know about, know how, know why, and

know with, allow classification of problem types. These can then be analysed

statistically to identify priorities for knowledge management strategy development.

According to Zack (2001), different types of knowledge problems are best processed

by differing knowledge and information systems strategies. Zack’s (2001)

knowledge problems are summarised in Table 2. As is illustrated, for each of his

knowledge problems, Zack (1999) suggests a number of information systems

strategies.

halla


Research Method

45

Information Knowledge

Lack of…

Uncertainty

Insufficient factual information about the goal, situation, or task, and some lack of confidence in the consequent inferences, estimates or predictions required.

Ambiguity

Inadequate knowledge (patterns/concepts) about, no explanation for, or understanding of a goal, situation, or task

Variety /

Diversity

of…

Complexity

Too many situational elements and relationships to coordinate or consider simultaneously

Equivocality

Multiple interpretations of a goal, situation, or task

Table 2 - Summary of Zack’s Four Knowledge Problem Model (Zack 2001, figure 1)

For problems of uncertainty, Zack (2001) suggests:

1) Providing central repositories to enhance the ability to locate codified and

documented information;

2) Providing automated capabilities to analyse large amounts of information;

3) Configuring communication networks in highly flexible ways to respond to

unpredictable information processing needs;

4) Enabling communication regardless of geography or time; and

5) Enabling broadcast at-large requests for information and knowledge, thus

eliminating the need to know precisely where it is located.

Becker’s (2001) strategy to reduce uncertainty involves acquiring increasing

amounts of information until the problem is diminished in terms of its uncertainty.

Becker also suggests that a problem defined as ambiguous will not be resolved by

this strategy; rather, it may make it worse. The information already at hand needs to

be processed in order to reduce an ambiguous problem to a complex one.

For problems of complexity, Zack (2001) suggests:

1) Auxiliary high-capacity memory for managing and analysing complex sets of

information rapidly, i.e. computer-based decision support systems, database

systems, and expert systems;

2) Develop searchable online repositories of explicit knowledge to leverage the

organisation’s experts;

3) Develop the ability to locate experts spontaneously and quickly; and

Research Method

46

4) Facilitate decentralised decision-making by making local information

available globally and global information available locally.

Utilisation of “embrained knowledge,” the cognitive skills and conceptual abilities

used by people to understand complex causations, is a strategy widely relied on that

releases the latent mental potential in problem solvers (Thompson & Walsham,

2004).

For issues of ambiguity and equivocality, Zack (2001) suggests the provision of

communication technologies to best support dialogue between a flexible and

responsive network of experts and associates. The purpose of the dialogue is to apply

the available knowledge resources to transform problems of ambiguity and

equivocality into problems of complexity and uncertainty. In some ways Zack’s

(2001) approach is similar in nature to Hansen et al. (1999), who describe their two

knowledge strategies of codification and personalisation, promoting the use of

information systems in their codification strategy, and communication technologies

in their personalisation strategy. The nature of the problem types (queries) directed at

the call centre affects the processes employed to respond to that problem.

Figure 10 - Generic Taxonomy of IT Helpdesk Calls (Pentland, 1991, p. 102)

halla


Research Method

47

A generic call classification system (Figure 10) was developed by Pentland (1991) in

his study of IT helpdesks. This hierarchical model initially divides between

“technical” and “non-technical” calls.15 Pentland (1991) in turn breaks technical calls

into “problem” and “question.” “Question” is then broken into “general,” “how to,”

and “compatibility.” These question calls mirror the problems of uncertainty

identified by Zack (2001). The “problem” calls are further subdivided into “ordinary

problem,” “bug,” “user error,” and “other vendor’s problems.” The difference

between a problem and a question is defined in terms of the callers’ expectations.

Pentland (1991) suggests that a caller does not have a problem until they have

attempted a solution and it did not work. A framed question (such as “How do I…?”)

makes the problem easy to identify, but this may become an issue if the helpdesk

response is a suggested solution already tried without success.

The problems will then rely on the abilities of the helpdesk staff to identify the

causalities, whether that of software or the user expectations of that software.

Complex problems, as defined by Zack (2001), would cover a range of calls

identified by Pentland (1991), but when callers have an expectation that the software

will perform certain functions (when in fact it will not or needs intervention from

another expert), it becomes a problem of ambiguity. Equivocality arises when the

system designers misinterpret functional requirements.

3.6. The Knowledge Management Strategy

The case study method also captures the strategies used by agents to resolve queries

to the call centres, allowing identification of relationships between roles, problem

types, and resolution methods. The works of Markus (2001) and Timbrell et al.

(2003) provide a guide to development of an appropriate knowledge management

strategy. The quality of advice given relies substantially on the relationship

established between the caller and receiver. For call-centre staff to be confident, they

need to be knowledgeable and responsible for their decisions. The inability to answer

a query immediately and accurately is not regarded as poor advice; rather, poor

advice comes with an attempt to cover for such a knowledge gap. Dixon (2000)

15 Pentland (1991) does not expand on the non-technical calls, since only technical calls were the focus of his study, which referred to a strategy of immediate referral back to a switchboard. His research indicates a fairly clear distinction between technical and non-technical calls, but admits some calls were of a hybrid nature.

Research Method

48

provides advice that a knowledge reuse strategy encompasses a “complete solution”.

Simply relying on technologies such as repositories and fast information technology

will not deliver improvement in performance. Also, design of incentives to increase

contribution to, and use of, information repositories is to be considered, since this

will provide support for the codification and storage of reusable information

identified as being essential to a successful knowledge reuse strategy.

The reuse situations identified by Markus (2001) and Timbrell et al. (2003) include

activities by data miners who do not interact directly with callers in that role, and so

will not be considered further in this study. Of the other roles identified by these

authors, shared-work producers are to be encouraged to make information available

to the whole team, rather than storing it for their own reuse (which is far simpler to

do). Shared-work practitioners also need to become producers, as they add value

though their work. Expert-seeking novices, as they extract relevant knowledge from

experts, are to make that knowledge available through community of practice

discussion, contribution to a repository, or by some other means. The goal is to

manage effectively both the codified knowledge of the organisation, and the expert

knowledge that resides within call-centre agents and every other employee of the

organisation. In each reuse situation, the information system that supports the

knowledge management initiative must be capable of accessing expertise as well as

the experts.

The role of the knowledge manager is of critical importance for the development of

organisational strategies. It is not sufficient to rely on call-centre staff to develop,

contribute, and disseminate relevant knowledge. Three key functions of such a role

include:

a) Ensuring valuable information is recorded in a knowledge repository;

b) Reviewing such information to ensure its currency and finally providing

assistance in locating; and

c) Linking historical information relating to areas of interest (Markus, 2001).

This case study investigates the existence of knowledge strategies in two Queensland

Government call centres and the roles individuals play in implementing them.

Research Method

49

3.7. Quality and the Call-Centre Response

Quality criteria, as identified by Brogowicz et al. (1990), are coded in the data

collection phase of the project. The three areas of focus are:

1) Training of staff, which is critical to ensure the determinants of quality

identified by Brogowicz et al. (1990) are addressed. These determinants of

tangibles (those visible signs of quality such as work space and contemporary

technology), reliability, responsiveness, assurance, and empathy are personal

qualities that require support from both technical and functional expectations.

It is anticipated that interviews with mangers and staff will provide data that

identifies training strategies.

2) The capability to incorporate knowledge captured in information systems into

general problem-solving and learning heuristics is referred to as the technical

determinant of quality. The technical determinants within a call centre are

substantially impacted by the individual staffer’s knowledge of the context of

the problem presented by the querist. Both individual and organisational

capabilities are able to be improved through experience, training, and access

to codified information. Physical resources identified as technical

determinants of quality include the information technology infrastructure,

communications infrastructure, and published materials.

3) Functional service quality, the expressive performance of a service, reflects

the way solutions derived as a result of technical capability are transferred

through personal attributes of attitude to service, accessibility, and public

relations (Gronroos, 1984, p. 39). Selective recruitment and training, and an

organisational culture overtly attuned to client service are precursors to

meeting the functional service quality gap.

Quality of advice is a core capability of successful service organisations. Leonard

(1995) proposes a model for the creation of capability (as shown in Figure 11)

through shared problem solving, internal implementing and integrating, future

innovation, and importing knowledge from external sources. Improvement in the

quality of advice provided by call-centre agents may be continuous and incremental

if many of the fundamentals of the model are in place. The requirement is for agents

to transform from the individualistic shared-work producers (who create knowledge

Research Method

50

primarily for their own later reuse) to team-based shared-work practitioners (who

acquire new knowledge from external sources in context). This suggests that

judicious prioritisation of knowledge reuse strategies would lead to improved

capability, and hence performance improvement in problem-solving exercises. As

knowledge management strategies are implemented to address each problem type,

successful outcomes are able to be embedded in practice, and unsuccessful ones

discarded or re-evaluated. The case study methodology provides an opportunity to

document the degree to which problem solving in the call centres is a shared or

personal activity.

Figure 11 - Capability-Creating Activities: Problem Solving (Leonard, 1995, p. 60)

3.8. Pilot work

Initial rounds of consultations were held with senior staff of the host agencies to

provide input into the further refinement of the framework. One call centre was

identified as having a complete recording of all customer conversations over the past

eight years. An initial sample of these calls was used to establish a reliable and valid

coding protocol (based on Zack’s (2001) problem types, Markus’s (2001) reuse

strategies and Brogowicz et al.’s (1990) quality assessments). Team leaders from the

call centres took part in workshops to review the protocols and processes, and to

develop an instrument to assist in the documentation of the information gathered.

PRESENT

INTERNAL

FUTURE

Shared Problem Solving

Improving Knowledge

Experimenting

Implementing and Integrating

EXTERNAL

Core

Capabilities

Research Method

51

3.9. Implementation

3.9.1. Participants

Three cycles of the Structured Case Methodology (Carroll et al., 1998) make up the

high-level processes in the study and focus on different staff groupings, as shown in

Table 3.

Stage 1

Establishing

the

Context

Stage 2

Embedding

the

Process

Stage 3

Performing

the

Analysis

Table 3 - Three Cycles of the Structured Case Method as Applied to This Study

Senior executives responsible for the management of the call centres have been

consulted in Stage 1 of the research project and gave approval to proceed. Through

informal and unstructured interviews, they provided insights into the business drivers

Strategic Requirements

Business Drivers Quality Criteria Problem Categorisation

Identify expected individual and organisational

behaviours

Unstructured interviews

Why does the call-centre

exist?

Link knowledge strategies to

expected outcomes

Operational Protocols

Operational Constraints Personal Impacts Problem Resolution

Develop data gathering tool for

call analysis

Structured interviews and

meetings

Gain support of call-centre

staff

Test and agree on coding protocol &

Pilot

Implementation

Literature (Zack, Brogowicz, Markus etc.) Insights Heuristics

Secondary code in NVivo to identify relationships and chain of causality

Collect and code the data

Identify relevant sources of

research data & information

Reflect on the research findings

Research Question

Resolution

Research Method

52

behind the call centres and the knowledge strategies they employed in relation to

knowledge reuse and its contribution to agency outcomes (being improved advice).

The executives contributed further by providing feedback on the initial conceptual

framework proposed to analyse call-centre performance.

Call-centre supervisors were interviewed to develop the data-capture process, and

also to assist in the validation of coding. Particularly in the pilot phase, these team

leaders were used extensively to ensure the coding rules reflected the nature of the

work being performed by the agents. They also assisted in the identification of key

phrases which provide “hints” for that coding process. Another role for the team

leaders has been to ensure that their staff members were comfortable with the

research project. To this end, several staff briefings were organised for each call

centre where the research project was presented, and queries relating to both

concerns and benefits were addressed. As well as providing call-centre operators

with general background of the project (a development exercise in its own right), this

ensured researcher’s access to staff during the data collection phases in a relaxed and

non-threatening environment.

The majority of the data has been taken from archival recordings taken by one of the

call centres. Live calls by the centre staff have also been monitored to assist in the

protocol development on both problem types and quality of advice. This is consistent

with a current practice where callers are advised that monitoring may take place,

primarily for the purpose of training. Callers have the option to veto such a practice.

3.9.2. Data Gathering

A rich source of data has been made available through allowing interviews with

managers, access to procedural and factual documents, recorded phone calls, and

archived materials covering the lives of the call centres. Over this period, several

strategies have been applied in order to improve call-centre performance. Anecdotal

evidence exists to indicate the success (or otherwise) of these strategies. Advice-and–

complaint-type calls have been sampled around these events and coded to capture:

1) Problem type (Zack, 2001);

2) Quality of advice (Brogowicz et al., 1990) and

3) Identification (if possible) of the resolution strategy used.

Research Method

53

In particular, the use of codified information (e.g., database, intuitive, training,

referred to tier 2) was recorded as it is a strong example of knowledge reuse. By

sampling around known interventions, significant changes in agent behaviours would

indicate a consequence of that event.

Pentland and Rueter (1994) have performed a detailed analysis of calls and responses

to helpdesks for software support of complex computerised systems in order to

develop a grammar to describe organisational routines. Their analysis defines the

processes specialists used to resolve problems, as shown in Figure 12, such as

transferring the problem to other areas, escalating it to engineering staff, making

reference to vendors, or by transferring back to the customer for further investigation

and information collection. Pentland and Rueter (1994) have devised and validated a

coding scheme for this process, as per Table 4. Their analysis focuses on non-routine

problem-solving work with a high variation in actual problems, but with a similarity

in call types (e.g., a familiar problem on a different platform). The problems

encountered have been classified as having low analysability owing to the user

uncertainty as to what caused the problem, whether a feasible solution exists,

inability to construct an explanation for the problem, and not always having

sufficient information to isolate the symptoms of the problem.

Code Explanation

O Open the call.

W Work on the call.

C Close the call.

I Declare problem inactive, usually because it can't be reproduced, so no further work can be done.

D Defer the problem, can't solve now.

T Tentative fix completed.

R Resolve problem (but leave call open for confirmation by customer).

F Fix given to customer.

E Explain reason for closing.

US Transfer responsibility to user.

PS Transfer responsibility to product support.

DV Transfer responsibility to development.

TR Transfer call to another functional area.

Table 4 -Coding Scheme for Software Support Process (Pentland & Rueter, 1994, p. 494)

hoshiko

Rectangle

halla

This table is not available online. Please consult the hardcopy thesis available from the QUT Library

Research Method

54

Figure 12 - Kinds of Responses to Calls to IT Helpdesk (Pentland, 1991, p. 115)

Using the Pentland and Rueter (1994) analysis as a model, the codes in Tables 5, 6,

and 7 have been proposed to incorporate the activities being investigated in this

study. Sufficient records are available to obtain statistical significance, but the coding

protocol will limit the amount of detail that will be recovered about any particular

interaction. This will not impact on the fit of the research data with the proposed

model being investigated, but will afford the necessary degree of confidentiality to

callers and call-centre agents.

hoshiko

Rectangle

halla


Research Method

55

Zack’s (2001) Problem Types

Code Type Description

A Ambiguity “Inability to interpret or make sense” (Zack, 2001, p. 21). This,“cannot be coded precisely into mutually exhaustive and exclusive categories” (Weick, 1995, p. 92)

C Complexity “A large number of parts that act in a non-simple way” (Zack, 2001, p. 18)

E Equivocality “Multiple meanings” (Zack, 2001, p. 21)

U Uncertainty “Lack of enough information to chose an exhaustive and well defined set of possible states, preferred outcomes and actions” (Zack, 2001, p. 19)

X Not Coded Not coded – does not fit the Zack (2001) model

Table 5 - Proposed Coding Scheme for Problem Type

Resolution Strategies


PD Took Personal Details

Personal details taken, such as name, phone, address.

CR Cyclical reframing

Cyclical reframing of the problem through testing partial solutions (Weick, 1995).

DP Decompose Decompose into smaller manageable problems and resolve (Zack, 2001).

MD Meaning through discussion

Meaning through discussion (trial and error, sounding out) (Weick, 1995).

RI Requires research

Requires time to research further information – will return call.

SC Social construction

Social construction and intervention through scanning and discovery. The environment is not just given, but rather can be “enacted” through the actions of agents, and produced by knowledge rather than the opposite (Weick, 1995).

TC Terminate Call Terminate Call due to unforseen factors such as caller hang-up, etc.

TU Transfer to user Transfer responsibility to user (more information or some action by caller needed).

TR Transfer to Functional Area

Transfer call to another functional area.

TS Transfer to supervisor

Transfer responsibility to call centre supervisor.

CI Use codified information

Access codified information (such as databases, fact sheets etc.).

Table 6 - Proposed Coding Scheme for Resolution Strategy

Research Method

56

Quality Characteristics


EA Ease of access Ease of access to the call-centre service (Brogowicz at al., 1990).

EC Empathy with customer Empathy with customer (Brogowicz et al., 1990).

CS Satisfactory closure Satisfactory closure of the call.

CT Technically competent Technically competent and knowledgeable (Brogowicz et al., 1990).

TQ Time waiting in queue Time waiting in queue < 30 seconds (Brogowicz et al., 1990).

RC Technical resource capability

Technical resource capability (eg access to Databases, IVR. etc.) (Brogowicz et al., 1990).

CB Not available - Call Back

Attempt to transfer call to a 2nd tier - not available so gave call back details.

NEA Difficult to access Negative - Access to the call centre service is difficult (Brogowicz et al., 1990).

NEC No Empathy with customer

Negative – Displays no empathy with customer (Brogowicz et al., 1990).

NQC No quality criteria used No quality criteria used.

NSC Unsatisfactory closure Negative - Call closed without an agreed resolution.

NTC Not Technically competent

Negative - Unable to address problem resolution in a logical manner (Brogowicz et al., 1990).

NTQ Excess Time waiting in queue

Negative - Time waiting in queue > 30 seconds (Brogowicz et al., 1990).

NTR No Technical resource capability

Negative - Lack of technical resource capability (eg no access to Databases, IVR etc.) (Brogowicz et al., 1990).

Table 7 - Proposed Coding Scheme for Quality Characteristics

3.9.3. Validity of Results and Data Analysis

This research has been limited to three cycles. The first cycle has been used to

validate the coding protocols and to establish a mechanism to ensure such a process

is learnable and repeatable. Thirty recorded calls were initially hand-coded by the

researcher. The results were then presented to the call-centre team leaders and

reviewed item-by-item. This process has also been used to identify “triggers” to be

used in the real-time coding of calls. Once an agreement was reached with team

leaders, a research assistant again coded the test set using the queues and other

information relating to the research topic. This exercise has been performed to ensure

that the researcher and assistant were able to code consistently. It has also established

that no more than five strategies have been applied for both resolution and quality

actions. The capture tool to be used in the major cycles has thus required only five

fields for storage of these codes. An ancillary benefit to this process has been the

Research Method

57

demonstration of the removal of identifying data necessary to meet the ethical and

confidentiality requirements of the agencies and the University. The test for construct

validity has been satisfied by the application of the restricted coding set (Yin, 1994).

Data collected in the second cycle is from recorded calls in one of the call centres.

Each call has been double coded for consistency. The coding process has been

assisted through the use of a database which provided timing, sequencing, and

coding automation (see the data definition in Table 8). Drop-down menus and

enforced metadata capture have meant that the coded data has been able to be later

scrutinised to confirm consistency and repetition of the process. Double-coding has

revealed few anomalies, which have been reviewed and changes agreed. These

results have again been discussed with call-centre team leaders, who have again

confirmed the coding to be consistent.

Field Name Data Type Description

time_Encoded Date/Time Default value = system time

call_Centre Number Either CC-A or CC-B (toggle)

call_Source Number To record whether a recorded or real time - recorded calls capture the media details of the recording (eg CD#99)

call_ID Text Unique sequential system identifier

call_Date Date/Time Default value = system time; able to be overwritten to reflect recorded call time

call_Topic Text General description - allows possible analysis of motivation for making the call

zack_Code Text Any of the 4 codes and option not to satisfy the Zack (2001) criteria (drop-down)

qualCode1 Text One of the 14 codes of the Quality criteria (drop-down)

qualCode2 Text Another of the 14 codes of the Quality criteria (drop-down) - code(s) eliminated if used in any previous quality field

qualCode3 Text As for qualCode2



resolveCode1 Text One of the 11 codes of the resolution criteria (drop-down)

resolveCode2 Text Another of the 11 codes of the resolution criteria (drop-down) - code(s) eliminated if used in previous quality field

resolveCode3 Text As for resolveCode2



call_Comment Memo Available to enter comments which may be pertinent to the call

Table 8 - Data definition for data capture tool

A final cycle has been used to capture real-time data from the second call centre.

This extra data has been gathered to compare and contrast data from different call

centres. The information collected has been analysed to establish its consistency

within the framework. The purpose of this action has been to establish the external

validity and reliability through replication (Yin, 1994).

Research Method

58

3.10. Summary

Chapter Three has been used to detail both the reasons for selecting a case study

methodology and the protocols used to enact the methodology. Detractors of case

study methodology point to a weakness associated with repeatability (Gable, 1994).

Since the framework proposed is a performance improvement model, it is not able to

provide an absolute measure for outcomes. It does, however, provide a model which

informs practice, and offers indicators for changes to outcomes. This aspect of the

research is repeatable in any related advice-giving environment. Yin (1994)

recommends a range of tactics to address validity and reliability criteria for quality

research design. The protocols developed for this study follow his recommendations.

The convergence of multiple sources of data, adoption of a structured approach, and

the incorporation of two organisations in the case study have ensured Yin’s (1994)

quality criteria have been addressed.

Work, Organisation, and Processes

59

Chapter 4

Call-Centre Work, Organisation, and Processes

4.1. Introduction

This chapter depicts life in two Queensland public-sector call centres. The systematic

observation of activity in two call centres identifies a variety of behaviours which are

characterised by different abilities to utilise knowledge strategies contributing to the

establishment advice-giving practices in the public sector. In effect, the approach

emphasises the commonalities and differences in systems and the behaviour of

agents. At the core is the focus on service quality by individual agents, and the fact

that knowledge is abundant, but that performance is limited by the ability to use it.

Management is, however, concerned with the allocation of scarce resources and their

alignment with localised business drivers.

4.2. A Broad View of Call Centres

Milner (2000) analyses the growth of public-sector call centres, and concludes that

the public’s familiarity with and access to telephones has allowed these centres to

flourish. She draws on the example of the Brisbane City Council, which has

articulated two key criteria for change needed to implement a call centre. These

criteria are “advantage to customer” and “demonstrate the achievement of value for

money”. Technology has provided the enabling role in the provision of information

access via the Internet and corporate Intranets, mobile and remote service delivery,

and in creating a sense of empowerment for users. The leveraged service quality

through consistency and ease of access to services supports the criteria for success

and provides a blueprint for the operation of public-sector call centres. The value-for-

money criteria is reflected in the knowledge self-sufficiency of call centres, i.e., the

ratio of queries resolved to the total queries within a call centre ( Timbrell et al.,

2003). The lower cost of problem resolution within the call centre, relative to tier-2

or specialist referral areas, means this measurable and meaningful ratio is indicative

of call-centre performance. It also supports the customer-advantage principle, since a

highly self-sufficient call centre has the capability to handle the majority of calls

without the need to redirect them to a third party.


60

Customer advantage also exists for callers in their ability to connect with the call

centre conveniently. Toll-free numbers (where the call centre accepts the charge) are

widely used, but a normal call (where the caller bears the cost of the call themselves)

still exists for some centres. Having connected with the call centre, the caller is

welcomed by either an agent or an interactive voice response (IVR) system. An IVR

system may take the caller through a layered menu of options, resulting in a

connection with a specialist agent, a general-enquiries agent, or an automated system

that requires no human intervention. Contemporary examples of this practice include

flight arrival information, recorded weather reports, and stock market prices. In

some cases, the caller can exit from the IVR system and connect directly to, or queue

for, an agent. The following provides an insight into the two call centres, designated

as CC-A and CC-B.16

4.3. Evolution of the Study: CC-A

The first call centre, CC-A, has been the initial focus of the study. This study

commenced in early 2002. CC-A was (at the time) the focus of efficiency drives, as

well as responding to changing legislation that contributed to significant growth in

the number of calls to the agency. Management was primarily concerned with

throughput, and allowed team leaders considerable freedom in the day to day

operations of the call-centre. CC-A’s operation is characterised by technology-driven

improvement, with performance reports reflecting the volume and time spent

responding to calls. However, a structural change in the organisation caused the

appointment of a new Executive Manager whose responsibility covered certain

aspects of the call centre.

External performance measures were instigated in an attempt to gauge the service

quality of a major component of the agency, and in particular the services delivered

via CC-A.

16 The identities of the organisations, staff, and callers in this study are disguised through coding. Although people familiar with the operations of the departments may identify the particular call centres being studied, no individual will be exposed. There is no intent to make value judgements on performance; rather, observation of activities and responses is documented in order to develop the frameworks necessary to provide the repeatability of findings. Nothing is lost in this anonymity since the assurance of confidentiality has allowed participants to behave in “normal” ways, and to be honest in face-to-face interviews.


61

Another organisational restructure transferred responsibility for the call centre to a

senior officer who initiated a number of changes to improve its performance. From a

knowledge perspective, each initiative has had a positive impact:

a) The introduction of “teams of teams” encouraged the sharing of knowledge

across the agency (including regional offices) via constant communication,

thereby ensuring all agents were kept current. This is a strategy referred to as

“knowledge streaming” (Timbrell et al., 2005b). These teams were initially

facilitated externally to share their vision and develop action plans in a two-

day workshop. Six-monthly face-to-face meetings were also planned, with a

commitment to improving intra-office communications.

b) A mentoring (“buddy”) system was also introduced where junior officers

were linked with senior members. Each mentor had only one buddy. These

buddies were nominated by management to ensure appropriate match of

experience, skills, and personalities. The knowledge issues addressed by

these strategies is both the sharing of knowledge, in much the same way as

apprentices learn their trades through socialisation and internalisation

(Nonaka & Takeuchi, 1995), and also to address the potential loss of

knowledge to the organisation through retirement of experienced officers.

This is of concern as the age demographic shows 24 percent of the agency is

50 years or older (i.e., potential retirees), while 40 percent are over 45 years

of age. See Figure 13 for a graphic representation of the age demographic

existent in the agencies.

c) The final initiative was the employment of a Training Officer, to maintain the

static database of information used, design, and deliver induction and

training, and to document processes and standards of operation formally

within the call centre. This strategy targeted the management of explicit

knowledge as identified by Nonaka and Takeuchi (1995) in the

externalisation and combination modes of knowledge conversion. In an

interview in December 2002 (Interview 1b - Annotated notes, December 23,

2002), the team leader suggested this regime was aimed at multi-skilling the

call-centre workforce, and that it has had a positive impact on both the

motivation of individual agents and the performance of the call-centre

operations.


62

Agency Age Distribution cf State Employed Population

0%

5%

10%

15%

20%

15-

19

20-

24

25-

29

30-

34

35-

39

40-

44

45-

49

50-

54

55-

59

60-

64

65-

74

75 +

Age Range

%

State %

Agency %

Figure 13 - Age Demographics of staff employed by the agency hosting CC-A

In an unstructured interview with the Executive Manager of CC-A in early 2003

(Interview 1a - Annotated notes, January 29, 2003), plans to obtain hard data were

discussed. Issues of concern were identified by the manager as a result of unsolicited

feedback from clients. These concerns primarily related to the consistency and

accuracy of advice provided by the call-centre and other front-line service staff as

they interpreted and applied the range of regulations being administered. The

absence of reliable data to support these concerns led to the instigation of a project to

obtain hard data capable of guiding decision making for improvement strategies.

Employing a technique developed by a sister agency in another Australian State, an

external analyst was engaged to develop, deliver, and analyse a “mystery shopper”

exercise. This had the intention of testing call-centre staff on matters of accuracy

and consistency of advice being offered. Approximately 300 telephone calls and 50

email queries were performed. Call-centre staff were assessed on greeting, call

times, helpfulness, and accuracy, as compared to model answers. Client exit

interviews and mystery shopper activities produced sets of data which were analysed

and reported on by the external analyst. In order to ensure the cooperation of the call-

centre staff, however, an embargo was applied on the findings by the agency. The

scenarios presented were made available but due to the focus on restricted function-


63

specific issues sets17 (primarily recall with little problem solving), they failed to

provide the breadth required for a valid18 and repeatable research exercise. The focus

of the research then changed to CC-B due to its reputation for being a best-practice

model within the Queensland public sector.

4.4. Evolution of the Study: CC-B

In the initial meeting with the manager and the team leader of the second call centre,

CC-B, the strategic role for the call-centre was discussed. The management of

people, processes, and performance was focussed on delivery of a service based on

quality principles, rather than efficiency. The team leader made the point that in this

organisation “quality is more important than time.” The call centre has fewer than

ten permanent staff members who have been recruited on the basis of attitude rather

than experience. This recruitment is followed by a training period of up to six

months, which allows the agent to develop a deep knowledge of the organisation.

CC-B has a well-documented set of processes covering both inbound and outbound

call functions. The provision of “tele-research” is the major outbound service offered

by CC-B on a fee-for-service basis. Call-centre agents work closely with clients to

develop survey questionnaires to ensure that the business group’s goals in

undertaking the survey are met. Consultancy regarding the timing, accuracy, and

type of questions is provided before data collection begins. Management has

emphasised the importance of the business group liaison with the interviewers, in

order to provide an in-depth understanding of the questions and intent of the survey

and the profile of the client base. A database has been created for the clients’

responses. If a hard copy is required by the business group, records can be updated

from paper-based questionnaires. Before the survey instrument is universally applied,

a pilot is performed on a sample of five to ten clients, and a post-pilot report is

provided. This report includes client response, length of time to complete the survey,

clients’ understanding of questions, and accuracy of information obtained. On

17 The scenarios developed were mainly problems relating to uncertainty in Zack’s (2001) framework and tested either accurate agent recall or the use of the dedicated database for scripted answers only. 18 An interesting phenomenon reported was that some of the experienced call-centre agents were able to identify the mystery shopper calls through the inconsistent responses to in-depth questioning by the agent. The knowledge distance was apparently greater than normally expected of callers to the call centre.


64

completion, a final report and synopsis is provided to the business group in electronic

format. This report contains the final figures, percentages, graphs, and any other

information requested by the business group. The synopsis contains interviewer

observations and comments, and adds anecdotal information regarding the conduct of

the survey for the information of the business groups. Although market research has

not been a conventional undertaking for public sector agencies, it is clear from the

activities of CC-B that an agency has the capability to deliver services both to

external and internal clients. Such a role will always be supplementary to the general

information/advice functions of government.

Functions performed

Preparation Negotiation & Performance

Acceptance

Client

Call Centre

Resp

on

sib

le

Are

a

Business Area

Table 9 - General information processes in CC-B

General information processes in CC-B are documented in process charts having a

similar format to Appendix 1, and reflect a matrix structure assigning function

preparation, negotiation, performance, and acceptance to client, call centre, or

business area. Calls to the call centre are managed in a way that is consistent with the

Query-Response Cycle (Figure 14), where they are either resolved via the knowledge

resources of the agent (primarily the “Preparation” column of the documented

process in Table 9), or consigned to tier 2 (primarily the “Negotiation and

Performance” column of the documented process in Table 9). Other documented

processes, with variations on the general information procedure, include the straight

transfer of calls to nominated agency staff or services, the provision of specialist

advice/information, mail-out processes (caller-instigated), non-English speaking

caller processes, complaints, and an emergency response hotline process. The

performance levels relating to these processes are documented in a Service Level

Agreement (SLA).


65

Figure 14 -Query-Response Cycle (Timbrell et al., 2005b, p. 548)

The SLA provides a framework for the development of a constructive and effective

relationship between CC-B and its Business Units. The framework has evolved from

an examination of the business needs of the call centre, and describes the levels of

service to be provided to meet those needs “and how they will be monitored,

evaluated, measured, and managed” (Service Level Agreement for Provision of

Services, 2004).19 Call-centre business process service levels are encompassed in the

general responsibilities of the call centre, business units, and specifically-nominated

roles within the agency. This reflects the symbiotic relationship between the call

centre and the wider agency, and specifies the type of personal dialogue to ensure the

maximum self-sufficiency of each. This apparent paradox balances the knowledge

needs of the call centre to reach its goal of resolution of 80 percent of calls without

referral, and the recognition that business units do not wish to be interrupted by

19 It is not intended that this agreement should have legal consequence; rather, that it serves the mutual benefit of both parties by providing a clear understanding of agreed operating arrangements and performance criteria. It is expected that it will evolve over time.

Knowledge

Problem

Query

Response

Query Reformulation

Search Strategy

Knowledge

Sources

Response Set

Pay off

Assessment

Knowledge Distance Adjustment

CONSIGN RE-QUERY

RESOLVE

ABANDON

Service Quality


66

“trivial” issues, such as the dissemination of information able to be codified and

reused. Whilst it is the responsibility of the call-centre operator to communicate

accurate, complete, and up-to-date information to callers, it is the overall

responsibility of the business units to ensure the information within the relevant

databases is accurate, complete, and up-to-date. This is achieved through annual

reviews undertaken by the business units, and continual updates from clients. Figure

15 indicates how callers access the call centres and are managed in the initial phase

of the Q-R Cycle.

The operational aspects of both call centres studied, as documented below, were

presented to the 15th Australasian Conference on Information Systems in December

2004, and published in the conference proceedings (Schefe & Timbrell, 2004).

Research

and

Return call

CC-A Only

CC-A Only

The Operational Context

Caller

Integrated Voice Response

General agent

Specialist agent

Consign to other

Supervisory or

Specialist staff

Measure

Service

Quality}CC-B Only

Legend:

communication flow

is limited to the call-centres indicated.

communication flow

in both call-centres.

Figure 15 - Operational Context of the Call-Centres Studied

4.5. Operational aspects of CC-A

CC-A is a call centre in a Queensland government department that builds consumer

and business confidence in fair marketplace outcomes. CC-A is an integral

component of a Customer Service Centre (CSC), which also includes counter staff

both in the capital’s central business district and at regional centres. Calls to the

widely publicised 1300-number (which allows calls from anywhere in the state to be


67

charged at local call rates) are routed to the nearest available reception point, which

will be either a regional office or the call centre located in the central business

district. The IVR system then routes the call based on whether it is a general query

or complaint (which will be more complex) or a transaction relating to the regulation

of business. Transactional functions are performed in real time, while client issues

may involve further research, be consigned to second-tier experts, or investigation’s

officers where legislative breaches have occurred.

Prior to a reorganisation in October 2002, the CSC comprised approximately 60

Brisbane-based staff. Regional staff generally processed over-the-counter enquiries

or took calls directly. The CSC consisted of teams with clearly defined roles

separating transaction processing (lower-level staff), complaints via the phones,

written complaints, and other business services, such as data entry and bank

reconciliation of payments. The situation was characterised by very little

communication between teams, limited written procedures, and no ability to load

balance (the process of distributing effort to areas currently experiencing the lowest

resource demand) during peak times. There was an imbalance in supervision loads

which generated some resentment. Training was minimal: it was on-the-job and ad-

hoc.

After October 2002, the organisation corrected the (both real and perceived)

inequities in work through the formation of work clusters and multi-skilling of all

staff in the CSC. The supporting information systems in the regional offices were

integrated with the Brisbane systems, and staff performing customer-service roles in

the regions were included in the CSC division. The PABX system commenced

routing callers to the nearest available service staff, both to decrease communications

costs and also to provide local recognition for the services offered by the

Department. Strategies included rotation of staff, production of procedure manuals,

targeted training (including regular meetings), and other technology upgrades.

The CSC consists of front counter staff members who predominantly serve face-to-

face customers, a call centre taking IVR-directed calls to areas of expertise, and the

business process area which data-enters payments and manages other finance-related

issues. This model is consistent with that of other service industries (Graumann,

Arnold, & Beltjes, 2003). The CSC established a matrix of teams across functional


68

areas and regional offices. These teams are encouraged to share knowledge and

discuss decisions across their lines of responsibility.

The call-centre management software used in the central business district is

Symposium. This system displays call-queue information and generates a database

for performance statistics. It produces reports and offers real-time displays of current

status. The regional offices use a less sophisticated application, which does not route

calls based on skills and experience of the particular call-centre agent. It does,

however, collect statistics and give real-time displays of calls waiting and time-in-

queue. Management sees technology as the best option to drive efficiency while still

providing advice and management registers to support Queensland businesses and

the general public. All CSC staff has access to the Departmental Intranet, which is

based on Microsoft Team Services. The Intranet includes staff contact lists, self–

publishing capabilities, threaded discussions, and event-driven capabilities. Specific

work groups have also created sub-Webs for their individualised use.

Transaction systems include registers for the management of business details and

history of client complaints. These systems are accessible to all CSC staff. A static

database has been developed in-house to hold hierarchical lists of information,

categorised by functional area so as to allow fast access to the information.

Information is entered into this database by staff with a degree of expertise in a

particular area, and only contains statements of fact; no interpretive information is

provided. This information forms the basis for a scripted response, although it is not

kept in a dialogue format.

Reported performance measures include:

a) Response rates where the call-centre software displays and stores a range of

statistics on number of calls by type, time-in–queue, and abandonment rates.

Approximately half a million calls are processed per year.

b) Quality issues are addressed by “mystery shopping” exercises that have been

performed to test a range of criteria including the consistency and accuracy of

advice and service focus.

c) Staff satisfaction, although no measures were available. It was noted that as at

December 2002, the longest serving employee in one team had four months’

experience.


69

An attempt to embed continuous performance improvement through training has led

to the appointment of a dedicated Training Officer. This role includes usual formal

training sessions, but also includes responsibility for maintenance of the information

database. Other strategies for performance improvement include regular team

meetings and externally facilitated workshops, to build trust and shared vision

through identification of improvement strategies by the teams. Action plans

developed and reviewed on a six-monthly basis include the identification of

knowledge gaps and the processes to address these. New staff operate in a “double

jacking” configuration (a dual mode where two agents are able to concurrently

participate in the query-response process) with an experienced operator for two

weeks prior to going solo, while each operator has a buddy nominated by

management to match experienced staff with the new staff for mentoring and

support. The CC-B call centre has a formal process for monitoring call quality. Team

leaders decide on the number of calls to be sampled, and assess the agent’s

performance against seven communication skills20 and sets of technical skills based

on the functional area21 of concern to the caller. Technical requirements are tailored

to the particular functional service area, and include:

a) Search the relevant business database correctly;

b) Perform the business transaction correctly;

c) Give the correct advice;

d) Explain forms and related matter accurately;

e) Use the information database to give consistent advice; and

f) Correct the referral process when required.

The results of these evaluations are used only to provide feedback to the agent, and

are not incorporated into a whole-of-unit performance regime. CC-A calls are not

20 Self-orientation, greeting, courtesy, listening, questioning, call termination, and dealing with difficult customers are each assessed against documented criteria which do not change. 21 Eight specific functional areas are possible. Only those areas of concern to the caller can be assessed. Some agents generally develop more expertise in one area than another through previous experience and the impact of the IVR system. Each call has similar assessment criteria, such as accurate advice, use of codified knowledge, and correctly identifying the database, and search methods employed.


70

recorded, but double jacking22 is accepted as a standard practice both for training as

well as for evaluation and research.

4.6. Operational aspects of CC-B

CC-B operates within a government department whose principal business is to

maximise the economic potential for agricultural industries on a sustainable basis.

Assistance is provided to both domestic and commercial clients, with some services

being provided on a fee-for-service basis. Agents in CC-B address queries on a wide

range of scientifically-based topics. Any caller can contact CC-B for the cost of a

local call, either to a direct number or via a 1300-prefixed number.

Eight information officers (not scientists), a team leader, and the centre manager staff

the call centre. Laid out in cubicles, the call-centre operations room is quite

spacious, creating a relaxed environment for the workers. Calls are managed by a

call board and system (VU-ACD/100 for Windows), which continuously display a

range of status information on calling queues. All calls are recorded.

At the time of its establishment, CC-B’s management implemented a policy not to

use IVR technology to route calls. The policy that people, and not information

systems, will handle call diversions is designed to heighten callers’ perception of the

functional service quality (Brogowicz et al., 1990). The most significant investment

in technology is the infrastructure used to manage call-centre operations and

statistics. Continuous displays of calls in queue and average time-in queue monitor

centre performance in terms of call management. A range of databases and Intranet-

based information sets are used to support the agents. Scientists are encouraged to

provide information to the call centre to allow them to address “hot” issues.

Operators record a small amount of metadata about the conversation on the call

logging system (Figure 16), but minimal information about the resolution of the

problems raised is logged to assist future related queries.

22 Early in the research project the researcher was invited to spend as much time as necessary listening in to calls. The agents in CC-A were proud of their capabilities and seemed keen that time should be spent with them. A concern for the researcher was that such listening may influence agent response behaviours, but no evidence of changed behaviour has been identified over the three years of the study. The actual calls used as base data for the research were coded towards the end of the research period.


71

Figure 16 - Call logging system – CC-B

The Queensland Government department which hosts CC-B makes a conscious

effort to maintain its Internet site with the most up-to-date information. However,

due to issues such as the lack of access to the World Wide Web and degree of

comfort with the technologies by the Department’s principal constituents (the

agricultural sector), querists prefer to access the call centre for advice and

information. Tier-2 support is provided by scientists, who are often located at

research stations outside the capital’s central business district.

Although responses to callers are not scripted, agents are instructed not to give

advice, but rather to provide only information sourced from their standard

information systems. They are supported by a range of fact sheets and databases

populated by the scientists. These resources continually evolve through interaction

between the scientists and the call-centre agents. Agents have responsibility for

meeting with scientists and business groups to identify future events, report on issues

being addressed by the call centre, satisfaction levels with the call centre, and to

request additional or updated fact sheets from scientists and other informants within

the organisation.

The scientists were initially reluctant to support the call centre. They felt that this

communication medium with their customer base was inappropriate; that the agents

would not have the expertise to answer the queries and could mislead their customers


72

(Interview 2a - Annotated notes, May 27, 2003). When the CC-B call centre started

operations in 1995, however, scientists found it led to a decrease in “trivial” queries

normally attended to by them, allowing them to concentrate on research and other

higher value work. Today, generic calls managed by the call-centre account for up to

60 percent of calls, which would have previously been referred to scientists.

Owing to the technical and scientific nature of calls, substantial effort is placed on

recruitment and training of the call-centre staff. The key selection criteria for call-

centre agents contain the requirement to display superior functional service quality

attributes by using techniques such as active listening, patience, and “when and how

to speak” (Thompson, Warhurst & Callaghan, 2001). Superior technical knowledge

of the Department’s business minimises the time to reach acceptable levels of

competence (typically between six and eight months). Maintenance of the call-

centre staffs’ capabilities is achieved via weekly meetings at the close of operations

that allow discussion of important issues able to be identified from records of the

calls, future events, and other issues of concern.

The focus of the call centre is on client satisfaction and grade of service (in which 80

percent of calls are answered within 20 seconds). Normal days involve each operator

taking approximately 80 calls. In peak times caused by staff shortages or increased

activity from emergent events, however, CC-B can cope with up to 200 calls per

operator per day. All calls are recorded, with a library of calls dating back several

years.23These recordings are used primarily for confirmation of later queries and are

occasionally used to develop a profile of call-centre agent performance. The major

quality monitoring process, however, involves real-time evaluation of agents by the

team leader who “double jacks”24 to an operator in order to gauge the quality of

23 The recording of calls often raises the industrial issue of “spying” on staff for the purpose of management of diminished performance. It also has implications for privacy of caller information. The former issue is regarded by call-centre agents as a safety net against vexatious complaints – an indication of the maturity and confidence staff have in their capability and the trust in their peers. An often-cited example is one of a caller who complained to the relevant Minister that the call-centre agent was rude and unhelpful. A check of the recorded calls showed quite the opposite, where it was the caller who was rude and aggressive.

The second issue (privacy) is covered by advising callers that the call is being recorded, and supporting this with a strict policy of access to this recorded set. The researcher feels privileged to have been given access to this data. 24 Double jacking as explained in section 4.5 is the activity whereby a second agent can listen in to the call without being able to influence the outcome. For the purpose of the study, the researcher was only able to listen i.e. the microphone capability was turned off.


73

responses via a sampling process. The process is documented and reflects

assessments over eight (8) calls on a monthly cycle. Evaluative criteria include the

accurate capture of data about the caller in the call-centre information system, a

range of nine categories in the area of communications skills (assessing greeting,

manner, determination of client’s needs, empathy, courtesy, professionalism,

verbalisation of actions, active listening, and confirmation strategies, which are not

too dissimilar to the Brogowicz et al. (1990) model), and five categories in call

management (call control, protocols, duration, resolution, and follow-up as the

effectiveness measures used). The process for the cycle ends in a face-to-face

feedback session where the previous month’s action plans are reviewed and new

ones are agreed for the next cycle. At this meeting, a graph of “true calls per hour”

(an efficiency measure which filters out calls incorrectly routed to the agent) versus

quality assessment is discussed.

This graph (Figure 17) shows the position of all agents (anonymously, except for the

individual being reviewed), with the objective being to identify best practitioners,

and to provide a visual personal goal of moving to the top right of the represented

area. These assessments form part of the broader (whole-of-centre) performance

report produced for management.

PQ MATRIX../../00 - ../../00

Agent 2

00.5

11.5

22.5

33.5

44.5

55.5

66.5

77.5

88.5

99.510

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

PRODUCTIVITY - True Calls Per Hour

QUA

LIT

Y

Figure 17 – Agent Performance Graph (Interview 2b - Full transcript, June 16, 2004)

Performance measures which are collected and used in reporting include response

rates (grade of service and occupancy statistics most used in reporting and decision

making), quality issues (from assessment of individual agent performance), and

client satisfaction. The vision for the future of CC-B is to have 80 percent of all calls


74

completed by its staff (i.e., 80 percent knowledge self-sufficiency). In other words,

only 20 percent of calls will be consigned to second-tier support groups.

4.7. Summary

CC-A has evolved due to business pressure to cope with increasing demand for

timely and accurate advice. Efficiency continues to be the major business driver in

CC-A. CC-B has developed from a planned strategy to provide accurate and timely

information as part of an integrated approach to client management. Based on a

personalised service and supported by training and coaching, the approach is to

protect clients from the technology driving the call centre. Table 10 summarises the

strategies employed in each of the call centres.

Strategy CC-A CC-B

Use of Integrated Voice Response / Call Routing

Yes No

Call Rate Up to 200 calls per person per day

80 calls per person per day

Information recorded on calls

Minimal – for reporting and return contact purposes

Minimal – for reporting and return contact purposes

Updating of information repository

Not involved Meetings with scientists / 2nd Tier operatives

Internet site Limited information available All information available

Consignment Policy Research often performed by agent prior to consignment

Consignment if query cannot be resolved with standard information

Level of remuneration for operators

Government clerk level 2 & 3 Government clerk level 4

Training 2 weeks 6-8 months

Table 10 - Summary of the Strategies Employed by Both Call Centres

The two call centres studied, although both within the Queensland public sector,

reflect their own unique business issues and underpinning service principles. Each

can be easily mapped onto the Q-R Cycle (Table 11), and provides clear performance

expectations of the call-centre agents. Customer access criteria contribute

significantly to decisions and strategies employed. The IVR system of CC-A has

been designed to direct callers to clusters of expertise within the call centre, and

hence reduces the need to refer on to another agent. Intelligent routing technology

has also been employed to allow regional callers to connect with local agents who


75

may have a better understanding of local issues. CC-B has employed a standard-of-

service measure weighted strongly in favour of the reduction of waiting time-in-

queue. A premium has been allocated to this, owing to the need to allow callers to

develop a relationship with the agent, particularly when complex issues are being

addressed. Value-for-money at CC-B is reflected in the proportion of calls able to be

resolved without referral to (relatively more expensive) experts who are more

usefully occupied in other value-adding roles. CC-A had no benchmark available, but

as demand for their services has increased, the marginal costs of servicing this

increase have been able to be absorbed into the accepted funding model. A

framework to analyse these observations is developed further in the next chapter.

Phase of Q-R Cycle CC-A Strategy CC-B Strategy

Accept Query Caller directed to functional experts via IVR. Average time-in-queue 73 seconds (target of 60 seconds) – see Appendix 2.

All calls are accepted by generalist agents. Average time-in-queue < 20 seconds.

Query Reformulation Training provided in introduction, listening and paraphrasing.

Training provided in introduction, listening, and paraphrasing.

Search Strategy/ Knowledge Sources

Dedicated database with script support is the primary knowledge source.

Range of knowledge sources including the Internet, fact sheets, and specific databases.

Response Set Presented factually but embellished with context provided by the caller to create the link between the regulatory basis for the advice and the situation described by the caller – may be taken off-line for further research and more formal response.

Responses are highly tuned to client situation and expectations. Referral to experts is done with handover briefing to maintain client care.

Pay-off Assessment/ Knowledge Distance Adjustment

Unless extended by further caller queries, the call is terminated once advice has been given.

Significant proportion of time spent prior to termination in ensuring the outcome is understood and the caller is satisfied.

Table 11 - CC-A, CC-B, and the Q-R Cycle

A Framework for Analysis

76

Chapter 5


5.1. Introduction

The case studies introduced in the previous chapter have identified the work,

organisation, and processes used by the study’s call centres. Actual problem

resolution methods are not clear from these observations, and a further level of

analysis has been performed using Zack’s (2001) Four Knowledge Problem Model.

This two-dimensional model is simple to apply, but powerful in its ability to

categorise behaviours of call-centre agents. The analysis of these behaviours and

comparison with strategies recommended by Zack (2001) will allow both validation

of the model and the identification capability of gaps and redundant processes.

5.2. Zack’s (2001) Four Problem Model

The problem of “uncertainty … represents a lack of information, or factual

’knowledge about’ current and future states, preferences and appropriate actions”

(Zack, 2001, p. 20). CC-B recognises the high proportion of problem types which

fall into this category, and has developed processes to maintain up-to-date electronic

and paper-based information sources. These are designed and populated by experts

(scientists), so that callers with knowledge problems of uncertainty are generally able

to be satisfied quickly. Development of intellectual capabilities within CC-B to

predict, infer, estimate, and learn (as part of the training strategy), assists in the

management of uncertainty. CC-A has a similar strategy for maintenance of

electronic information systems (also populated by experts), but does not replicate this

information on the Internet. The IVR system also contributes to the efficiency of

CC-A, since it streams callers directly to functional experts who, through their

experience in dealing with specific business problems, are able to recall factual

details quite independently of the electronic systems.

Zack (2001, table 1) suggests the response to complexity is either to “increase a

firm’s capacity to process it or to reduce the level of complexity.” Both CC-A and

CC-B commence dialogue with the caller by following an initial scoping-questioning

technique. CC-A takes longer (due to the need to respond to the IVR prompts), but

directs callers with complex problems to the more capable staff within the team. The


77

superior, yet more complex, IVR programming provides this capability. The system

generates monthly call-time reports for CC-A staff. Their objective is to minimise

call time. The impact of this is that the time available to decompose complex

problems during the call is reduced. Once the CC-A agent decides time will limit

resolution, the problem is consigned either to tier-2 specialists, or it may be set aside

for further research by the agent at a later time. CC-B operators do not specialise in

particular functional areas, and are not considered to be fully trained until they have

undertaken several months of training and mentoring. The operator’s new-found

knowledge increases the call-centre’s capacity to deal with complex problems.

Higher levels of remuneration also increase the likelihood that more talented and

experienced operators will be both attracted to the call centre and will stay on longer.

In addition, since call time is not a critical success factor in CC-B, extra resources

can be committed in real time to decomposing complex problems. The option exists

for CC-B agents to consign to second-tier experts, but the goal to reduce referrals

encourages resolution within the call centre.

Provision of rich interactive conversation is the key capability in management of

ambiguity, according to Zack (2001). CC-B has addressed this in its strategy of

providing ample time for call-centre operators to discuss the problem with the caller,

and to attempt to reframe the problem into a scenario that is more meaningful. The

agent is able to repeat several cycles of interpretation and explanation until either the

problem becomes resolvable (complex or uncertain), or is consigned to an

appropriate scientist whose education and experience provide the skills to solve such

problems. CC-A agents do not have the time to perform this type of dialogue, and

the consignment policy of postponing it to resolution by off-line research does not

allow for reframing, interpretation, or explanation.

The multiple meanings and interpretations which are characteristic of problems of

equivocality require negotiation to converge on one meaning (Zack, 2001). With its

policy of generous time allocation for agent dialogue with callers, combined with

capable staff, CC-B is well placed to resolve such problems. The knowledge strategy

employed is not one of access to information, but rather one of process which

provides the capability to cycle interactive discussion, negotiation, and interpretation

until a resolution is achieved. The IVR system used in CC-A may become a liability

with problems of equivocality, since the confusion of meaning may even commence


78

with the initial options presented to the caller. This may end in an incorrect referral

to an agent who may have preconceived ideas on the context of the call. These

incorrectly directed calls are generally addressed quickly either through

abandonment or further referral to a more appropriate tier-2 agent.

CC-B has the goal of achieving 80 percent call-centre self-sufficiency. The reporting

and analysis of consigned calls serves as a trigger to identify strategies to reduce

them. Past strategies have included requests for fact sheets from scientists, briefings

by specialists, and incorporation of contemporary topics into formal training

sessions. CC-B also has sufficient redundancy in its resources to maintain a grade of

service which is interpreted as zero abandoned calls. In times of high load due to

staff illness or a significant event, calls can be lost. These lost calls are not analysed

due to the infrequent occurrence of such events. CC-A has a complex call-routing

system, which attempts to locate available local resources, and progressively skips to

the next available agent. This is also supported by an IVR system to stream callers

vertically into functional pools based on their business need. In this process, many

calls are abandoned. Reporting in CC-A includes (by business function) total

number of calls, number of referrals beyond the call centre, proportion of calls settled

with advice, and the percentage of all calls sampled for quality monitoring. The

analysis of this data is primarily used to identify the division of business functions

and the proportion of effort directed to this. It is also used to pre-empt future client

needs and educational or promotional requirements.

Within both call centres, the majority of problems fall within the

uncertainty/complexity problem type. This means that knowledge strategies based on

codified knowledge sources allow agents to converge quickly to a viable response

set. The effort put into the development of fact sheets and databases serves this

requirement well, since uncertainty is removed through access to current knowledge,

while complexity is reduced through online searching and supported by access to

experts via referral. CC-A, however, has a significant proportion of callers whose

problems fall within the ambiguity problem type. These generally emerge through

the IVR system as a complaint and are managed by the more experienced team

members. Such calls tend to take much longer to resolve and involve either referral

to experts or are managed off-line due to the requirement for further research.


79

Another issue of this dispersed agent model has been identified by Brooke (2002).

Workers not involved in the physical call centre (e.g., located in regional offices), but

who receive IVR-directed calls may not share the knowledge and meanings generally

agreed by the core group in the head-office environment. The continuous payoff

assessment by callers for these problem types generally demands much more

resource allocation, since the problems often have a personal/corporate financial

consequence for the caller. Staff working in this field do receive higher pay, but are

also subject to much more stress, both due to the intellectual challenges of the

problem, and also to the awareness of the need to reduce time per call (an efficiency

measure used both at an individual level as well as at a team level).

Analysis of CC-B calls where the researcher tracked the IVR path of a caller,

suggests that the IVR system effectively directs callers to the relevant area of

expertise. The message that greets the caller provides options for the major business

functions, which predominantly reflect uncertain problem types. The IVR system

routes the call to agents with expertise to resolve the problem, either through recall or

access to codified knowledge. Other options route the predominantly complex (due,

mainly, to legislative interpretation) and ambiguous problems to agents who employ

a range of resolution strategies in order to achieve successful problem resolution.

A F

ram

ework

for

Anal

ysi

s

Ta

ble

12

- S

um

ma

ry o

f K

now

led

ge

Str

ate

gie

s U

sed

in

th

e C

all

Cen

tres

by

Pro

ble

m T

yp

e

Zack

’s (

2001)

Kn

ow

led

ge

Pro

ble

m T

yp

es

Str

ate

gy

Call

C

entr

e U

nce

rtain

ty

Com

ple

xit

y

Am

big

uit

y

Eq

uiv

oca

lity

CC

-B

No

IV

R

No

IV

R

No

IV

R

No

IV

R

Use

of

Inte

gra

ted

V

oic

e R

esp

onse

/

Cal

l R

outi

ng

CC

-A

Po

siti

ve

imp

act

- in

itia

l cl

assi

fica

tio

n b

y c

alle

r p

rovid

es q

uic

k a

cces

s to

te

chnic

ally

cap

able

ag

ents

Pro

bab

le n

egat

ive

imp

act

due

to t

he

call

er

nee

din

g t

o s

elec

t fr

om

p

red

eter

min

ed o

pti

ons.

Pro

bab

le n

egat

ive

imp

act

due

to t

he

call

er n

eed

ing t

o

sele

ct f

rom

p

red

eter

min

ed o

pti

ons.

Pro

bab

le n

egat

ive

imp

act

due

to t

he

call

er n

eed

ing t

o

sele

ct f

rom

p

red

eter

min

ed o

pti

ons.

CC

-B

Po

siti

ve

imp

act

- al

low

s ti

me

for

clar

ific

atio

n b

y

acq

uir

ing a

dd

itio

nal

fa

cts

Po

siti

ve

imp

act

- al

low

s ti

me

for

pro

ble

m

dec

om

po

siti

on a

nd

re

buil

din

g

Po

siti

ve

imp

act

- al

low

s ti

me

to r

efra

me

a p

rob

lem

si

tuat

ion i

nto

so

met

hin

g

mo

re m

eanin

gfu

l

Cal

l R

ate

CC

-A

Po

siti

ve

imp

act

- as

sist

ed b

y

IVR

, ca

ller

s ar

e d

irec

ted

to

sit

uat

ional

kno

wle

dge

exp

erts

Neg

ativ

e im

pac

t -

pre

ssure

to

lim

it c

all

tim

e im

pac

ts o

n a

gen

t’s

abil

ity t

o i

den

tify

the

inte

rrel

ated

var

iab

les,

so

luti

ons

and

met

ho

ds

Neu

tral

sin

ce t

he

stra

teg

y t

o

carr

y o

ut

furt

her

re

sear

ch o

ff l

ine

is n

ot

fact

ore

d i

nto

cal

l ra

te

stat

isti

cs

Neg

ativ

e im

pac

t -

the

con

flic

ts

cause

d b

y m

ult

iple

in

terp

reta

tio

ns

req

uir

es

sig

nif

ican

t ti

me

to

conver

ge

on a

shar

ed

pro

ble

m s

tate

men

t

80

A F

ram

ework

for

Anal

ysi

s

Zack

’s (

2001)

Kn

ow

led

ge

Pro

ble

m T

yp

es

Str

ate

gy

Call

C

entr

e U

nce

rtain

ty

Com

ple

xit

y

Am

big

uit

y

Eq

uiv

oca

lity

CC

-B

Po

siti

ve

imp

act

- sm

all

imp

act

since

all

ow

s an

alysi

s o

f tr

end

s to

sup

po

rt o

ther

kno

wle

dge

stra

tegie

s

Nil

Po

siti

ve

imp

act

- ca

ller

det

ails

al

low

s ag

ents

to

re

vis

it/f

orw

ard

cl

arif

yin

g i

nfo

rmat

ion

Nil

Info

rmat

ion

reco

rded

on

call

s

CC

-A

Po

siti

ve

imp

act

- sm

all

imp

act

since

all

ow

s an

alysi

s o

f tr

end

s to

sup

po

rt o

ther

kno

wle

dge

stra

tegie

s

Nil

Po

siti

ve

imp

act

- gen

eral

ly

add

ress

ed o

ff l

ine

wit

h

form

al n

ota

tio

n o

f re

sear

ch i

ssues

, b

ut

no

t li

nked

to

the

call

-cen

tre

IS.

Nil

CC

-B

Po

siti

ve

imp

act

- ro

bust

and

cu

rren

t re

po

sito

ry e

asil

y

acce

ssed

by a

gen

ts

Po

siti

ve

imp

act

- sm

all

imp

act

since

so

me

fam

ilia

r ro

uti

nes

are

d

ocu

men

ted

Nil

N

il

Up

dat

ing o

f in

form

atio

n

rep

osi

tory

CC

-A

Po

siti

ve

imp

act

- ro

bust

and

cu

rren

t re

po

sito

ry e

asil

y

acce

ssed

by a

gen

ts

Po

siti

ve

imp

act

- b

oth

o

per

atin

g p

roce

dure

s an

d h

euri

stic

s ar

e co

ded

in t

he

call

-ce

ntr

e IS

.

Po

siti

ve

imp

act

- P

arti

cula

rly

hel

pfu

l in

reg

ula

tory

ar

eas

Nil

CC

-B

Po

siti

ve

imp

act

- ex

ten

sive

and

curr

ent

site

eas

ily

acce

ssed

by b

oth

cal

lers

an

d a

gen

ts

Nil

N

il

Nil

Inte

rnet

sit

e

CC

-A

Neu

tral

– n

ot

a m

ajo

r so

urc

e o

f use

ful

fact

ual

in

form

atio

n

Nil

N

il

Nil

81

A F

ram

ework

for

Anal

ysi

s

Zack

’s (

2001)

Kn

ow

led

ge

Pro

ble

m T

yp

es

Str

ate

gy

Call

C

entr

e U

nce

rtain

ty

Com

ple

xit

y

Am

big

uit

y

Eq

uiv

oca

lity

CC

-B

Po

siti

ve

imp

act

- m

ajo

rity

re

solv

ed i

n f

irst

cal

l w

itho

ut

refe

rral

; in

form

atio

n s

up

pli

ed b

y

exp

erts

do

cum

ente

d i

n

fact

shee

ts

Po

siti

ve

imp

act

- re

ferr

al

net

wo

rk t

o e

xp

erts

b

ased

on f

unct

ional

ar

ea

Po

siti

ve

imp

act

- re

ferr

al

net

wo

rk t

o e

xp

erts

bas

ed

on f

unct

ional

are

a

Po

siti

ve

imp

act

- re

ferr

al

net

wo

rk t

o e

xp

erts

bas

ed

on f

unct

ional

are

a

Co

nsi

gn

men

t P

oli

cy

CC

-A

Po

siti

ve

imp

act

- m

ajo

rity

re

solv

ed i

n f

irst

cal

l w

itho

ut

refe

rral

; ac

cess

to

exp

erts

thro

ugh

kno

wle

dge

of

org

anis

atio

nal

un

its

Po

siti

ve

imp

act

- re

ferr

als

to

mo

re e

xp

erie

nce

d s

taff

w

ithin

the

call

-cen

tre

Po

siti

ve

imp

act

- sp

ecif

ic u

nit

s es

tab

lish

ed t

o d

eal

wit

h

refe

rred

cli

ents

(b

ased

o

n c

onsi

gn

men

t ru

les)

Po

siti

ve

imp

act

- sp

ecif

ic u

nit

s es

tab

lish

ed t

o d

eal

wit

h

refe

rred

cli

ents

(b

ased

o

n c

onsi

gn

men

t ru

les)

CC

-B

Po

siti

ve

imp

act

- re

lati

vel

y

hig

h t

o a

ttra

ct a

nd

ret

ain

inte

llec

tual

res

ourc

es

Po

siti

ve

imp

act

- re

lati

vel

y

hig

h t

o a

ttra

ct a

nd

re

tain

in

tell

ectu

al

reso

urc

es

Nil

N

il

Lev

el o

f re

mu

ner

atio

n

for

op

erat

ors

CC

-A

Neg

ativ

e im

pac

t -

rela

tivel

y

low

; hig

h t

urn

over

m

eans

loss

of

inte

llec

tual

res

ourc

es

Neg

ativ

e im

pac

t -

rela

tivel

y

low

; hig

h t

urn

over

m

eans

loss

of

inte

llec

tual

res

ourc

es

Neg

ativ

e im

pac

t -

exp

ecta

tio

n

of

com

ple

tio

n o

f o

ff l

ine

rese

arch

, al

tho

ug

h

pro

vid

ing v

arie

ty,

also

cr

eate

s st

ress

Nil

82

A F

ram

ework

for

Anal

ysi

s

Zack

’s (

2001)

Kn

ow

led

ge

Pro

ble

m T

yp

es

Str

ate

gy

Call

C

entr

e U

nce

rtain

ty

Com

ple

xit

y

Am

big

uit

y

Eq

uiv

oca

lity

CC

-B

Po

siti

ve

imp

act

- fo

cuse

s o

n

kno

wle

dge

acq

uis

itio

n,

info

rmat

ion s

earc

h

met

ho

ds

and

ser

vic

e q

ual

ity c

rite

ria

Nil

N

il

Nil

Tra

inin

g &

m

eeti

ngs

CC

-A

Po

siti

ve

imp

act

- fo

cuse

s o

n

kno

wle

dge

acq

uis

itio

n,

info

rmat

ion s

earc

h

met

ho

ds

and

ser

vic

e q

ual

ity c

rite

ria

Po

siti

ve

imp

act

- ag

ents

en

coura

ged

to

uti

lise

m

etho

ds

to r

eso

lve

com

ple

x p

rob

lem

s

Nil

N

il

83


84

5.3. Summary

Neither of the call centres studied employs a knowledge strategy based on any overt

theoretical framework. The strategies employed have evolved due to the current

business drivers and continue to develop based on the historically bureaucratic

organisational structure designed to deal with complexity. However, the process of

participation in this study indicates that each organisation is prepared to develop a

capability to handle a broader set of knowledge problems. Table 12 provides a

summary of the impact of current knowledge strategies,25 and identifies the gaps that

need to be addressed if the model proposed by Zack (2001) applies. This summary

highlights the general practice of resolution of complex/uncertain problem types by

the call-centre agent, and a referral system for those aligning with Zack’s (2001)

ambiguous/equivocal types.

Zack’s (2001) four-problem framework has been used to analyse the policies and

actual practices of the call centres, CC-A and CC-B. Cross et al. (2005) suggest that

call centres use routine response social networks, since these are best suited to

situations where problems and solutions are fairly predictable and resolved with a

degree of collaboration. This is mirrored in the strategies espoused by CC-B and

CC-A, which have a robust tool set for resolving uncertain and complex problems.

Methods of attack for ambiguous and equivocal problems are not well supported

within the call centres, but consignment policies do exist to address them. In Chapter

Six, a closer investigation of call archetypes and their responses reveals patterns of

response strategies used by the agents.

25 A nil entry in any particular cell may mean either that the strategy is not appropriate for that problem type, or that the particular call centre does not employ that strategy.

Patterns from Observations, Interviews, and Responses

85

Chapter 6


Qualitative research relies on patterns and commonalities extracted from rich case

data; however, the sensitive interpretations from this complex data are difficult to

represent in a numeric way. The seemingly infinite and unrelated data elements

originally encountered in the study of the two call centres have been used to develop

a new understanding of the organisational behaviours exhibited in this micro-

community of the Queensland public sector. A forms-driven database has been used

to encode calls, and a qualitative data analysis tool called NVivo has provided the

capability to link recurring concepts in interviews, observations, literature, and

individual call conversations. The outputs from these tools allowed the management

and synthesis of ideas into well-defined patterns that have emerged from the

relatively unstructured primary data sources. The relationships between components

of this research activity are shown in Figure 18.

Management Views

Literature Search

CC1 Management

CC2 Management

PS Call Centres

Quality Principles

KM principles

Data Collection

Ancilliary documents

Method

PhenomenographyYin

Analysis

Nvivo

Research Structure

Coded Calls by Zack Type

CC1Leader_Interv iew

CC2Leader1_Interview CC2Leader2_Interview

Figure 18 - Research components


86

6.1. Executive Management Interviews

Data has been collected from a wide variety of sources. Early interviews with

Executive Managers were used to:

a) Identify the business drivers for the call centre;

b) Gauge the Executive Managers’ degree of familiarity with the principles

which underpin knowledge management; and

c) Obtain a commitment from the Executive Managers to support the research

program (by providing access to the staff, operations, and statistics).

Both the desired and the actual practices and procedures in operation were

documented during the cycles of meetings with call-centre managers, team leaders,

and staff that followed the interviews. During this period, calls were monitored in

real time,26 and retrieved from archival systems to develop a complete picture of the

interrelationships among management expectation, policy compliance, procedural

activity, and individual responses. The theories of knowledge reuse and their

application to repository design and intervention postulated by Markus (2001)

extended the work of Dixon (2000), who investigated criteria for successful

knowledge transfer. The roles of shared-work producers, shared-work practitioners,

expert-seeking novices, and secondary data miners, and their contributions and

interactions with repositories formed the basis of the structured interviews with

managers and team leaders.

6.2. Management and Team Leader Interviews

The role of repositories is of interest to contemporary knowledge managers

(Davenport & Prusak, 1998; Markus, 2001). Of interest to the later analysis of calls

is the resolution strategy employed by agents. In many cases this involves identifying

an appropriate information source, which leads to resolution of the particular query.

The postulate that knowledge reuse (the combined search strategy, information

source, and application in context) improves call-centre performance is central to this

study. In interviews with management and team leaders, at least 64 references were

26 Staff were quite comfortable with having the double jack arrangement, since it is used for training, performance monitoring, and the promotion of the work done by call-centre agents.


87

made to knowledge repositories. These have been coded as External Knowledge,

Structured Internal Knowledge, or Informal Internal Knowledge repository types

(Davenport & Prusak, 1998, p. 146), as shown in Table 13.

Table 13 - Frequency of Repository References in Interviews

Call-centre agents and supervisors have identified the knowledge sources that

contributed to these repository types as including:

External: “Library; Web sites of other Government Departments; general Internet

sites; legislation; and referral indexes”;

Informal Internal: “palm notes; book where they’ve got common contacts and stuff;

email; daily news; just a matter of writing it down; we keep half folders of

information; tagged and marked so they can readily refer to things; some

will store things electronically”; and

Structured Internal: “Knowledge Base; Infotech; Brands database; telephone listing;

Call Centre information system; fact-sheets; flowchart procedures on how

to answer the calls; printed documents; Inform database system; Intranet;

training manuals; handbooks” (Interview 1a - Annotated notes, January 29,

2003; Interview 1b - Annotated notes, December 23, 2002; Interview 1c -

Full transcript, June 9, 2004; Interview 1d - Full transcript, June 11, 2004;

Interview 2a - Annotated notes, May 27, 2003; Interview 2b - Full

transcript, June 16, 2004).

The roles undertaken within the call centres have different responsibilities for the

development, access, and maintenance of these repositories. Team-leader roles

across both agencies have aligned extremely well with Markus’s (2001) shared-work


88

producers, who generate knowledge for their own later reuse. Their title reflects their

primary purpose of the development of an effective operational environment. Team

performance has been mentioned regularly in the interviews, with clear operational

performance measures being defined, collected, and reported. The following extracts

from interviews with team leaders indicate their expectation that agents be aware of

performance targets:

“And they can see what they do for the day, for the week. They can

see what their peers do for the week. And in a way, it's sort of

y'know, encourages them to do that little bit better too.”

"If, see Symposium tells us, for individuals, it tells us how many calls

they’ve taken, what their talk time is. I actually had one girl in my

team that was taking like 50 calls a day, but her talk time was four

minutes."

"And our target is three minutes and under. So I actually sat with her

and I also sat at my desk and listened in on her calls. She knew I was

doing it. Listened in on her calls and then at the end of it, I went

through it with her and I said “Well y'know.” She had a problem, her

customer service was too much."

"And most staff actually know their stats by the end of the day

themselves."

"We refer to it [performance statistics] regularly and report on it. I

report on it weekly and monthly."

"And I give feedback weekly to the team and that’s usually on the

number of calls received per day, per week. The average wait-time."

This is consistent with findings in literature which indicate a predominance of

measurement against benchmarks by technology solutions (Anton & Gustin, 2000).

Easily measured and reported, strategies have been applied to ensure individual

performance is within the acceptable boundaries for the agency. The mentoring role,

although not referred to in the interviews, was a significant component of call-centre

staff’s daily work noted in real-time observation. The team leaders have been

generally regarded as a source of general and contextual knowledge that relates to the

contemporary issues of the time. This reuse situation comes from the historical


89

experiences of the team leaders, who have referred to the development of the call-

centre practices and procedures and reasons behind such actions.

The following passages reflect the characteristics of shared-work producers

described by Markus (2001, p. 65) that, “Users need to know what is done, how it

was done, why it was done and how it can be done better”.

An understanding of the single view of Queensland Government illustrates why

things are done in a particular way:

"And it's not only knowledge outsiders who ask what happened within

[the host agency], we have to have a broad knowledge of what’s

happening in the other Government Departments as well."

Also the intuition exists to maintain historical information for personal use, but

shared with other agents including both what and why a decision was made:

"And that’s what we do. Cause I mean a couple of years ago we had

something, …. I kept the number because..."

Practices are established to identify how things are done and where

improvement opportunities exist:

"Team leaders perform coaching mentoring role and are the

information expert. They also produce performance statistics and are

able to report by subject by day."

How things were done in the past is passed on by the experienced staff to later

recruits:

"Some of us might know more information, cause we’ve been here

longer. So yes, we’re going to have more information, but you don’t

keep that information to yourself."

"We’ve got people who build up expertise over time. Some of the

people have information that goes back a fair way, more so than

someone who’s just joined us. And they’re often relied on as the

knowledge experts and the information is gleaned from them."

Although these examples support the Markus (2001) model of shared-work

producers, some characteristics of shared-work practitioners are also displayed. This


90

is to be expected, since shared-work practitioners produce knowledge for sharing

with the team and leaders drop into and out of the agent role from time to time. They

have a strong affinity with the team, and hence do not see their role as being outside

that team. Call-centre agents generally fall into the shared-work practitioner

category, and “acquire new knowledge that others have generated,” and “get advice

about how to handle a … situation that is new to the team” (Markus, 2001, p. 64).

The only overt strategy used to support this predominant role in both call centres has

been the regular staff meeting/training activity. However, this has not significantly

contributed to the production of knowledge for the use of other agents. Since these

small call centres have been designed in a way that allows personal interaction

between agents, having access to staff with tacit knowledge is an efficient process

(when compared to codification strategies) (Hansen et al., 1999). This strategy has

simply evolved, and has been taken for granted. CC-B managers have referred,

however, to a redundancy strategy27 in their scheduling, which arguably is a

necessary precondition for knowledge transfer (Davenport & Prusak, 1998).

Emphasis has been placed on team harmony, with the consequence that agents noted

the mutual benefits associated with knowledge sharing. Markus (2001) refers to the

producer, intermediary, and consumer roles in the knowledge reuse process and notes

that each role “can be performed by the same individual or group, different

individual or group, or some combination” (2001, p. 61). The contribution of both

individuals and groups within the call-centre team are valued on the assumption that

it provides benefit to call-centre performance, as seen in the comments below:

“It's the culture of the call centre. The culture that’s been fostered here is just

like a family.”

“If you don’t have it, you know for a certainty that somebody in the call

centre will.”

“Oh yeah, I had this fellow was talking about this and y'know and I found this

out. And then they just, a lot of it is word of mouth between themselves.”

27 Redundancy implies a strategy that allows staff time-out from the routine of work to communicate with their peers, allowing knowledge transfer to occur due to a shared/common view of that knowledge. The value of conversation around the water fountain is an example.


91

“I always not only tell them the answer, I show them where in the legislation

they can find that answer. And next time they don’t have to come back or

next time they can show somebody elsewhere to find it as well.”

“… it's still shared initially anyway. So if [agent named] finds a really good

site or something, she’ll share it with the team and then it's up to the team

what they do with it.”

“… [An agent] will just stand up and say, ‘Hey, can anybody, Does

everybody know this, that and this?’”

“Or if we get something come through at the counter that people aren’t aware

of then we send it by email to all staff.”

“Staff talk all the time amongst themselves and luckily, we’re very lucky that

we have a lot of staff that really get on well together.”

“So we have a meeting with them once a fortnight and any issues that are

raised by Team Leaders from their teams at team meetings.”

“It’s just a matter of writing it down and if it's something new they think

needs recording or discussing with colleagues, they’ll do that. And perhaps

do that at the end of the day.”

“A lot of people often go to her because those people who store things

electronically may not know where it is, but [agent named] can put her finger

on it with the hard copy.”

These examples illustrate the impact of social networks. Cross et al. (2005) provide a

framework for the identification and nurturing of informal networks in a productive

fashion. Such a knowledge strategy which aligns technology, processes, and human

capability, already fundamentally, yet informally, exists in the call centres studied.

Management intervention that encourages these collaborative networks to flourish

would improve call-centre performance in areas of self-sufficiency.

The goal of 80 percent self-sufficiency in CC-B and the strategy of taking tier-2

work off-line to research stretch the shared-work practitioner role to one of expertise-

seeking novice for call-centre agents, particularly with the goal of minimising the

need to access the more expensive tier-2 experts. The majority of actors in the call

centre defined as expertise-seeking novices are, however, the callers who are seeking


92

to develop new knowledge about unfamiliar problems, or are unable to contextualise

correctly the issues they are investigating. Calls made to each of the call centres have

been monitored to understand both the processes and interactions of the agents, as

well as the nature of the incoming queries.

Calls examined early in the process28 have been used to test the coding methodology

and to train a research assistant in its application. Archived calls for the final analysis

were randomly selected on the basis of 25 calls for each of the six-month periods

from 1 January 2000 to 30 December 2004, giving a total of 200 calls. They were

retrieved by call-centre staff from recordings and reconstructed to enable them to be

copied to compact disc.

The ability to take them off site for analysis has had several advantages, including:

a) No interruption to the operations of the call centre;

b) Ability to double-code the calls to ensure consistency of approach; and

c) The easy revisitation of individual calls for further detailed analysis.

The major disadvantage encountered in this approach has been the error rate in the

reconstruction process. This has occurred as a result of the recording system used to

archive calls using a time-slicing algorithm, and hence has recorded calls as a block.

Calls that were incomplete at the end of that time-slice have in some cases been

difficult to re-link, and hence have had to be discarded from the sample.

Of the calls selected, 143 have been incorporated into the final analysis for this

research. Rejected calls generally fall into categories of incomplete29 or incorrectly

routed, often from within the host agency. The latter category of calls falls outside

the scope of this research. Based on strategies developed by Di Gregorio (2000), the

output from the database of calls was automatically coded using NVivo, a tool with

search and coding functionality based on predetermined criteria.30 Such coding has

allowed information collection relating to the problem type, resolution, and quality

themes to be explored. This is primarily due to the structured nature of the output

28 These calls did not form part of the final analysis. 29 Due to a quirk in the call-centre recording software, some calls became fragmented in the process of copying them to CD ROM for off-line analysis. 30 Future research could utilise these processes to encode much larger information sets giving a higher degree of confidence and a capability to perform more rigorous statistical validation.


93

from the original data-collection tool. Each line refers to a particular call which

already has had the predefined codes for problem type, resolution strategy, and

quality. The algorithm used to code problem types automatically has involved

searching for a particular code and allocating the associated node to the whole line

which due to the coding process equates to a paragraph for the purpose of NVivo

analysis.

For response and quality codes, only the particular code was associated with the

relevant node, since the unique codes for these could be used not more than once for

each call. This has allowed frequency analysis to be presented more easily, since it

ties the quality and resolution strategies to the problem type, rather than to the unique

call identifier.

Transcripts of interviews and case notes from direct observation have also been

coded using this tool. Each document has been imported into NVivo and manually

coded to ensure that the context of the relevant node is captured. NVivo's search tool

has been used to identify combinations of themes (such as knowledge reuse and

strategic position of the call centre) during interviews. Simple searches referring to

the predetermined nodes have revealed references to areas of consideration. The

Boolean matrix intersection and difference searches provide the composite tables

(Table 14) used below to identify the patterns of interest for the actual calls.

Pat

tern

s fr

om

Obse

rvat

ions,

Inte

rvie

ws,

and R

esponse

s

94

R

esp

on

se S

trate

gie

s f

rom

Co

ded

Call

s

Zack T

ypes

Cyc

lic

al

refr

am

ing

D

e-

co

mp

ose

Me

an

ing

th

rou

gh

d

isc

uss

ion

R

eq

uir

es

re

se

arc

h

So

cia

l c

on

str

uc

tio

n

Te

rmin

ate

C

all

To

ok

P

ers

on

al

De

tail

s

Tra

ns

fer

to

us

er

Tra

ns

fer

to

Fu

nc

. A

rea

Us

e

co

dif

ied

in

form

ati

on

T

OT

AL

C

all

s

Am

big

uit

y

7

0

14

1

2

1

1

7

2

4

15

Co

mp

lex

ity

5

9

11

2

0

0

8

8

3

7

17

Eq

uiv

oc

al

0

0

0

0

0

0

0

0

0

0

0

Un

ce

rta

inty

9

1

63

1

3

2

52

28

37

63

111

Am

big

uit

y

46.7

%

0.0

%

93.3

%

6.7

%

13.3

%

6.7

%

6.7

%

46.7

%

13.3

%

26.7

%

10.5

%

Co

mp

lex

ity

29.4

%

52.9

%

64.7

%

11.8

%

0.0

%

0.0

%

47.1

%

47.1

%

17.6

%

41.2

%

11.9

%

Eq

uiv

oc

al

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n

/a

Un

ce

rta

inty

8.1

%

0.9

%

56.8

%

0.9

%

2.7

%

1.8

%

46.8

%

25.2

%

33.3

%

56.8

%

77.6

%

Q

uality

Cri

teri

a f

rom

Co

ded

Call

s

Zack T

ypes

Clo

su

re

Sa

tis

fac

tory

D

iffi

cu

lt t

o

ac

ce

ss

E

as

e o

f a

cc

ess

Em

pa

thy

wit

h

cu

sto

me

r

Ex

ce

ss

Tim

e-i

n-

qu

eu

e

No

Em

pa

thy

wit

h

cu

sto

me

r C

all

Ba

ck

Te

ch

nic

al

res

ou

rce

c

ap

ab

ilit

y

Te

ch

nic

all

y

co

mp

ete

nt

Un

-s

ati

sfa

cto

ry

clo

su

re

TO

TA

L

Ca

lls

Am

big

uit

y

10

0

0

14

0

0

0

0

3

1

15

Co

mp

lex

ity

11

1

0

16

1

0

1

6

7

0

17

Eq

uiv

oc

al

0

0

0

0

0

0

0

0

0

0

0

Un

ce

rta

inty

71

0

3

90

0

2

12

31

36

2

111

143

Am

big

uit

y

66.7

%

0.0

%

0.0

%

93.3

%

0.0

%

0.0

%

0.0

%

0.0

%

20.0

%

6.7

%

10.5

%

Co

mp

lex

ity

64.7

%

5.9

%

0.0

%

94.1

%

5.9

%

0.0

%

5.9

%

35.3

%

41.2

%

0.0

%

11.9

%

Eq

uiv

oc

al

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n

/a

Un

ce

rta

inty

64.0

%

0.0

%

2.7

%

81.1

%

0.0

%

1.8

%

10.8

%

27.9

%

32.4

%

1.8

%

77.6

%

Ta

ble

14

- F

req

uen

cy o

f R

esp

on

se S

tra

teg

ies


95

6.3. Responses to Calls

Uncertainty

The predominant problem type presented to the call centres falls into the category of

uncertainty, as defined by Zack (2001). Such problems are characterised by lack of

knowledge about current and future states. Often the caller simply states, “I am

looking for information about XYZ” at the commencement of the call. This generally

indicates an uncertain problem type. Not all calls are so easily categorised, however,

and in some cases this simple lead-in statement points to another problem not so

easily resolved. A typical uncertain call follows the following pattern of this

transcript of an actual conversation:

Caller: “We’re on a property and we want to get into wild flowers growing. Is

there any information that you can send me that will help me out?”

Agent: “I do have some information that’s on wild flowers – it will depend on

what species you are talking about. Any in particular that interests you?”

Caller: “Oh – No, nothing really, we just sort of rung because we want to make

it sort of work for us – that’s all …. and somebody suggested wild flowers you

know they seem to be quite in the export you know side of it….”

The sample of calls contains 111 with uncertain problem types. Significant strategies

employed by call-centre agents in answering these calls are enumerated below:

a) Meaning through discussion (57 percent) is used primarily to clarify the real

needs of the caller (Weick, 1995). This strategy also links to the quality

characteristic of empathy. The consequence of this apparent display of

interest in the problem by the agent is that it provides a perception of quality

to the caller as well as confirmation of the real problem to be solved.

b) The agents access codified information (57 percent) in the resolution of a

majority of calls, an expected tactic based on interview data, strategies

recommended by Zack (2001), and the investment in structured internal

knowledge repositories. These repositories not only provide the necessary

information, but also the architecture to ensure (routine) successful search

outcomes, and hence ensure consistency of advice.


96

c) Personal details are recorded (47 percent) as a matter of policy in CC-B to

satisfy the requirements of the call-centre management software. In CC-A,

although not formally recorded, notes are regularly made both to prompt the

use of the caller’s name in conversation, and also to fulfil a contract such as

the mail-out of fact sheets if required. This again supports the quality criteria,

but is also used by CC-B to identify the nature of previous calls when a

follow-up call is received and referenced by the caller.

d) Transfer to functional area (33 percent) occurs when the required knowledge

lies outside the expertise of the agent and not apparently available from the

structured internal knowledge repositories. Agents who identify a need for

knowledge on recurring topics are able to request fact sheets, briefings, or

other information to reduce the need to rely on the transfer call strategy.

e) Transfer to user (25 percent) normally indicates a lack of necessary

information able to be provided by the caller. These calls are recorded as

successfully closed since the call centre has lost control of the process, but is

able to provide direction on the next steps towards obtaining a satisfactory

solution (such as how to locate a vehicle identification number prior to

providing advice on the status of the vehicle).

f) Cyclical reframing (8 percent) reflects a trial–and- error approach where the

initial discussions indicate a lack of clarity for the agent (Weick, 1995).

Although not a widely used strategy for handling problems of uncertainty, it

is useful for the alignment of perceptions of both the caller and agent.

g) Other documented strategies have not been employed to the degree that they

are of significance in the analysis of uncertain problems.

Complexity

Problems of complexity are characterised by a diversity of issues, a number of

variables, and their interrelationships. This implies a potential multiple–method

analysis, producing many (differing) solutions. Often characterised by statements

such as, “I just don’t know where to start,” these problem types are surprisingly easy

to identify (based on the 100 percent correlation between researcher and assistant in

the coding process). A typical complex problem call follows the following pattern of

this transcript of an actual conversation:


97

Caller: “I’m trying to identify two types of grasses that grow near my house

but I have no idea who I would sort of ring or how to find out this

information.”

Agent: “OK Now these grasses are a problem or...”

Caller: “No we’re actually trying to identify them because my dog has an

allergic reaction to either one of them and I’m trying to find information…”

This has become more than just a simple grass-identification exercise which would

have resulted in a reference to a taxonomy for grasses. The multiple problems and

the relationship between these led to a resolution based on a range of branched

decisions, such as, “If the initial reaction started during a particular time of the year,

then investigate this, else ...”

The sample of calls contained 17 having complex problem types. Significant

strategies employed by call-centre agents in addressing these problems are specified

below:

a) Meaning through discussion (65 percent) has been used primarily to identify

the major issues involved in the problem. This strategy also links to the

quality characteristic of empathy. The consequence of this apparent display of

interest in the problem by the agent is that it provides a perception of quality

to the caller as well as confirmation of the real problem to be solved.

b) Decomposition (53 percent) is the strategy recommended by Zack (2001).

This strategy requires the larger problem to be broken down into smaller

ones, with the reintegration of the simplified elements directed towards

achieving a satisfactory solution.

c) Personal details are recorded (47 percent) as a matter of policy in CC-B, to

satisfy the requirements of the call-centre management software. In CC-A,

although not formally recorded, notes have been regularly made both to

prompt the use of the caller’s name in conversation, and also to fulfil a

contract, such as mail-out of fact sheets if required. This again supports the

quality criteria, but is also used by CC-B to identify the nature of previous

calls when a follow-up call is received and referenced by the caller.


98

d) Transfer to user (47 percent) normally indicated a lack of necessary

information able to be provided by the caller. These calls have been recorded

as successfully closed, since the call centre has lost control of the process, but

has been able to provide direction on the next steps to obtaining a satisfactory

solution. This proportion initially appears to be high, but considering the

nature of the decomposition process, it is likely that the resolution of some

sub-problem is likely to require further investigation by the caller.

e) The agents access codified information (41 percent), since once simplified (to

the degree that they become familiar and solvable), the sub-problems often

present as being uncertain. The strategies then reflect those which have been

discussed above. These findings indicate a hierarchical relationship between

problem types, as per Figure 19, where the application of resolution strategies

to a complex problem may either provide a resolution directly, or produce a

problem of uncertainty, which in turn requires an appropriate strategy.

Figure 19 - Complex Problem Devolution

f) Cyclical reframing (29 percent) is used more widely in solving problems of

complexity than those of uncertainty, since the method of attack in

decomposition of a complex problem often requires a “trial-and-error”

strategy (at least in the initial stages). As suggested by Zack (2001), this

convergent process leads to identification of sub-problems which are either

familiar or able to be solved by a prescribed method.

g) Transfer to functional area (18 percent) occurs when the required knowledge

lies outside the expertise of the agent, and is not apparently available from the

structured internal knowledge repositories. In terms of transfer-to-user

strategy, it is likely that a sub-problem will require knowledge from an expert

who has experience in the area being addressed.

Complex Problem

Complex Problem Strategy Uncertain

Problem

Uncertain Problem Strategy Problem

Resolved


99

h) Queries requiring further research (12 percent) are not considered closed until

the outcome of the research is relayed back to the caller. These calls reflect a

knowledge gap on the part of the agent, but the tier-2 support available is

unclear. Rather than commencing a referral cycle,31 the agent takes the

problem off-line, having a personal commitment to resolve it and to

communicate the outcome to the caller. Although the response addresses the

quality criteria of empathy and competence, it is not registered as

satisfactorily closed at the time of the call.

Ambiguity

The final problem type encountered in the call centres is one of ambiguity (Zack,

2001). These problems present in calls which tend to skip around a range of issues

without any apparent clear request from the caller. Often many questions, seemingly

unrelated either to each other or to the functions of the organisation, are asked,

without the opportunity to provide a logical solution. The following actual transcript

of a conversation illustrates the ambiguous problem type.

Caller: “My...my son….I’m I’m divorced….my son has been taken .. he sort

of ...he’s unaware it is illegal. He’s only 7 years of age but um when I report

it .. if I ask the Boating and Fisheries... um if this guy is caught …because

my son’s involved in it. Would they notify me?”

The example above relates to an anonymous caller trying to get the partner of his ex-

wife into trouble with the Boating and Fisheries authorities, but without implicating

his son.

Problems of ambiguity are generally associated with a degree of anxiety as well. This

reflects findings by Weick (1995) that such problems cause concerns about the

rationality and orderly nature of their environment. From the sample of calls, 15 have

been classified as being ambiguous, with one of these calls being terminated early

(i.e., undergoing unsatisfactory closure) owing to the inability of the agent to be able

to engage the caller in a rational discussion. In these problem types, the ability to

31 An issue regularly referred to in discussions with staff has been the desire to remove “telephone tag”, where callers may wait in a queue for long periods only to be told they are in the wrong area and be transferred to another holding area awaiting response. Both management and agents have expressed a desire to minimise referrals predominantly for this reason.


100

make sense from what is being sought by the caller is of more importance than the

actual knowledge demands. Once agreement is reached on what the actual problem

is, appropriate knowledge strategies are able to be employed as per the Q-R cycle

(Timbrell et al., 2005b) The resolution methods require cycles of interpretation

without the introduction of new materials (Weick, 1995; Zack, 2001), and are

reflected in this case study. Below is an analysis of the ambiguous problems raised

in the call set.


the major issues involved in the problem. This strategy also links to the

quality characteristic of empathy, which is essential in removing the feelings

of irrationality of the caller. These discussions then provide the opportunity to

clarify the issues in the mind of the caller.

b) Cyclical reframing (47 percent) follows the initial discussions. The aim of

this strategy is to converge what the caller actually required to a single

meaning or understanding, and then to bring about an acceptable solution.

c) Transfer to user (47 percent) indicates a lack of necessary information able to

be provided by the caller. This strategy reflects the expected behaviour of

callers who are, as a result of the clarification process, able to attack the

problem with their own resources supported by a more rational decision-

making context.

d) The agent’s access codified information (27 percent) predominantly from

external sources, rather than structured internal knowledge repositories used

for complex and uncertain problems. This often reflects confusion on the part

of the caller about the responsibilities of the agency. Once the problem has

been clarified, however, agents take the opportunity to access publicly

available information (e.g., via other public-sector Web sites) in order to

deliver a satisfactory closure. The frequency of use of this strategy again

reinforces the hierarchical nature of the problem-solving strategies used

where ambiguous problems may be translated into either complex or

uncertain problems which are then resolved with appropriate strategies. This

is illustrated in Figure 20.


101

Figure 20 - Problem Devolution Hierarchy

e) Social construction (13 percent) is significant in that the resolution of

ambiguity may involve convincing the caller that the agent is pursuing the

right objectives in the right way. This active process utilises the caller’s

senses of perception, attention, and memory to construct “new” knowledge

through collaboration and cooperation.

f) Other documented strategies have not been employed to the degree that they

are of significance in the analysis of ambiguous problems.

Problems of equivocality have not been encountered in the study. This phenomenon

is a consequence of the nature of business in public-sector bureaucracies, where rules

and policy-based activities are thoroughly researched – precisely to remove

ambiguity and equivocality. While such multiple meanings or conflicting

interpretations may exist at the political level, the departmental delivery of services

aligned with public policy is generally clearly articulated in business rules. Hence,

presentation of equivocal problems will likely be met with emphatic, “Our policy

position is that XYZ is the case.” This is a strategy that raises issues of service

quality for the enquirer.

6.4. Service Quality

Table 15 is a composite representation of the frequency of the identified quality

criteria for each problem type derived from the NVivo matrix search of coded calls.

Figure 21 graphically represents these frequencies to identify the variations in their

application by agents.

Complex Problem

Complex Problem Strategy

Uncertain Problem

Uncertain Problem Strategy

Problem Resolved

Ambiguous Problem

Ambiguous Problem Strategy


102

Quality Criteria from Coded Calls

Zack Types Satisfactory Closure

Empathy with

Customer

Needs more Info - Call

Back

Technical Resource Capability

Technically Competent TOTAL

Ambiguity 10 14 0 0 3 15

Complexity 11 16 1 6 7 17

Uncertainty 71 90 12 31 36 111

143

Satisfactory

Closure

Empathy with

Customer

Needs more Info - Call

Back

Technical Resource Capability

Technically Competent TOTAL

Ambiguity 66.7% 93.3% 0.0% 0.0% 20.0% 10.5%

Complexity 64.7% 94.1% 5.9% 35.3% 41.2% 11.9%

Uncertainty 64.0% 81.1% 10.8% 27.9% 32.4% 77.6%

Table 15 - Frequency of Quality Criteria

Strategies to develop empathy with the caller directly reflect the emphasis placed on

training in call-centre procedures. The development of a service culture has been

observed to be very strong in both call centres, with scripted introductions followed

by ad-lib discussions clearly in evidence in the strategies used. The early

establishment of rapport with callers with problems coded as uncertain (81 percent)

satisfies the “empathy quality” criterion. The problem types that require rich

interactive conversations (complexity, 94 percent; and ambiguity, 93 percent) reflect

a high correlation between the response strategy of developing meaning through

discussion, and the development of empathy with the caller.


103

Figure 21 - Graphical Representation of Frequency of Quality Criteria

Technical resource capability32 has contributed to the quality of responses to

problems of uncertainty (28 percent). This is displayed through well-defined

search/navigation processes surrounding the familiar structured internal knowledge

repositories which relate to agencies’ core business. For complex problems, this

capability (35 percent) has been used relatively more often. It has extended to

broader knowledge sets due to the lesser degree of familiarity with the issues raised

by the callers.

Technical competence33 is reflected in the ability of an agent to recall facts and past

experiences relating to the query. This ability is seen to be involved in the resolution

of problems of uncertainty (32 percent), and is aligned with the business of the

agency.

32 Technical competence is the accumulated impact of the machines, materials, facilities, technologies, and databases available for use in resolving the problem. 33 The knowledge and ability to attack the range of problems directed at the agent.

Quality Criteria from Coded Calls

Satisfactory closure

Empathy with customer

Needs more information - Call back Technical resource capability

Technically competent

Ambiguity

Complexity

Uncertainty


104

A further aspect of technical competence is the ability to apply rules and heuristics to

the ambiguous (20 percent) and complex (41 percent) problems in order to realise the

necessary preconditions (clarification and simplification) for resolving the query. All

agents in CC-B demonstrated this competence, while the IVR technology of CC-A

effectively routed the complex and ambiguous calls to specialist sections within the

call centre.

Satisfactory closure (approximately 65 percent) is coded as a quality action only if an

active statement is used by the agent or caller to indicate a satisfactory resolution has

been obtained. Agents often specifically ask if they can be of further assistance, or if

the advice has met their client’s requirements, while callers have often thanked the

agent for their assistance. The noted low frequency of unsatisfactory calls represents

the number of times either the agent or the caller terminated the process prematurely.

This coding rule has also been used for other negative criteria; unless specifically

encountered, however, these criteria are not encoded.

6.5. Summary

Although much time has been spent in building a rich description of the

organisations, their business drivers and management expectations, the actual calls to

the call centres CC-A and CC-B provide the best insight into the ability of agencies

to provide quality advice in an efficient way. The nature of the business of the

public-sector departments studied has determined the types of calls received, while

the different technical and organisational structures within the departments have

influenced management responses to problem solving within their call centres.

Although having different underpinning philosophies and policies in order to provide

satisfactory responses to callers almost 100 percent of the time, both call centres

have systems and procedures backed by a range of knowledge repositories. The

theoretical models and resolution strategies of Zack (2001), Weick (1995), Pentland

(1991), Markus (2001), and Brogowicz et al. (1990), have all been displayed in the

activities of the call centres, indicating at least at an intuitive level that the

management and staff are pursuing practices which align with their business needs.

The technical elements have been coded and reported with predictable or justifiable

frequencies, while the social aspects of the interactions are not so well described by

the models and taxonomies used in the data-collection processes.


105

The relationship between calls and responses is documented in Table 14 - Frequency

of Response Strategies and Table 15 - Frequency of Quality Criteria. The application

of the Zack (2001) problem categorisation to caller problems, and analysis of

response strategies by agents provide a description of work practices of these public-

sector call centres. This view of the organisation is transparent to both callers and

agents, but when both groups are specifically questioned on strategies employed to

deal with the problem types encountered, an almost subconscious response is

triggered. The agents display a natural ability to recognise the method of attack in an

instant, which has a close alignment with the findings of Zack (2001) in terms of

resolution, and with Brogowicz et al. (1990) for quality issues. Although the coding

reflects activity at a quite general level, the results reflect the critical elements

necessary to support the business of the department in taking calls and providing

responses.

A Pragmatic Approach to Call-Centre Strategies

106

Chapter 7


A traditional model for call-centre performance, represented in Figure 22 below,

suggests a simple independent relationship between quality and call-time. This study

has revealed a variety of both documented and enacted performance goals dependent

on the role of the performer. The managers of CC-B have indicated that call-time

has not been an issue, but have relied heavily on the grade of service reporting which

is primarily based on time-in-queue. They considered quality of service to be the

major determinant of individual performance, but monthly reviews were aimed at

moving all agent performance to the fast resolution and high quality quadrant in

Figure 2234.

Figure 22 - Resolution Rate - Quality Matrix

CC-A management placed an emphasis on measurable activities displayed on the

call-centre display boards (time-in-queue, talk-time, abandonment rate), and seemed

much more concerned with throughput and efficiency. However, all observations

place agent activity in the right half of the matrix meaning that they were striving to

deliver a high quality service. The study has found that no evidence of time-

watching existed in the behaviours of either call-centre agents, suggesting that for

their purpose, the concern has simply been to deliver a quality service with an

34 The Resolution Rate - Quality Matrix was used by one of the call centre supervisors to record and report on relative performance of agents. The origin of the tool is unknown.

Quality (as per Quality Monitoring Process

Fast, but lacking quality

Slow and lacking quality

Fast and high quality

Slow, but high quality

Ca

ll R

es

olu

tio

n R

ate


107

expectation that management would address any unit-wide issues around hard

metrics.

The incorporation of time-based components in call-centre performance cannot be

ignored; it is important in the drive for efficiency as the stated business driver for the

formation of CC-A. Timbrell et al. (2005b) posit a model that links service quality to

pay-off assessment on the part of the caller. The pay-off assessment is the critical

decision to determine if the query has been sufficiently addressed (successful

response); has not been adequately answered, but wishes to continue with the

interaction (the Query-Response Cycle continues another loop); or does not appear to

be converging to a solution (unsatisfactory conclusion to the call). Hence, it is the

caller who indirectly determines the number of cycles traversed, and consequently,

the time spent on an issue. This time has a cost, both to the caller and the agency. It

is this cost which determines the inertia of the interaction, and has been quantised in

customer loyalty literature as the perceived investment in time, money, and effort by

the client (Ranaweera & Neely, 2003). Given that the services provided by the call

centres studied are primarily free of charge, the formula for pay-off assessment is

reduced to demands on the time and the effort of the caller. Quality measures used as

a basis for improvement in call-centre performance in the public sector need to

account for such demands on callers.

7.1. Service Value

A positive pay-off assessment does not necessarily imply a satisfied customer.

Service value (as perceived by the caller), a combination of service quality, and

client satisfaction is the initiator of desired client behaviours (Dedeke, 2003).35 One

of the objectives of this study has been to identify improvement strategies without

the need to rely on exit satisfaction-type surveys. Service value better equates to pay-

off assessment, since it incorporates the cost perceptions of callers through their

behaviours (to terminate or continue with a call). The proposition, then, is that an

improvement in service value will deliver an associated improved call-centre

performance.

35 These include displays of trust, assurance, and either acceptance of the advice or permission to

continue.


108

Personal qualities

Service value is derived from a range of characteristics. The personal qualities of

call-centre agents - empathy, reliability, responsiveness, assurance, and tangibles

(such as a clear voice) - directly contribute to caller satisfaction (Parasuraman,

Zeithaml, & Berry, 1988). In their 1999 study, De Ruyter et al. identify drivers of

customer behavioural intent, finding that they differed markedly across six different

service industries, as listed in Table 16 (Dedeke, 2003, p. 284). The strategy implied

is to deliver on these drivers in the same proportion as expected by the caller, and to

improve on this marginally to exceed that expectation. Below is an assessment of

both the service criteria used by De Ruyter et al. (1999) and their impact in the

Queensland Government call-centres being studied.

a) At an individual level, the ability to establish an empathetic relationship with

callers quickly was evidenced across both call centres. To some degree this

has been a consequence of the problem-resolution process, but its pervasive

existence represents a group norm which is either identified as part of the

recruitment process, or established during the pre-service induction. De

Ruyter et al. (1999) have found expectations of consumers of government

services in establishing empathy occurred in only 5 percent of responses. This

is quite a low benchmark: one which is well exceeded by staff in CC-B and

CC-A. This implies no anticipated improvement from strategies to improve

the call-centre staff’s empathy with callers.

b) Reliability of government services (rating 17 percent in the 1999 De Ruyter

study) ranked second to public transport in criticality. This issue has not been

investigated in this call-centre study, since the continued existence of the

service is only threatened by a technical catastrophe (for which business-

continuity plans exist), or a political decision to alter the service.

c) The grade of service measures in place directly addresses responsiveness

characteristics (rating 27 percent in the 1999 De Ruyter study). In general,

Government has the highest rank of all services when measured against this

criterion. Any improvement strategy aimed at this area should deliver

improvement in client satisfaction. Time-in-queue and abandonment rates are

the hard metrics to be minimised, while the ability to address concerns of

callers once connected to an agent provides scope for enhancement. Such


109

strategies are closely aligned with the technical determinants addressed

below.

d) Assurance (rating 15 percent in the 1999 De Ruyter study) is ranked as

second lowest of the six services. The decision to omit this attribute from this

study is supported by the unique business position of each department. The

two agencies have no competitors in the marketplace, and each has evolved to

a point where the organisations are respected, particularly in key

representative stakeholder groups. Supporting this position is the lack of any

negative responses to the quality criteria used in the data-collection process.

No real benefit is likely to be achieved through investment in strategies to

improve assurance.

e) The nature of call centres means that clients interact remotely, and hence do

not experience the ambience of the built environment. Improvement in

tangibles (rating 10 percent in the 1999 De Ruyter study), outside the

generally-accepted standards of telephone communications and published

materials sent to callers, would serve little purpose in increasing perceived

service value for the caller.

Tangibles Reliability Responsiveness Assurance Empathy Total

Shops 12 (13.2%) 13 (11.1%) 40 (20.4%) 38 (16.8%) 28 (17.5%) 131

Restaurants 48 (52.7%) 13 (11.1%) 28 (14.3%) 52 (23%) 8 (5%) 149

Government 9 (9.9%) 20 (17.1%) 53 (27%) 33 (14.6%) 8 (5%) 123

Public transportation 13 (14.3%) 41 (35%) 20 (10.2%) 13 (5.8%) 30 (18.8%) 117

Banking 7 (7.7%) 16 (13.7%) 27 (13.8%) 44 (19.5%) 39 (24.4%) 133

Health care 2 (2.2%) 14 (12%) 28 (14.3%) 46 (20.4%) 47 (29.4%) 137

Total 91 (100%) 117 (100%) 196 (100%) 226 (100%) 160 (100%) 790

Table 16 - Analysis of critical incidents (De Ruyter, Wetzels, & Van Birgelen, 1999, p. 1139)

Technical determinants

Technical determinants provide the largest opportunity for improvement in the

delivery of service value. The ability to resolve issues more quickly will improve

responsiveness (a critical quality contributor to caller satisfaction for government

clients), reduce cost to the caller (time and effort), and reduce demand on call-centre

resources (efficiency principles). Hence, although a practical limit exists on the

ability to effect efficiencies in the problem resolution process, knowledge-based

hoshiko

Rectangle

halla

This table is not available online. Please consult the hardcopy thesis available from the QUT Library


110

strategies provide opportunities to improve call-centre performance when applied to

the following technical determinants.

a) Competence surrounding problem resolution has not been considered as a

specific component of any of the in-service strategies employed by the call-

centre managers, and is instead a by-product of the exposure to the

experience of solving problems. The heuristic knowledge which results from

the creative processes of interpretation is a direct outcome of that process

(Tsoukas & Vladimirou, 2001). This is also reflected in the findings of

Pentland (1991) where the software support group in the call centre were

found to “use rules but do not merely follow rules” (Pentland, 1991, p. 216).

A recursive relationship exists between structure and behaviour, which in turn

determines the rituals, systems, and competence reflected in organisational

knowledge (Timbrell, Delaney, Chan, Yue & Gable, 2005a). The work of

Zack (2001) and the findings of this study suggest the problem-solving

competencies are able to be developed, and that active management

intervention is required to target these competencies.

The formal pre-service strategies employed by CC-B are directed towards

learning the policies, procedures, and standards (which are further expanded

later in this section), the business of the department (facts, contacts, and

responsibilities), and the development of a service culture (again, further

addressed later this chapter). The large percentage of problems of uncertainty

encountered in CC-B indicates the emphasis on knowing facts about the

organisation in pre-service programs is an effort well spent. Given the

boutique nature of this relatively small organisation, such an in-depth

knowledge of the Department is able to be achieved within the extended

period of pre-service and mentored initial period of service. Agents develop

their competence in addressing complex or ambiguous problems as a

consequence of the mentoring program, rather than as a result of a planned

management intervention. The group norm based on engendering empathy

with the caller is primarily responsible for addressing ambiguous problems

through active dialogue (“meaning through discussion”). The system of

referral to experts allows novice agents to refer complex problems based on

their knowledge of the organisational structure; the referral decision may be


111

an uncertain problem type for the agent. With experience, agents develop an

ability to recognise partial solutions (based on factual knowledge) to complex

problems, which in turn encourages the problem decomposition process.

Higher rates of pay and investment in the emotional intelligence of the team

is aimed at maintaining a stable group of agents, which is a necessity given

the strategy employed for problem-solving competency development (Urch

Druskat & Wolff, 2004).

CC-A receives a relatively higher proportion of ambiguous and complex

problem types compared to those of uncertainty. The short pre-service

training process reflects a strategy of becoming familiar with systems and

processes and an in-depth knowledge of the primary information system. This

system contains the information basis for scripted responses, which are based

on a hierarchical structure of functions and related facts, legislation, and

policies. The recruitment process of CC-A does favour applicants with a prior

knowledge of the organisation, but interpersonal experiences are reported as

being more important to selection. The department serviced by CC-A has a

history of regular and significant structural change. The strategies employed

by call-centre management (who also regularly change) mirror the

characteristics of an agile organisation. Very little investment is placed in the

development of the emotional intelligence of the group, which is reflected in

the high turn-over rates of staff. However, the centre manages to operate

effectively. The problems of uncertainty are able to be addressed through the

reasonably static, but highly robust information systems. Complex and

ambiguous problems are able to be referred to agency experts, who contrast

the demographic of the call centre in that they have been working in the area

for a long time (see Figure 13 - Age Demographics of staff employed by the

agency hosting CC-A). They may also result in the agent taking the call off-

line in order to perform further research, which is an option not available in

CC-B.

b) An important factor in the success of knowledge-reuse strategies is the

development and maintenance of knowledge repositories (Markus, 2001).

Generally, consideration is only given to formal, relatively static knowledge

sources such as Web sites, functionally specific databases and fact sheets,


112

libraries, and archival records. Search strategies associated with these sources

are well-defined, and mitigate against concerns of information overload.

Development of a successful knowledge repository strategy requires

significant effort, since it may be based on a creator documenting for their

own needs only, for others in the call centre (having similar needs), or for a

different audience altogether. Motivation of call-centre agents to contribute to

knowledge repositories may also be a limiting factor. Pressures on time, ease

of data capture, confidence in personal abilities, and pay-back for the effort

will all impact on the rate of contribution to shared repositories by authors. In

both call centres studied, agents had effective methods of capturing

knowledge for their own reuse (often showing as paper-based processes). The

knowledge sources produced for dissimilar audiences have primarily been the

responsibility of other organisational units with management processes in

place for their maintenance. Markus (2001) provides direction in the

development of a knowledge repository for capture and reuse by agents for

agents in their roles as:

i. Shared-work producers (maintain context, support searching, document

the rationale);

ii. Shared-work practitioners (publish context along with content, package

expertise, push knowledge to recipients, provide incentives for

contribution and reuse);

iii. Expert-seeking novices (support re-contextualisation of knowledge in

the local context, capture knowledge using understandable terminology,

and training); and

iv. Secondary knowledge miners (metadata stored for secondary use,

structure the knowledge base, develop analytical and synthesis skills,

verify results).

The dilemma faced when considering the development of a knowledge

repository is the degree to which it will contribute to knowledge reuse. CC-B

agents developed individual approaches to documenting responses to new

problems. The ability to stand up and call out “Does anyone know anything

about XYZ?” and expect a positive response suggests that investment in


113

technology to capture and make available such knowledge may deliver little

benefit. However, team leaders have suggested that many issues are cyclical

(season-dependent), and currently rely on the agent’s ability to recall past

experiences in order to respond to calls successfully. As a trade-off, the

implementation of a policy aimed at the capture of relevant responses to

(identifiable) cyclical issues would assist in obtaining consistency of advice

on matters as they recur over time. No new technology would be needed,

since CC-B has call registering software (log the caller, issue, etc.). It does

not, however, record resolution strategies. Current analysis is limited to

frequency reporting on particular issues, which may initiate a management

response, such as seeking further input from departmental experts. The nature

of the evolution of CC-B means that no single (integrated) knowledge

repository (or linking technology) exists. Hence, the agent’s knowledge of the

organisation is an essential pre-condition of successful knowledge repository

use in CC-B. The requirement to open many databases simultaneously has

been a source of nuisance to agents, both because of the cluttered desktop, but

also because of its impact on performance. The specification of an

information architecture capable of identifying subject matter across

databases should be the goal of CC-B management. This would allow the

incremental implementation of technologies to provide a consolidated view of

the codified knowledge held in the agency.

The maintenance of the single knowledge base for functional matters in

CC-A is the responsibility of the Training Officer. Changes to legislation,

feedback from agents, and pro-active planning by the CC-A team initiate

updates to this information source. However, this system is not capable of

capturing the organisational logic underpinning its use. Neither actual data

nor metadata on the advice given to a complainant is captured for future

reference. This is consistent with the current procedure documented in the

knowledge base. Formal (written) complaints are addressed with more rigour,

but the electronic information systems supporting this process do not make

such information easily accessible to call-centre agents. This provides the

potential for the provision of conflicting advice to similar queries, and loses

the opportunity to provide assistance for similar queries in the future.


114

Changes to incorporate such a strategy in systems would be expensive, but

the dynamic nature of CC-A indicates a cost and quality benefit would

accrue.

c) Davenport and Prusak (1998) advise that a process perspective be adopted in

the creation of a knowledge environment. This approach incorporates the

formalisation of routine procedures embedded in normative activities

specified by rules and standards, as well as core team knowledge processes of

knowledge auditing, development, mutual updating, briefing, and reviewing

(Eppler & Sukowski, 2000).

Documented procedures exist in CC-B in the form of flowcharts to assist

agents in working through a generic problem-resolution process. Although

these procedures show the steps involved, they are unable to provide

information on the significance of the intellectual activities encompassed in

each step. The culture of trust and sharing displayed by the agents in CC-B

helps fill this void. Consequently, a consistent standard of service is

displayed, and rituals reinforcing such behaviour are common-place. No

formal process exists for knowledge auditing. However, given the relatively

small size of the unit and the high amount of interaction between staff, a

broad knowledge of agents’ skills, experiences, expectations, and motivations

is shared among team members. Team knowledge development is achieved

through adherence to formal responsibilities to maintain dialogue with

external reference groups. The early-warning system for campaigns and

responses to requests for fact sheets on topical issues ensures CC-B agents’

knowledge and practice are current. Such interaction with external groups is

supplemented by regular and frequent team meetings designed to provide the

opportunity for mutual updating and briefing. The one area able to deliver

benefit to CC-B is the systematic reflection on its performance, critical

incidents, and recurrent activities.

A comprehensive list of training manuals provided by CC-A evinces a

systematic approach to call-centre (and other service-centre) activities.

However, the approach to work in the call centre is more on an individual

basis, with each agent developing strategies for their own use. Work is

assigned to CC-A staff in a way that attempts to maximise individual agent


115

effort utilisation. This reduces any opportunity for development of emotional

intelligence (Urch Druskat & Wolff, 2004). The stress associated with the

constant demands of the telephones is somewhat offset by the opportunity to

take work off-line to be completed at a more leisurely pace. The general

impact of the high work rate is that very little time is given to social activities

able to develop a sense of group efficacy. Knowledge development and

mutual updating are facilitated by regular formal team meetings; however,

knowledge auditing and reviewing do not exist. Eppler and Sukowski (2000)

refer to a range of tools supporting a systems approach that would benefit

team knowledge management in CC-A. These include:

i. Team Knowledge Auditing – A team matrix provides a framework to

determine the present and required knowledge state of an organisation,

while expert web provides transparent access to experts and their

external contacts.

ii. Team Knowledge Development – Coordination of knowledge

development and alignment with business goals is achieved by applying

the pyramid principle which decomposes particular business hypotheses

into manageable knowledge development tasks. The systematic

mapping of knowledge present in team discussions is captured using a

Toulmin map (Toulmin, 1958) which disaggregates arguments into

explicit functional components of claim, grounds, and rebuttal etcetera.

iii. Mutual Updating and Briefing – Visual protocols assist in visually

arranging the main issues addressed in meetings. Team members are

able to be kept up-to-date using flight plans which tabulate situational

tasks against the size/impact of the task and chart the status of all team

tasks in a single place.

iv. Reviewing – Case study and the lessons learned inventory are effective

tools for team review.

d) Communications infrastructure is also identified as a technical determinant

for service quality. Most organisations have technologies such as email and

Intranets that allow users to interact with the changing contextual knowledge

of the organisation. These tools encourage performance improvement


116

through collaboration by capturing, representing, and interpreting their

knowledge resources (Fensel, 2001).

Communication behaviours in CC-B are characterised by the development of

a culture of mutual support. In this environment, informal communications

activities account for a significant proportion of staff interaction.

Conversations over the partitions and in meetings are supplemented by email.

This personalisation strategy reflects the findings of Hansen et al. (1999), in

that the computer is used to “help people communicate knowledge, not store

it”(1999, p. 107). The databases and other published materials are

supplementary to the main organisational modus operandi.

The reliance on the knowledge base within CC-A reflects a strategy of

codification. Resources are applied to ensuring the knowledge base is current

and complete, with an officer assigned to the role of ensuring a consistent and

structured approach to both the insertion of new material and maintenance of

current details. The systems are regarded as the primary source of

information, and are applied where scripted responses are available to the

problems at hand at many levels. The information used to populate these

knowledge bases is also replicated on the Internet,36 but in a different format.

The challenge is thus to keep these sources consistent. The trade-off in the

drive for efficiency is the effort required to build, maintain, and develop

efficient search and navigation methods.

Neither of the knowledge strategies employed is any better than the other.

The goal is that CEOs understand the strengths and weaknesses of both

strategies, thereby choosing the one that best suits their decisions about

knowledge management, and will persist with that approach. Both CC-A and

CC-B reflect the advice of Hansen et al. (1999), in that they predominantly

employ either a codification or personalisation strategy, but also direct some

effort to the subservient strategy. The codification strategy of CC-A relies

36 Customers are encouraged to find the information on the Internet without reference to the call centre. Although the use of the Internet is increasing, the complexity of interpretation of this information means the call-centre agents are needed in many instances.


117

heavily on IVR technology, a tightly controlled information database, and the

relevant transaction processing systems. There is a much smaller reliance on

referrals based on personal expertise. CC-B, on the other hand, applies

significant resources to training new staff in the business of the agency.

Although documents are gathered on specific topical issues, a far greater

reliance is placed on personal knowledge of the situation, or on the ability to

refer to specific expertise.

Functional service offerings

Functional service offerings provide the cohesion between individual personal

qualities and the system-imposed technical determinants. This allows the customer

to synthesise a perspective of the call centre. The importance of empathy and

responsiveness to client satisfaction, and the problem-solving ability’s contribution

to service value are realised in the degree to which the organisational culture is

attuned to a service culture. This occurs through:

− Software tools, techniques, methodologies, routines, and best-practice support

call-centre activities, and the information required to develop the sorts of

indicators and reports that would be as meaningful for agents as they are for

management;

− Contextual awareness to include a consideration of agents’ aptitudes and skills

of willingness to document information, attention to detail, analytical ability;

and

− Willingness of the agent to share personal caller intelligence, and to engage in

dialogue with others about ways to optimize the use of the encultured norms and

practices (Thompson & Walsham, 2004).

The combination of the attributes above contributes to an organisational culture

overtly attuned to service delivery. The following provides an analysis of the two call

centres and their functional service capability.


118

a) Unlike organisation-wide measures37 such as financial performance, gross

transaction counts, and quantum of service, process measures provide detail

on tasks and activities at the work-unit level. Meyer (2004) suggests a

guiding principle underpinning effectiveness of teams is that measurement

systems are to allow them, rather than management, to gauge their

performance. By implication, this requires the team to be intricately involved

in the design and implementation of the system. The focus of these measures

is on the identification of strategies by team members to improve their own

performance.

Performance measurement in CC-B has a hierarchical structure. At the work-

unit level, the grade of service levels are reported on a regular basis.

Performance below the nominated benchmark is investigated, but normally

reflects an unplanned event which either dramatically increases the call-

centre load, or reduces the available pool of agents. The policies and practices

of CC-B, which include sufficient resource redundancy to allow for

predictable fluctuations in demand, ensure that the desired grade of service is

effectively achieved. At an individual level, the call rate and quality of call

are measured against a standard criteria set, and plotted against (anonymous)

others to provide a graphic of relative performance compared to all agents.

The protocol-based processes are assessed, and reports and

corrective/improvement actions are negotiated for the next period. In line

with the recommendations of Meyer (2004), the development and

maintenance of call protocols is carried out in team meetings.

The critical reporting criteria in CC-A are call-resolution rates, time-in-queue

and abandonment rates. Agents are scheduled between on-line and off-line

tasks, having the ability to load-balance by moving staff between these

modes. The relatively high expected call rates of CC-A leads to tension to

meet the benchmark, and has the consequence of increasing stress in the call

centre. Individual performance is assessed primarily in terms of average talk-

time (with the goal of reducing this), but is supplemented by a comprehensive

37 These measures are predominantly of interest to senior management. In a public-sector organisation, staying within budget is important, as is reporting on the number of services delivered (e.g., number of operations performed in a health district).


119

list of performance criteria, which cover skills as well as knowledge of

particular functional issues. Feedback on performance and an associated

action plan are provided to the agent after each appraisal.

b) Agent aptitudes and willingness to document information, their attention to

detail, and their analytical ability contribute to the functional service quality

of an organisation (and hence contribute to perceived service value). Agents

will display various degrees of commitment towards publishing their

knowledge for the use of their peers, owing to time, quality, and social

pressures. Successful knowledge reuse requires the provision of proper

incentives for knowledge producers in the development of quality repositories

(Markus, 2001). Given a policy to publish is actively encouraged, the

robustness of the resulting document will reflect the authors’ moods,

demonstrating that they are certain that the content is true, believe that the

content is true, are indifferent to the truth value of the content as either from a

second-hand and indefinite source, or are uninformed about the truth value

content (Gotel & Finkelstein, 1995). Hence, an ability to identify implicit and

derived group contributions in addition to the explicit ones is an analytic skill

required by all agents.

The only knowledge collected by CC-B operators is basic identifying

information and the call’s topic to facilitate call-back if required. Referrals to

previous calls only serve to identify frequency and cyclical timing issues that

are also highlighted in reporting cycles. Agents within CC-A are not required

to formally collect or publish any call-related information, unless it is the

subject of a particular marketing campaign. An opportunity exists in both call

centres to expand the agent’s capability to capture and publish resolution

strategies and outcomes in order to exploit knowledge transfer and reuse.

7.2. Summary

The common elements of this study are limited to call centres and the Queensland

public-sector environment. The differences in goals (effectiveness versus efficiency)

and organisational components (workload, performance management, technology)

are reflected in predominantly evolutionary knowledge strategies (personalisation

versus codification). In each situation, the business drivers have quite rightly


120

determined the strategies implemented. Personal and technical determinants have

contributed most significantly to each call centre’s performance, with problem

resolution reflecting the methods proposed by Zack (2001) in relation to uncertain,

complex, and ambiguous problem types. Functional service offerings, although

supported by performance management systems, are able to be improved through the

untapped knowledge reuse opportunity to capture and share resolution strategies.

Conclusion

121

Chapter 8

Conclusion

8.1. Introduction

This final chapter is a discussion of the overall learnings from the study. A

conceptual link runs from the presentation of a problem in the first instance, through

a search of the literature to identify related theory and practice, which is followed by

the adherence to a defensible research methodology, to the development of

appropriate research protocols. This case study of two Queensland public-sector call

centres is directed by this protocol (incorporating access to sources of data, sampling

strategies and ethical standards), and by an analytical framework applied to converge

the sources of evidence into patterns and explanations. Certain implications,

limitations, and opportunities for further research are also discussed.

The genesis of the study is based in seeking the improvement of performance of

advice-giving in the Queensland public sector, with call centres being the study’s

focus, since they provide a stable and manageable set of operational variables. A

review of the literature has indicated that:

a) Very little research existed to inform call-centre managers on performance

improvement using principles of knowledge management;

b) Public-sector service delivery has different business drivers from the private

sector;

c) Theories, models, and frameworks existed for organisational problem

solving, role definition, performance measurement, and quality improvement,

but no model existed for their integration into a call-centre environment; and

d) Principles underpinning knowledge management provided fundamental

guidelines in the search for improvement strategies in service delivery.

The structured case study methodology has provided an opportunity to evaluate

evolving theories and knowledge. A cyclical process of collection of evidence and

analysis and reflection allows further development of these themes until a resolution

of the research problem is achieved. Essential to the evidence collection process has

been the Query-Response Cycle, as proposed by Timbrell et al. (2005b), which has

Conclusion

122

provided a model for problem resolution within a call centre. A taxonomy of

problem types and resolution strategies provided by Zack (2001), and quality criteria

defined by Brogowicz et al. (1990) forms a fundamental framework for the study,

with Pentland (1991) providing a coding method to capture the detailed behaviours

in the resolution of individual calls. Finally, Markus (2001) provides the model used

to distinguish roles within a call centre and strategies for performance improvement

based on the knowledge reuse of agents as they perform any or all of these roles.

A broad set of sources has been used in the collection of the data. Guidance from Yin

(1994) has been used to design the protocols and case study activities to ensure tests

for construct, internal and external validity, as well as reliability have been met.

These sources have included direct observation, open-ended interviews, structured

interviews, archival documents, and recorded phone calls. Data from these sources

has been coded according to problem type, resolution strategy, and quality criteria. It

has been analysed through documents generated using relationship identification

capabilities of NVivo qualitative research software.

An analysis of the data in Chapters Six and Seven has indicated that, although each

call centre has differing underpinning philosophies and policies, they both have

systems and procedures to provide satisfactory responses to callers almost 100

percent of the time. The theoretical models and resolution strategies of Zack (2001),

Markus (2001), and Brogowicz et al. (1990), have all been displayed in the activities

of the call centres, indicating, at least at an intuitive level, that the management and

staff are pursuing practices which align with their business needs.

8.2. Research Question




The process used to answer this question has been to:

a) Identify the business drivers of the organisations from a management

perspective;

b) Determine the degree of alignment of the call centre with those drivers,

through observation and interviews with call-centre agents; and

Conclusion

123

c) Collect and organise evidence to describe the processes used by agents in

their daily work.

CC-A is primarily organised to deliver efficiency benefits to the host department

through a low-cost service-delivery model. Technology plays a large part in CC-A’s

operations, with IVR being used to direct calls to agents with experience in those

matters. CC-B is primarily organised to deliver service quality benefits through the

personalised provision of information. Performance targets require calls to be

answered promptly. Since no pre-screening has occurred, any agent may receive a

query on any matter. Uncertain, complex, and ambiguous problem types have been

encountered during the study; their resolution methods are summarised in Table 14.

The sample of calls contains 111 with uncertain problem types. Significant strategies

employed by call-centre agents in their resolution are as follows:

a) Meaning through discussion (57 percent) has been used primarily to clarify

the real needs of the caller (Weick, 1995);

b) The agents accessed codified information (57 percent), which is an expected

tactic based on interview data, drawing on strategies recommended by Zack

(2001); and

c) Transfer to functional area (33 percent) has occurred when the required

knowledge lies outside the expertise of the agent.

The sample of calls contains 17 with complex problem types. Significant strategies

employed by call-centre agents are identified as follows:


the major issues involved in the problem;

b) Decomposition (53 percent), as the strategy recommended by Zack (2001);

c) The agents accessed codified information (41 percent), since once simplified

(to the degree that they become familiar and solvable), the sub-problems

often present as being uncertain; and

d) Cyclical reframing (29 percent) is used more widely in solving problems of

complexity than uncertainty, since, as suggested by Zack (2001), this

convergent process leads to the identification of sub-problems, which are

either familiar or able to be solved through a prescribed method.

Conclusion

124

From the sample of calls, 15 are classified as being ambiguous, and are resolved by

the following strategies:


the major issues involved in the problem;

b) Cyclical reframing (47 percent) has been used to converge to a single

meaning or understanding of what the caller actually required, and then to

achieve an acceptable solution; and

c) Transfer to user (47 percent) reflects the expected behaviour of callers, who,

as a result of the clarification process, are able to attack the problem with

their own resources, supported by a more rational decision-making context.

Although the sample is relatively small, the results confirm both the efficacy of

Zack’s (2001) taxonomy, both in its ability to describe problem types, and in the

problem resolution strategies associated with the uncertain, complex, and ambiguous

problem types. The study has also found that technical competence of the public-

sector call-centre agents aligns closely with this taxonomy, even though they have

had no prior knowledge of it.

Problems of equivocality have not been encountered in the study, so no claims are

able to be made with respect to their place in the framework.

8.3. Implications for Theory

How well the theory of problem resolution in call centres is informed by this study is

limited to the consequential observations of the range of practices being performed.

The existence of heuristics employed by the range of agents without specific training

targeted at that resolution strategy implies a social phenomenon exists to support

learning. No reference has been made to resolution-specific coaching or training by

either call-centre manager. Emphasis has been placed on engaging the caller, use of

the script by CC-A agents, and organisation-specific knowledge by CC-B agents.

The aim of this study has not been, however, to develop new theories. Rather it has

been to validate and test a model which integrates a set of pre-existing theories,

models, and principles in their application to a public-sector call centre. To this

extent, the Query Response Cycle (Timbrell et al., 2005b) provides a robust

description of call-centre activity, and is able to be used to investigate the technical

Conclusion

125

and functional determinants (Brogowicz et al., 1990) of the quality problem

resolution in public-sector call centres.

8.4. Implications for Practice

Zack’s (2001) Four Knowledge Problem Model provides a powerful lens through

which to analyse knowledge-based management strategy within a call centre.

Combined with the Query-Response Cycle proposed by Timbrell et al. (2005b), the

technical capability of two public-sector call centres has been documented and

analysed. Although the strategies associated with problem resolution have not been

overtly managed, they have evolved to reflect the framework proposed by Zack

(2001). Even with the relatively small sample of calls analysed, it is clear that the

problem types encountered have reflected the nature of the business of the agency. In

this regard, one of the call centres studied has a complaints-management role, a

factor which has contributed to higher proportions of complex and ambiguous

problem types and the development of strategies to resolve these in an efficient way.

The other, having a policy of providing information, not advice, has developed

significant skills in the solution of problems of uncertainty.

Neither of the managers of the call centres investigated has seen the need to address

their work-day issues explicitly from a knowledge perspective. However, the systems

and processes in place reflect an intuitive appreciation of knowledge management

principles, and management have been keenly interested in the outcomes of the

study. In each case, the reuse strategies relate only to static (i.e., factual and process)

knowledge. Investment in technology to allow the easy capture and location of

successful problem resolution would improve the performance of the call centres.

Markus (2001) recommends the inclusion of specialist intermediaries in both the

design and codification of knowledge processes, since they are able to maintain the

repositories and facilitate the reuse process, and to address problems associated with

incomplete records and indexing.

The government-wide drive for efficiency is effective from the point of view of each

call centre; however, the organisations’ more valuable knowledge resources (tier-2)

are being used more frequently than needed, thus leading to an overall higher cost of

service. By managing the service more closely from a knowledge self-sufficiency

viewpoint, the overall cost of this service to the organisation could be reduced.

Conclusion

126

Examples of this achievement include extending the time allowed for problem-

resolution in CC-A, improving the knowledge resources available to agents,

providing better training in the technical areas, improving agents’ information

literacy skills (i.e., to improve their search capabilities), and developing better

internal processes for handling problems of complexity and ambiguity. These

strategies would all improve the performance of call-centre CC-A. Such a

rebalancing of workforce and technical strategies could improve not only the

efficiency of the call centres’ function, but also improve their effectiveness, thus

leading to higher service quality.

8.5. Further Research Opportunities

a) The research methodology - a structured case research method (Carroll et al.,

1998) and an extension of the coding work of (Pentland & Rueter, 1994) - is

able to be applied to other similar environments. Extending this research to

include private-sector organisations and other service-centre interactions

(such as counter service) would demonstrate the generalisation of the

frameworks used.

b) By extending the tools developed to capture and code the call data to include

the ability to interrogate calls in real time using voice recognition technology,

broader research questions could be investigated. Examples of these include a

time analysis of strategies, and how they evolve from the perspective of new

agents, changes to call-centre policies, or other organisational developmental

instances.

c) The existence of heuristics employed by a range of agents without specific

training targeted at that resolution strategy implies a social phenomenon

exists to support agent learning. An investigation of this may well inform

both academics and practitioners alike on the activities that enforce transfer

of critical knowledge (socialisation, according to Nonaka and Takeuchi

(1995)).

Conclusion

127

8.6. Summary

The research described in this thesis has used a range of models to investigate the

operations of public-sector call centres. The models have been entirely congruent

with the behaviours and practices of callers and responding agents. The applied

nature of this study means it is easily repeatable in any similar agency wishing to

obtain a better understanding of its activities and its problem-resolution capabilities,

as defined in the taxonomy of Zack (2001). At the same time, the thesis provides

recommendations on ways both to measure and improve call-centre performance

through the implementation of repositories to facilitate knowledge reuse (Markus,

2001).

References

128

References

Alavi, M., & Leidner, D. E. (2001). Review: Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly, 25(1), 107-136.

Anton, J. (2000). The past, present and future of customer access centers. International Journal of Service Industry Management, 11(2), 120-130.

Anton, J., & Gustin, D. (2000). Call Centre Benchmarking: How Good is "Good

Enough". West Lafayette: Ichor Business Books/Purdue University Press.

Arzbacher, S., Holtgrewe, U., & Kerst, C. (2002). Call Centers: Constructing Flexibility. In U. Holtgrewe, C. Kerst & K. Shire (Eds.), Reorganizing

Service Work: Call Centres in Germany and Britain. Burlington: Ashgate.

Baker, K., & Badamshina, G. (2002). Knowledge Management. In E. L. Malone (Ed.), Managing Science as a Public Good: Overseeing Publicly Funded

Science: United States Department of Energy.

Becker, M. C. (2001). Managing dispersed knowledge: Organizational problems, managerial strategies, and their effectiveness. Journal of Management

Studies, 38(7), 1037-1051.

Blackler, F. (1995). Knowledge, knowledge work and organizations: An overview and interpretation., 16(6), 1020-1046.

Blue is the colour. (1998, 1998/06/06/). Economist, p. 65.

Bradbury, A. (2000). Develop Your NLP Skills (2nd ed.). London: Kogan Page Limited.

Brogowicz, A., Delene, L., & Lyth, D. (1990). A Synthesised Service Quality Model with Managerial Implications. International Journal Of Service Industry

Management, 1(1), p 27-45.

Brooke, C. (2002). Information Systems at callcentre. International Journal of

Information Management, 22(5), 389-401.

Brown, J., & Duguid, P.(2001) Knowledge and organization: A social-practice perspective. Organization Science, 12(2); 198-214

Bryar, J. (2001). Taxonomies: The Value of Organised Business Knowledge. Retrieved 29/07/2002, 2002, from http://www.newsedge.com/materials/whitepapers/taxonomies.pdf

Caldwell, F. (2004). KM Adoption Model Highlights Critical Success Factors

Research (Research Article): Gartner Research.

Carlsson, S. (2001). Knowledge Management in Network Contexts. European

Conference on Information Systems ECIS 2001, Bled, 616-627.

Carroll, J., Dawson, L., & Swatman, P. (1998). Using Case Studies to Build Theory:

Structure and Rigour. Proceedings of the 9th Australasian Conference on Information Systems, ACIS'98, Sydney.

Choo, C. W. (1996). The knowing organization: How organizations use information to construct meaning, create knowledge and make decisions. International

Journal of Information Management, 16(5), 329-340.

References

129

Close, W. (2003). Contact Centre Quality Management User Survey Results: Gartner.

Cross, R., Liedtka, J., & Weiss, L. (2005). A Practical Guide to Social Networks. Harvard Business Review. 83(3), 124-132.

Davenport, T., Eccles, R., & Prusak, L. (1992). Information politics. Sloan

Management Review, 34(1), 53-65.

Davenport, T., & Prusak, L. (1998). Working Knowledge: How Organisations

Manage What They Know. Boston: Harvard Business School Press.

Davis, G., & Weller, P. (2001). Are You Being Served?: State, Citizens and

Governance. Crows Nest, N.S.W. Allen & Unwin,.

De Ruyter, K., Wetzels, M., & Van Birgelen, M. (1999). How do customers react to critical service encounters?: A cross-sectional perspective. Total Quality

Management 10, 1131-1145.

Dedeke, A. (2003). Service quality: a fulfilment-oriented and interactions-centred approach. Managing Service Quality, 13(4), 276-289.

Di Gregorio, S. (2000). Using NVivo for your literature review. Proceedings of the Strategies in Qualitative Research: Issues and results from analysis using QSR NVivo and NUD*IST, September, London.

Dixon, N. (2000). Common Knowledge: How companies Thrive by Sharing What

They Know. Boston: Harvard Business School Press.

Duder, J. C., & Rosenwein, M. B. (2001). Towards "zero abandonments" in call center performance. European Journal of Operational Research, 135(1), 50-56.

Eppler, M., & Sukowski, O. (2000). Managing Team Knowledge. European

Management Journal, 18(3), 334-341.

Estelami, H., & Maeyer, P. D. (2002). Customer reactions to service provider overgenerosity. Journal of Service Research, 4(3), 205 - 216.

Feinberg, R. A., Kim, I.-S., Hokama, L., de Ruyter, K., & Keen, C. (2000). Operational determinants of caller satisfaction in the call center. International

Journal of Service Industry Management, 11(2), 131-141.

Fensel, D. (2001). Ontologies: Silver Bullet for Knowledge Management and

Electronic Commerce. Berlin: Springer-Verlag.

Gaballa, A., & Pearce, W. (1979). Telephone sales manpower planning at Qantas. Interfaces, 9(3), 1-9.

Gable, G. (1994). Integrating case study and survey research methods: an example in information systems. European Journal of Information Systems, 3(2), pp. 112-126.

Goh, S. (2002). Managing effective knowledge transfer: an integrative framework and some practice implications. Journal of Knowledge Management, 6(1), 23 - 30.

Gotel, O., & Finkelstein, A. (1995). Contribution structures. Proceedings of the Second IEEE International Symposium on Requirements Engineering,, York, England.

References

130

Graumann, M., Arnold, H., & Beltjes, N. (2003). Call centers: A case study on the interplay between organization and information technology. The Geneva

Papers on Risk and Insurance, 28(1), 111-126.

Gronroos, C. (1984). A service quality model and its marketing implications. European Journal of Marketing, 18(4), 36-44.

Gualtieri, J. (2004). When the whistle blows. CA Magazine, 137, 43-45.

Handzic, M., & Chaimungkalanont, M. (2004). Enhancing organisational creativity through socialisation. Electronic Journal of Knowledge Management, 2(1), 57-64.

Hansen, M. T., Nohria, N., & Tierney, T. (1999). What's your strategy for managing knowledge? Harvard Business Review, 77(2), 106-116.

Harris, C. M., Hoffman, K. L., & Saunders, P. B. (1987). Modeling the IRS telephone taxpayer information system. Operations Research, 35(4), 504-523.

Harvey, N., Harries, C., & Fischer, I. (2000). Using advice and assessing its quality. Organisational Behaviour and Human Decision Processes, 81(2), 252-273.

Humphrey, W. (1989). Managing the Software Process. Reading, Massachusetts: Addison-Wesley Publishing Company.

Interview 1a - Annotated notes, January 29, 2003.

Interview 1b - Annotated notes, December 23, 2002.

Interview 1c - Full transcript, June 9, 2004.

Interview 1d - Full transcript, June 11, 2004.

Interview 2a - Annotated notes, May 27, 2003.

Interview 2b - Full transcript, June 16, 2004.

Jonassen, D. H. (2000). Toward a design theory of problem solving. Educational

Technology, Research and Development, 48(4), 63.

Jungermann, H. (1999). Advice Giving and Taking. Proceedings of the Proceedings of the 32nd Hawaii International Conference on System Sciences - 1999, Hawaii.

Kaplan, B., & Duchon, D. (1988). Combining Qualitative and Quantitative Methods in Information Systems Research: A Case Study., MIS Quarterly,12, 570-586.

Kjellerup, N. (2004). Call Centre Know How Essays.. Productivity, Measurements &

Benchmarks. Retrieved 18/06/2004, 2004, from http://www.callcentres.com.au/resource.htm

Kreizman, G., Kotsikopoulos, D., & Shah, J. (2005). How to Build a Business Case

for Service-Oriented Development of Applications in Government: Gartner Research.

Kucza, T. (2001, 9/01/2002). Knowledge Management Process Model. Retrieved 6/08/2002, 2002, from http://www.inf.vtt.fi/pdf/publications/2001/P455.pdf

Leonard, D. (1995). Wellsprings of knowledge: building and sustaining the sources

of innovation. Boston: Harvard Business School Press.

References

131

Ling, T. (2002) Delivering joined–up government in the UK: dimensions, issues and problems, Public Administration, 80(4), 615-642

Mahon, J. (2003). Social issues in management conference symposia abstracts. Academy of Management Proceedings, pp. 1-14.

Majchrzak, A., Neece, O. E., & Cooper, L. P. (2001). Knowledge reuse for

innovation-The missing focus in knowledge management: Results of a case

analysis at the Jet Propulsion Laboratory. Proceedings of the Academy of Management 2001, Washington, D.C.

Managing for Outcomes - Strategic Governance Principles & Indicators. (2003). Retrieved 2/02/2005, from http://www.treasury.qld.gov.au/office/knowledge/docs/manage-for-outcomes/mfo-governance.pdf

Markus, M. (2001). Toward a theory of reuse: Types of knowledge reuse situations and factors in reuse success. Journal of Management Information Systems,

18(1), 57-93.

Meyer, C. (2004). How the Right Measures Help Teams Excel. Harvard Business

Review on Teams That Succeed, 27-51. Boston: Harvard Business School Press.

Milner, E. M. (2000). Managing Information and Knowledge in the Public Sector: New York: Routledge.

Nonaka, I. (1994) A dynamic theory of organisational knowledge creation. Organisation Science, 5(1).

Nonaka, I., & Takeuchi, H. (1995). The Knowledge Creating Company: How

Japanese Companies Create The Dynamics Of Innovation. New York: Oxford University Press.

Parasuraman, A., Zeithaml, V., & Berry, L. (1988). SERVQUAL: a multiple-item scale for measuring customer perceptions of service quality. Journal of

Retailing, 64(Spring), 12-40.

Pentland, B. T. (1991). Making the Right Moves: Toward a Grammar of Software

Support Hot Lines. Massachusetts Institute of Technology, Massachusett.

Pentland, B. T., & Rueter, H. H. (1994). Organizational routines as grammars of action. Administrative Science Quarterly, 39(3), 484-510.

Performance Management Framework. (2002). Retrieved 2/02/2005, from http://www.treasury.qld.gov.au/office/knowledge/docs/manage-for-outcomes/mfo-performance.pdf

Peters, T. J., & Austin, N. (1985). A Passion For Excellence : The Leadership

Difference (Fontana paperback ed.). London: Fontana/Collins.

Polanyi, M. (1967) The tacit dimension. Garden City, New York: Anchor.

Ranaweera, C., & Neely, A. (2003). Some moderating effects on the service quality-customer retention link. International Journal of Operations & Production

Management, 23(2), 230-248.

Riggs, B. & Thyfault, M. (1999) The Modern Call Center. Retrieved 6 February 2005, from http://www.informationweek.com/755/call.htm

References

132

Rogers, A., Hassell, K., Noyce, P., & Harris, J. (1998). Advice-giving in community pharmacy: variations between pharmacies in different locations. Health and

Place, 4(4), 365-373.

Rus, I., Lindvall, M., & Sinha, S. (2001). A State of the Art Report: Knowledge

Management in Software Engineering. Retrieved 6 February 2005, from http://www.dacs.dtic.mil/techs/kmse/kmse.pdf

Schefe, N., & Timbrell, G. (2004). A Tale of Two City Call Centres. Proceedings of the 15th Australasian Conference on Information Systems, Hobart TAS.

Senge, P. M. (1997). Communities of leaders and learners. Harvard Business

Review, 75(5), 30-32.

Service Level Agreement for Provision of Services. (2004). Brisbane.

Standards Australia (2003) Interim Standard for Knowledge Management (AS 5037(Int)-2003), Retrieved 10 October 2004, from Standards Australia Online database.

Teicher, J., Hughes, O., & Dow, N. (2002). E-government: a new route to public sector quality. Managing Service Quality, 12(6), 384-393.

Thompson, M., & Walsham, G. (2004). Placing knowledge management in context. J Management Studies, 41(5), 725-747.

Thompson, P., Warhurst, C., & Callaghan, G. (2001). Ignorant theory and knowledgeable workers: Interrogating the connections between knowledge, skills and services. Journal of Management Studies, 38(7), 923-942.

Timbrell, G., Andrews, N., & Gable, G. (2001). Impediments to Inter-firm transfer of best practice: in an Enterprise Systems Context. Australian Journal of

Information Systems, Special Edition, 116-125.

Timbrell, G., Delaney, P., Chan, T., Yue, A., & Gable, G. (2005a). A Structurationist

Review of Knowledge Management Theories. Proceedings of the International Conference on Information Systems, Dec 11-14, Las Vegas, USA.

Timbrell, G., Koller, S., Schefe, N., & Lindstaedt, S. (2005b). A knowledge process infrastructure hierarchy model for call-centre processes. Journal of Universal

Computer Science, 11(4), pp. 546-564.

Timbrell, G., Nelson, K., & Jewels, T. (2003). Knowledge Reuse in an Aplication Service Provider. In Knowledge Management: Current Issues and

Challenges. Hershey PA: IDEA Publishing.

Timbrell, G., & Shepperd, B. (2002). Improving Advice and Support Services using

Applied Knowledge Management Strategies. Proceedings of the Enabling the Information Future Conference, 16-17 October, 2002, Brisbane.

Todd, R., & Southon, G. (2001). Educating for knowledge management: Perceptions of library and information professionals. Australian Library Journal, 50(4).

Toulmin, S. (1958) The Uses of Argument. Cambridge University Press, Cambridge.

Tsoukas, H., & Vladimirou, E. (2001). What is organizational knowledge? Journal of

Management Studies, 38(7), 973-993.

References

133

Tuten, T. L., & Neidermeyer, P. E. (2004). Performance, satisfaction and turnover in call centers; The effects of stress and optimism. Journal of Business

Research, 57(1), 26-34.

Urch Druskat, V., & Wolff, S. (2004). Building the emotional intelligence of teams. In Harvard Business Review on Teams That Succeed (pp. 27-51). Boston: Harvard Business School Press.

Van de Walle, S., & Bouckaert , G. (2003). Public service performance and trust in government: The problem of causality. International Journal of Public

Administration, 26(8-9), 891-913.

von Krogh, G., Nonaka, I., & Aben, M. (2001) Making the most of your company’s knowledge: A strategic framework. Long Range Planning, 34, 421-439

Weick, K. (1969). The Social Psychology of Organizing. Reading, Mass: Addison-Wesley.

Weick, K. (1995). Sensemaking in Organizations. Thousand Oaks: Sage Publications.

Wiig, K. (1999). Comprehensive Knowledge Management - Working Paper. Retrieved 6 February, 2005, from http://www.knowledgeresearch.com/downloads/compreh_km.pdf

Wilde, C. (2000). Web Call Centers Benefit Customers and Businesses - Linking web

sites and call centers can improve service and customer retention. Retrieved 23 March, 2004, from http://www.informationweek.com/809/calls.htm

Yin, R. K. (1994). Case Study Research, Design and Methods (2nd ed. Vol. 5). Newbury Park: Sage Publications.

Zack, M. H. (1999). Developing a knowledge strategy. California Management

Review, 41(3), 125-145.

Zack, M. H. (2001). If Managing Knowledge is the Solution, then What's the Problem? In Y. Malhotra (Ed.), Knowledge management and business model

innovation (pp. 16-36). Hershey, PA: Idea Group Publishing.

Appendices

134

Appendices

Appendix 1 - General Information Processes in CC-B

GENERAL INFORMATION PROCESS

RESPONSIBILITY PREPARATION NEGOTIATION AND

PERFORMANCE

ACCEPTANCE

CLIENT

CALL CENTRE

BUSINESSAREA

Client calls

Call Centre

Does

operator

have enough

information toanswer the

inquiry

Log call

details

Determine General

Information is

required

Identify the correct

contact number

Explain call details

to Business Area

and transfer call

Identify the

location of the

information

Answer inquiry

using all information

available in CCIS

Cut and paste

call details into

email and send

Business Area

provides general

information

Random call

back to verify

client

satisfaction

Provide feedback

to Call Centre

END

N

Y

Identify the

subject nature of

the call

Is the

Business Area

available

Business Area

documents inquiry

for CCIS

N

Y

Information

provided to

client

Log call

details

Appen

dic

es

135

A`

pp

end

ix 2

- A

ver

age

Ad

vic

e L

ine

Wait

Tim

e – C

C-A

38

Bri

sb

an

e C

usto

mer

Serv

ice C

en

tre

0

20

40

60

80

100

120

140

Jul-04

Aug-0

4S

ep-0

4O

ct-

04

Nov-

04

Dec-0

4Ja

n-0

5F

eb-0

5M

ar-

05

Apr-

05

May-0

5Ju

n-0

5

Seconds

BC

SC

(R

EV

S)

BC

SC

(C

ON

S)

BC

SC

(B

US

)B

CS

C A

VG

MP

S S

TD

BC

SC

(B

US

)B

CS

C A

VG

YTD

38 A

bb

revia

tio

ns

are

no

t ex

pan

ded

sin

ce t

hey

sim

ply

rep

rese

nt

dif

feri

ng f

unct

ional

gro

up

s w

ithin

the

agen

cy.

MP

S S

TD

– t

he

org

anis

atio

nal

tar

get

AV

G Y

TD

- t

he

aver

age

for

yea

r-to

-

dat

e per

form

ance

.

135