1

Collaborative Key Performance

Indicators

R. Gene Richter Business Award Submission

Category: Process Tuesday, March 11, 2008

Willem van Oppen willem.vanoppen@kpn.com Chief Procurement Officer,

KPN Telecom, The Netherlands

Prepared by Joe Thornton

joe.thornton@kpn.com Operational Buyer

KPN Telecom The Netherlands

2

Executive Summary on Collaborative KPI’s KPN provides telephone, Internet and television services to personal customers through its fixed network in

The Netherlands. KPN also provides a range of voice, Internet and data services and fully-managed outsourced ICT solutions for business customers. KPN also offers mobile services in The Netherlands, Germany, Belgium and Western Europe. As of March 31, 2007, KPN served 6M voice subscribers, 8.7M mobile customers, 2.3M Internet customers, and 0.3M TV-customers in the Netherlands, and served 15.6 million mobile customers in Germany and Belgium. The company was founded and incorporated in 1989, and is based in The Hague, the Netherlands, and its shares are listed on the stock exchanges of Amsterdam, New York, London and Frankfurt.

In the retail telecom service setting, there are multiple independent companies that need to collaborate

closely via a host of interrelated IT systems and business processes to serve the needs of the end customers. In these settings, the “conventional” buyer-supplier performance metrics and contracts are insufficient and can even be counter-productive. There is a key assumption for conventional contracts that does not hold: there is no clear distinction between the buyer who delegates work to the supplier, who does the work. Buyers and suppliers co-produce the service to the end customer, and the performance of the suppliers is strongly dependent on the actions of the buyer. Also, the complexity of the dozens of IT systems and business processes is such that one cannot determine the cause of the performance glitch to the final customer. (see figure 5)

The case we’ll examine stems from our relationship with Atos Origin, our outsourced IT provider.

Originally, Atos secured two safeguards regarding revenues: • Revenue guarantee - if KPN demand was less than a certain threshold, KPN would pay Atos

penalties ranging between 25 - 50 % of the gap between guaranteed and actual revenues. • First-Call-Last-Bid clause - ensured KPN confidence for market conformity, and ensured that Atos

would have a chance to match any competitive offer from the market. With a changing market, the relationship between KPN and Atos was deteriorating. The original

outsourcing deal was counterproductive against improved performance and partnership. KPN's cost-down drive resulted in low IT expenditures, which invoked penalties under the Revenue Guarantees. To counter these, KPN would enter into ill-founded projects with Atos, not for competence-based reasons, but financial ones (to avoid penalty costs). Over time, KPN devised a solution.

The initiative that KPN developed to address this was first introduced August 2003 and began to have significant and measurable outcomes starting in autumn 2004. This continues to produce positive results for KPN today, and has become a company “best practice”.

This “best practice” is called “Collaborative KPIs”. First a significant part of the monetary value of the

contract is set apart in escrow. Secondly, a set of performance measures is developed for the suppliers, which links their performance directly to the service provided for the end-customers, even if this performance is influenced by others. Thirdly, a two-way set of KPIs are developed, containing performance targets for the supplier as well as for the buyer. How much of the escrow budget is paid to the suppliers is determined by how well both suppliers and buyer meet their respective targets.

This initiative made joint process improvement opportunities more visible and created a shared mindset to

address these. This led to higher quality execution, reducing the “rework” and labor costs, and improving customer satisfaction with service delivery. This manifested higher revenues from end-customers for KPN, and higher revenues for the supplier who is rewarded for improvements in end-customer service performance. This win-win approach has further improved the buyer-supplier relationship.

This process has also changed the managerial orientation on both sides fundamentally, reflecting a

successful partnership model: long term, quality, and end customer driven. The main transition made here was from a relationship of buying services into a relationship based on the assumption of contracting performance.

KPN is spearheading this cutting edge innovation in buyer-supplier relationships. Through

collaborative KPI’s, both buyer and suppliers see their performance evaluated, and see this evaluation taking place on the basis of what affects service to end customers, not on what happens under their direct control. In this situation the relationship between buyer and supplier becomes a true partnership, where both sides fail if the end customer is not served well, and both parties share profits if end customer is delighted. KPN is leading the way with this innovative and unique practice, and is regarded as a state of the art initiative to manage critical buyer and supplier relationships.

3

With Atos, the area of Operational & Enabling KPIs was most affected by this initiative. This KPI saw drastic improvements of the complaint level (Figure 2). In 2005 it took only 6 months for the complaint level to drop from 14 % to 8 %. This was key for customer satisfaction, however one should not underestimate the cost impact of such improved operational performance. After all, hundreds of people are involved with fixing problems in a company like KPN. If the error level halves, so shall the number of FTE required to deal with quality issues.

With KPN, the Operational and Enabling KPIs were also affected. For the KPN processes to run better it

was essential not to limit the challenge to Atos only, but to extend it to the KPN colleagues in the process. Thus, this KPI was in every aspect, linked to those of Atos, as can be seen in Figure 3.

In the autumn of 2004, it became clear that significant improvements had been made in the actual

operational performance as well as the quality of the buyer-supplier relationship, with clear improvements in terms of trust and transparency. In February 2005, the first payment of the “drawing rights” took place. Over the year 2004, Atos was awarded 79 % of its potential under the Atos KPIs. KPN had to remit 7.5 % to Atos of its potential 2004 exposure. This was final evidence that both organizations had successfully changed the rules of the game, the operational performance and, perhaps most importantly, the nature of their relationship.

As for sustaining this initiative, the market outlook is constantly changing, as are the expectations.

Traditional contracts will not be adjusted automatically and therefore become more and more dysfunctional in driving performance. Typically, realigning these requires substantial effort from mid-level management, with final endorsement by top management from both sides. KPN is currently striving to achieve this endorsement and implement Collaborative KPI’s with all of our top suppliers to improve service as well as revenue.

4

Supporting documentation

Figure 1: A causal loop diagram of propagation of errors in the network

BREAKDOWNS ofinterfaces between

systems in day process

Disturbances resultingfrom humanintervention

Time-outs ofsystems

Fullness of filesystems

Interruption interceptionof service order (as a

circumvention)

Ability to key incomplete order

Up-to-dateness ofdatabases

++

+

+

-

-

Intensity ofmonitoring of

interfaces

-

Attention given todealing with error

lists

-

Ability to sortout queries

Correctness ofconfirmation letter

Confusion withcustomer

Customer calls

Customer queries

Customer orderchanges

Agent queries

Agent timeinvestment

Work pressure

-

-

+

-

-

+

+

+

+

+

+

+

CUSTOMERSATISFACTION

-

No of critical customerprocess applications

(40)

- Appear on worklist for backoffice

-

Backoffice activity

Correction of errors

Quality of delivery

-

Revenues

Budgets

+

+

+ Upgrade skill levels

+

People'scompetence

Coding complexity+

Backup systemsavailable

Prevention quality

Stability ofinterfaces between

systems

No interfacesbetween systems

Focus onsystem-specific

performance

Owners ofinterfaces

-

SLAs on interfaces

+

-

Interrelatedness ofsystem performance

-

Control ofsystem network

Predictability ofconsequences of local

changes to other systems

-

+

-

-

Figure 1: This diagram reads as follows. At the top-left one can see a central element in “what can go wrong and corresponding results”: some interface between IT systems during the day breaks down. Consequently, the order entry systems are not, or only partially available. This makes that new customer orders cannot be (fully) entered, that systems cannot be updated and that pending queries cannot be resolved. There are many potential reasons for such a breakdown of an interface. Some of them are technical, others are human. The top-right part of the diagram explains why this happens so often. Most importantly, when a change is made to one of the 40+ critical systems in the delivery process, the effect on any of the other systems is unclear, because of the large number of systems involved, and lack of control over their interfaces. When errors are made people will try to correct them. This is described in the lower-middle part of the diagram. Doing so is a complex undertaking, requiring considerable skill. Much of that skill had either left the companies or was becoming obsolete as a result of the ongoing changes in all these systems. Budgets were too low to keep technical expertise up to standard. As a result, actual service deliveries were not “right the first time” far too often. The bottom-left of the diagram illustrates the vicious cycle the delivery process went into when such errors occurred. Mistakes in service deliveries could be caused by the wrong kind of conformation letter to the customer because of system errors, or because of data pollution in the systems as a result of inadequate error corrections. In all cases, mistakes lead to customer confusion, customer queries and therefore greater time pressure for the agents to sort out these queries, leaving even less time for dealing with the list of known errors to be resolved.

5

Such a delivery process evidently leads to low customer satisfaction, lower future revenues and further pressure on budgets necessary to improve performance; thus creating yet another vicious cycle of low performance leading to low revenues leading to low investments leading to even lower performance. This is shown in the bottom of the diagram. Figure 2: Development of operational performance as a result of joint improvements

Nazorg 2005 versus nazorg 2006

0%

2%

4%

6%

8%

10%

12%

14%

16%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

week

perc

enta

ge

nazorg 2005 nazorg 2006 Figure 2: This chart compares the average percentage of complaints per week of 2005 and 2006. This shows the drastic drop in complaints, and gives visibility on how overall customer service has been improved from 2005 to 2006.

Table 3: Two-way KPI with their weights over time KPI’s 2004 2005 2006 2007 Total Innovation & Redesign 3% 2% 2% 0% 6% Operational and Enabling KPIs 6% 6% 5% 3% 20% Total Cost of Ownership 8% 6% 0% 0% 14% KPN BU Client Satisfaction 2% 2% 2% 2% 6% Sell to and with Atos 3% 3% 2% 2% 9% Total of Atos-directed KPIs 21% 18% 9% 6% 55% Atos wallet share at KPN BUs 6% 12% 6% 3% 27% IT Governance 3% 6% 2% 0% 11% Operational & Enabling KPIs 3% 2% 2% 2% 8% Total of KPN-directed KPIs 12% 20% 9% 5% 45%

Table 1: This table displays two-way performance targets for KPN and Atos Origin from 2004 to 2007. Five KPIs (with a host of supporting KPIs via the so-called KPI Tree) applied to Atos’s performance, and three KPIs applied to KPN’s performance. Meeting the Atos Origin KPIs would lead to “drawing rights” for Atos and, in fact, payment in cash by KPN.

6

Partial satisfaction of KPIs would lead to a lower amount being paid out. Payment would take place on a year-by-year basis. The structure of the KPN KPIs was such that lest KPN would not be able to achieve these performance targets, it would have to remit cash to Atos. A joint KPI Office was set up that managed and monitored the main KPIs and their supportive KPI Trees. Figure 4: Figure 4 summarizes the main inputs, outputs and results of the Collaborative KPI approach, as well as their key interrelations.

Top ManagementStrategy Alignment

Meetings

Mid-level managementContract redesign

sessions

Operational Process & RootCause Mapping Workshops

Modelling financial impact ofoperational redesign

From:

� 1000s detailed &local SLAÕs

�Low Trust

�Managerial orientation

�Short-term

�Cost/sales-driven

�Localperspective

�0-Sum gameperspective

To:

� 7 integral/collaborative KPIÕs(for both supplier andcustomer)

�Medium to high trust

�Longer-term

�Quality/consumersatisfaction-driven

�Integral perspective

�Partnership perspective Workflows in sync

Visibility of processimprovementopportunities

Fewer errors/lessrework

Higher consumersatisfaction/lower costs

Higher customerrevenues

Higher supplierrevenues

Improved customer-supplier

relationship

�Knowledge of operational processes andinterdependencies

�Deep

Inputs Outputs Results

�Limited Figure 4: Figure 4 shows on the right the indirect results of the Collaborative KPI approach. Firstly, it made joint process improvement opportunities more visible and created a shared mindset to address these. This led to higher quality in execution. Consequently, this had the combined effect of reducing rework and thereby labor costs as well as improving customer satisfaction with service delivery. That then led to higher revenues from end-customers for the buyer side of the supply network. And that leads to higher revenues also for the supplier who is eventually rewarded for improvements in end customer service performance.

Figure 4 also lists the immediate outputs of the Collaborative KPI process. It points out that this entailed considerably more than a major reduction of the number of performance indicators or Service Line Agreements; however significant this change may have been. More importantly, we argue that this process has changed the managerial orientation on both sides fundamentally, much more in line with a successful partnership model: long term, quality and end customer driven, from an integral supply network perspective, and with deep knowledge of operational processes and interdependencies.

7

Figure 5: Visualization of IT Reality

Figure 5: Historically grown application landscape (>1000 applications) with stove pipe solutions and > 1700 point-to-point connections. This results in inconsistent data, high costs of operation, and poor customer satisfaction.

8

Figure 6: Results from managing the Delivery Process as a whole. (figures in millions of Euros)

13,4%12,9%

11,8%11,4%

10,9% 11,1%

9,9%9,3%

8,6%

10,3%

8,4%8,9%

7,2%

10,3%

8,7%

9,7%9,3%

11,0%

0%

2%

4%

6%

8%

10%

12%

14%

16%

2004.06 2004.07 2004.08 2004.09 2004.10 2004.11 2004.12 2004.13 2004.14 2004.15 2004.16 2004.17 2004.18 2004.19 2004.20 2004.21 2004.22 2004.23

Rewards Paid Out 2004 2005 2006 2007 TotalOperational & Enabling KPI's 2 (1,9) 2 (1,7) 1,5 (1,4) 1 (0,8) 6,5 (5,8)