sustaining growth in the modern enterprise

Sustaining growth in the modern enterprise:

A case study

Luis Rabelo a,*, Thomas Hughes Speller Jr.b,1

a Industrial Engineering and Management Systems, College of Engineering and Computer Science,

University of Central Florida, Eng. Bldg. No. 2, Office 312, 4000 Central Florida Blvd,

Orlando, FL 32816, USAbEngineering Systems Division, School of Engineering, Massachusetts Institute of Technology,

Building E40-248, 77 Massachusetts Avenue, Cambridge, MA 02139-4307, USA

Available online 7 November 2005

Abstract

This paper describes a model developed to demonstrate the constancy of and virtual need for change in a

sustaining organization. Several executives from different organizations were interviewed in order to

capture the dynamics structure. Essential variables were listed and causal modeling was used to obtain

important relationships. Equations of interrelationships among the variables were developed. Validation of

the model was performed at two levels: (1) validation of the structure as suggested by the causal modeling

process using a new methodology and (2) validation of the emergent behavior by using case studies of

industries which have different time rates of evolution.

# 2005 Elsevier B.V. All rights reserved.

JEL classification: O32; O33

Keywords: System dynamics; Growth; Strategic business units; New product development; Causal modeling

1. Introduction

Growth is an important concern for the modern enterprise, and industry focus has shifted more

towards solutions, not only products. Innovation, product development, and global awareness are

essential in this new focus. There are several reasons for this strategy:

(a) Macro economic trends are apparent:

(1) globalization of the world economy,

www.elsevier.com/locate/jengtecman

J. Eng. Technol. Manage. 22 (2005) 274–290

* Corresponding author. Tel.: +1 407 882 0091; fax: +1 407 823 3413.

E-mail addresses: [email protected] (L. Rabelo), [email protected] (T. Hughes Speller Jr.).

URL: http://web.mit.edu/thsjr711/www/1 Tel.: +1 617 253 9756.

0923-4748/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.jengtecman.2005.09.002

(2) the move to dynamic supply chains, modularization of business processes, and the trend

towards ever changing networks of buyer/supplier relationships.

(b) There are several types of differentiated selling models employed globally.

(c) Some of the biggest customers (i.e., very profitable corporate accounts) have become more

global than they were before. They operate in many countries and have migrated from a

basically centralized to a decentralized organization structure using common processes

throughout the world.

(d) Another important role of this hybrid product/service scheme is to better understand potential

opportunities and to identify which markets are growing, who are the competitors, what are

their activities in specific market segments, how will the competitors react to new

competition, and what other challenges they may pose.

Therefore, to study this new environment we decided to examine the factors involved in:

sustaining corporate growth while providing economic value-added, diversifying the corporation

by increasing the market segments served, increasing new product development rates, and

increasing order win rates. System dynamics was selected as the modeling methodology due to

the dynamic nature of the problem to be modeled and the impact of causal interrelationships. A

system dynamics model was developed first in loop form and then simplified in more of a stock

and flow diagram to depict and define the relationships between two companies within an

enterprise. Extensive analysis of the model was conducted to discern the behavior among the two

companies and the marketplace in addition to the enterprise as a whole. The lessons learned can

be extrapolated and applied to a larger enterprise or serve as a growth model for startup, small and

medium sized companies. The two companies described herein are separate yet synergistically

related strategic business units (SBU).

2. System dynamics

System dynamics is a method for studying the systems (and systems of systems) around us.

Senge (1994) states that system dynamics is a conceptual approach to facilitate the understanding

of complex problems. The analysis of sustaining growth in the modern enterprise is a complex

problem indeed. The central concept to system dynamics is the understanding of how all the

objects in a system interact (i.e., causal relationships) with one another. ‘‘A system can be

anything from a steam engine, to a bank account, to a basketball team. The objects and people in a

system interact through ‘feedback’ loops, where a change in one variable affects other variables

over time, which in turn affects the original variable, and so on.’’ Causal relationships exist in

which the behavior of one element influences that of another. System dynamics asserts that these

relationships form a complex underlying structure for any system. This structure may be

empirically or theoretically discovered and described. Through the discovery of the system’s

underlying structure, the causal relationships become clear, and predictions may be made of the

future behavior of the different agents in the system.

The creation of a formal dynamic model of a system requires the identification of the causal

relationships that form the system’s feedback loops (Forrester, 1971; Sterman, 2000). Generally,

feedback loops are thought to be either negative or positive. A negative feedback loop is a series

of causal relationships that tend to move behavior towards a goal. In contrast, a positive feedback

loop is self-reinforcing. It amplifies disturbances in the system to create high variations in

behavior. Causal loop diagrams are important tools for representing the feedback structure of the

systems. A causal loop diagram consists of variables connected by arrows denoting the causal

L. Rabelo, T. Hughes Speller Jr. / J. Eng. Technol. Manage. 22 (2005) 274–290 275

influence among the variables. The important feedback loops are also identified and displayed in

the diagram. From our own experience, building these causal diagrams represents a very efficient

and effective knowledge acquisition methodology (Fig. 1).

Now that we have elicited and captured the mental models by using causal loops, we can start

developing a stock and flow structure of the system (Fig. 2). Stocks are accumulations of

information. They characterize the state of the system and generate the information upon which

decisions and actions are based. Stocks create delays by accumulating the difference between the

inflow and outflow of a process. Flows (outflows, inflows) are rates which are added to or

subtracted from a stock. This graphical description of the system based on stocks and flows can

be mapped into a mathematical description. Once the model has been graphically and

mathematically described, it may be simulated by the computer.

3. Causal loops

The theme of the model is to study the dynamics of creating corporate growth with a positive

economic value-added (EVA) in perpetuity. So far, no company has overcome the forces limiting

corporate growth and making it vulnerable to ultimate merger, acquisition or failure. The model

incorporates product development as the means to fuel a diversification strategy with an

increasing portfolio of products to create sustainability. The key feedback loops in this study are

the systems producing organization, the services and spares parts business and of course the

marketplace. Interwoven in the systems producing organization is the product line order

fulfillment and the product development scheme to continuously replenish the company’s

portfolio of products and grow the enterprise.

The project herein started by investigating and modeling two loops of an organization that has

two strategic business units (SBU): SBU1 and SBU2. SBU1 and SBU2 are synergistically joined

at the center by the necessity for successful order fulfillment. The background for the SBU model

structure involved formal and informal interviews with executives and personal from different

organizations in companies such as Honeywell International, Gemcor Systems Corporation, and

L. Rabelo, T. Hughes Speller Jr. / J. Eng. Technol. Manage. 22 (2005) 274–290276

Fig. 1. Causal loop diagrams.

Delphi Packard. Causal models were developed in order to capture the variables and

interrelationships of interest. An iterative process was necessary. Additionally the word of mouth

(WOM) model was connected to represent the market introduction of new products and their life

expectancy.

The first loop depicted in Fig. 3 starts with an order received by the separate business unit

(SBU1), which requires order fulfillment leading to a level of customer satisfaction. With higher

customer satisfaction, more requests for proposals (RFP’s) are received and therefore more orders

are processed. This loop can be either spiraling upward or falling over time to zero. On the other

hand, SBU2 is a different business but has a similar pattern, shown in Fig. 4. Orders received must

be successfully fulfilled with a sufficiently high customer satisfaction in order to maintain or


Fig. 2. Stock and flow diagrams.

Fig. 3. Order growth of SBU1.

preferably increase the number of RFP’s to create a virtuous upward spiral. Otherwise, the business

volume over time may go to zero.

The causal loop diagram (Fig. 5) depicts the two separate business units that are strategically

and synergistically attached and exist within a larger enterprise organization. The order win rate

refers to the number of orders received by the SBU’s compared to the total number of bids as a

percentage of requests and is very important to create this organic growth. The goal is to win

>90% of all bids. It is the intention that these two SBU’s in this model and other SBU’s in the

organizational system grow at as fast a rate as possible while exceeding the required cost of

capital, creating a positive EVA while being composed of the portfolio of companies and their

respective portfolio of product lines. ‘‘Successful order fulfillment’’ and ‘‘customer satisfaction’’

are important factors in driving virtuous upward spirals of requested proposals, unit sales, and the

breadth of products offered by product development. Order fulfillment to be successful must

satisfy the customers’ and company stakeholders’ requirements to create this virtuous spiral.

Likewise, unsuccessful order fulfillment causes a downward spiral which cannot go below zero.


Fig. 4. Order growth of SBU2.

Fig. 5. SBU1 and SBU2 synergism via successful order fulfillment.

Orders must be successfully landed with customers ultimately pleased at their decision to place

the orders. The SBU’s must perform in compliance to the technical and contractual requirements

as well as meet or perform better than the budgets and schedules of the orders.

It is the organization’s intent to develop new product lines for different market segments, all of

which match the core competencies of a single SBU or the collection of SBU’s, as quickly as

possible. One purpose of this systems dynamics modeling effort is to create a template or

methodology for enhancing the quantity and quality of new product lines. It is the desire in

constructing and simulating this model to show management the importance of the SBU’s

(products and services) working very closely together in generating requests for proposals (RFP)

at a high rate, increasing the rate of new product launches and corporate growth. The expectation

prior to running the simulation is that the virtuous spiral loops will rise exponentially without

limit. Although no oscillations are expected in a perfectly running system, it is anticipated

because of lags or delays in different steps in the system that oscillatory behavior will occur. It is

expected that as the portfolio of SBU’s increases and as the portfolio of their product lines

increases, the oscillations will dampen. This behavior is analogous to the diversified stock

portfolio finance theory, that the more diversified a stock portfolio becomes, the less volatile

becomes its total value. Delays have been introduced in the time to go from high quality

proposals to actual unit sales, and separately in the time it takes to generate new products from

product development efforts.

The depiction of the causal feedback loops diagram is an excellent tool in team management

meetings for developing corporate strategy. Several other causal loops were developed to add

workforce training, supply chain, financial structures, and the competitive environment.

4. Stocks and flows

Stocks and flows were added to the model. The conversion from the causal loops to stocks and

flows required many iterations and calibrations to obtain the different differential and auxiliary

equations (and respective parameters—see Fig. 6). The model shown in Fig. 7 hereinafter


Fig. 6. Example of equations.

referred to as ‘the core model’ (this core model was expanded to reflect the supply chain and the

human resources issues) depicts the inflection points where variation would cause a spiral. The

‘‘core’’ model consists of more than 40 simultaneous equations (including differential and

auxiliary equations). Adding the different sets of human resources, workforce training, supply

chain, and other additions, the model rises to over 125 equations (with more than 20 differential

equations).

5. Eigenvalues and elasticities

There are no methodologies available to see if the stock and flow diagrams/equations reflect

the original causal loops. We introduced the notion of using eigenvalue analysis and elasticities to


Fig. 7. The core model.

validate this ‘‘mapping’’. The ‘‘core’’ model is very non-linear (as verified by the state equations

and the phase diagrams) and expresses non-linear behavior. Therefore, it was decided to

investigate the eigenvalues of the system at different times.

To do the analysis, we linearized the systems equations at a point in time of interest. By so

doing, we were able to express the model as:

dX

dt¼ AX þ b

with X being our state vector for the ‘‘core’’ model by all the levels:

X ¼

Actual customers

Disguntled former customers

Potential customers

SBU1 available products with service

SBU1 available products without service

SBU2 service for current products

SBU2 services for discontinued products

2666666664

3777777775

The dynamic matrix A, and in particular its eigenvalues, determines the behavior of the

system. The different behavior modes of a linear dynamic system will be driven by the

eigenvalues of the dynamic matrix. The real part of the eigenvalue will determine the mode

stability. A pure imaginary eigenvalue will indicate never-damping oscillations with the period

equal to 2p divided by the eigenvalue. A complex eigenvalue will identify oscillations and either

growth or decay, depending on the sign of its real part (Elberg, 2000). A negative real part will

indicate decay or goal seeking modes, whereas a positive eigenvalue will cause exponential

growth (positive or negative). In all cases, the associated time constant will be the inverse of the

eigenvalue’s real part. The time constant is determined by the real part while the imaginary part

determines the frequency of oscillation. This is shown graphically in Fig. 8. Note that complex

eigenvalues always come in conjugate pairs of the form b a j�, where j = �1. Therefore, the

polarity of the imaginary part is not relevant (Elberg, 2000).

The first step for the eigenvalue analysis is to obtain the dynamic matrix that represents the

behavior of the system at a given instant (because of the non-linearities this dynamic matrix A

will change with respect to time). This discrete dynamic matrix will be obtained by linearizing

the differential equations at every time simulation step of the dynamic simulator (e.g., Maple,

Vensim).

Assuming interest in the dynamics at time T = 1.5 and 12.5 years, the matrices A at time

T = 1.5 and 12.5 years and their respective eigenvalues are depicted in Figs. 9 and 10.

How can we know which is the actual eigenvalue behind the behavior of the variables ‘Actual

Customers’ and ‘Disgruntled Former Customers’? Are there any causality relationships among

them in the stock and flows diagram (that reflect the original causal loops)? In order to answer

these questions, one has to take a look at the sensitivity (called elasticity in this paper) of the

eigenvalues at the interlevel (i.e., interrelationships among stocks) strengths. In the case of the

first eigenvalue at T = 12.5 years, the contributions are depicted in Table 1. The elasticities are

obtained by changing the values of the different eigenvalues using the method of sensitivity

analysis. This sensitivity analysis process identifies which eigenvalues affect a specific stock by

taking into consideration the changes produced in the behavior of the stocks. This sensitivity

analysis will not only individually change a specific eigenvalue but also their combinations (as



Fig. 8. Behavior based on eigenvalues.

Fig. 9. Eigenvalues of the core model at T = 1.5 years.

dictated by design of experiments procedures). The non-linearities present in the model make this

sensitivity analysis necessary to be repeated at each interval when linearization is applied.

A complete analysis included the different eigenvalues and different representative times of

the simulation (we are simulating for the next 25 years). The conclusions for the core model were

very clear from these analyses. The eigenvalue analysis indicates that the dynamics in the model

are dominant in the first years. However, there are fewer dynamics in the following years. In

addition, the link elasticities (related to the causalities and loops) indicate that there are not many

links between the product development part of the model (SBU1 available products with service,

SBU1 available products without service, SBU2 services for current products, SBU2 services for

discontinued products) and the market part (actual customers, disgruntled former customers,

potential customers). These were issues that were addressed to improve the model and a better

reflection of the causal loops. This finding prompted more sessions with the different executives

consulted earlier and resulted in a review of the equations and parameters.

6. Validation using case studies

The emergent behavior of the model was validated by using examples of industries which

exhibit various rates of evolution (this rate of evolution is labeled ‘‘clockspeed’’ by Fine, 1998).

We know that each industry evolves at different rates, depending in some way on its product

clockspeed, process clockspeed, and organization clockspeed. For example, Intel has a product

known as the Pentium II that has a market life of 2–4 years. As for its ‘‘process clockspeeds, at

each time Intel sinks a billion dollars into building yet another microprocessor superfactory, it

expects much of that investment to be obsolete in little more than 4 years. That gives Intel a 4-

year window to recoup its outlay of billions of dollars in capital, plus achieve a return on that

investment.’’ Due to this clockspeed the service organization for Intel is expected to be very

limited. This contrasts with the automobile companies which ‘‘typically refresh their car and

truck models every 4–8 years. In the process domain, they expect that a billion dollars invested in

an engine or assembly plant will remain vibrant for 20 years or more.’’ The service organization


Fig. 10. Eigenvalues of the core model at T = 12.5 years.

will provide a good value to these companies. At the lowest end of the clockspeed scale are the

aircraft manufacturers. Boeing measures the clockspeed of its products by decades! Therefore,

their service organizations can create good profitable opportunities. As a matter of fact Boeing

started offering MRO (maintenance, repair, and overhaul) services just 5 years ago to capture the

value of services.


Table 1

Elasticities of eigenvalue 1 (simulation time = 12.5 years)

Link Sensitivity

Actual customers ! actual customers �1.66

Actual customers ! disgruntled former customers �1.66

Actual customers ! potential customers �1.66

Actual customers ! SBU1 available products with service Not connected

Actual customers ! SBU1 available products without service �1.66

Actual customers ! SBU2 services for current products Not connected

Actual customers ! SBU2 services for discontinued products Not connected

Disgruntled former customers ! actual customers Not connected

Disgruntled former customers ! disgruntled former customers 3.96

Disgruntled former customers ! potential customers 2.71

Disgruntled former customers ! SBU1 available products with service Not connected

Disgruntled former customers ! SBU1 available products without service Not connected

Disgruntled former customers ! SBU2 services for current products Not connected

Disgruntled former customers ! SBU2 services for discontinued products Not connected

Potential customers ! actual customers 2.099

Potential customers ! disgruntled former customers Not connected

Potential customers ! potential customers 1.042

Potential customers ! SBU1 available products with service Not connected

Potential customers ! SBU1 available products without service Not connected

Potential customers ! SBU2 services for current products Not connected

Potential customers ! SBU2 services for discontinued products Not connected

SBU1 available products with service ! actual customers Not connected

SBU1 available products with service ! disgruntled former customers Not connected

SBU1 available products with service ! potential customers Not connected

SBU1 available products with service ! SBU1 available products with service �1.66

SBU1 available products with service ! SBU1 available products without service Not connected

SBU1 available products with service ! SBU2 services for current products Not connected

SBU1 available products with service ! SBU2 services for discontinued products Not connected

SBU1 available products without service ! actual customers Not connected

SBU1 available products without service ! disgruntled former customers Not connected

SBU1 available products without service ! potential customers Not connected

SBU1 available products without service ! SBU1 available products with service �1.66

SBU1 available products without service ! SBU1 available products without service 0.995

SBU1 available products without service ! SBU2 services for current products �1.66

SBU1 available products without service ! SBU2 services for discontinued products Not connected

SBU2 services for current products ! actual customers Not connected

SBU2 services for current products ! disgruntled former customers Not connected

SBU2 services for current products ! potential customers Not connected

SBU2 services for current products ! SBU1 available products with service Not connected

SBU2 services for current products ! SBU1 available products without service Not connected

SBU2 services for current products ! SBU2 services for current products �1.09

SBU2 services for current products ! SBU2 services for discontinued products �1.66

SBU2 services for discontinued products ! actual customers Not connected

SBU2 services for discontinued products ! disgruntled former customers Not connected

SBU2 services for discontinued products ! potential customers Not connected

Therefore, we use aggregate data from several companies to analyze qualitatively the

emergent behavior produced by the model. The companies selected were Dell at the fastest

clockspeed (Farhoomand et al., 2000; Rivkin and Porter, 1999; Thomke et al., 1999), Intel (Bell

and Leamon, 2000; Brandenburger and Nalebuff, 1996; Casadesus-Masanell and Rukstad,

2001), and Boeing at the slowest clockspeed (Garvin et al., 1991; Imberman, 2001; Ovans,

2001). The model produced good results for the different sets of enterprises. For a global

enterprise with a slower clockspeed such as Boeing, SBU2 produced a sustaining growth

(reinforcing SBU1). In the case of a global enterprise with a faster clockspeed such as Intel,

SBU2 was not the key to sustaining growth (the emergent behavior of the model did not produce

a SBU2!). The innovation (i.e., requests for proposals) and WOM were important factors. With a

global enterprise such as Dell, the supply chain was a very important factor to sustain growth.

This validation was also performed concurrently with the validation using the eigenvalue

analysis. The eigenvalue analysis varied in the strengths of the causalities and strengths of the

feedback produced.

7. Results (development of policies)

The model now was exercised for an aerospace company. SBU1 produces complex, large-

scale automation systems for various markets and especially aerospace. Its parent has

investigated the acquisition market for growth and has decided instead to create growth, using

its custom design and project oriented core competencies, by developing product lines in each

market segment within the industry of assembly automation. SBU2 is the service organization

for the different product lines. The equilibrium quantities in the core model shown in Fig. 4 are

the inflection points where variation would cause a spiral. Any oscillations in the model are

quickly overcome by the strength of the two main positive reinforcing loops. Each variable in

the model must have a policy or set of policies aimed at preventing precipitous downward

spirals, and they must be designed to address these inflection points (Grove, 1996). The

methodology developed to obtain the policies was sensitivity analysis. In addition, non-linear

programming using objective functions and the corresponding adjustment of parameters to

modify the eigenvalues in order to minimize the ‘‘undesired’’ behavior was also used to

develop policies.

7.1. Critical balances and rates of change

Rate imbalances are sources of limits to growth or decline. An interesting question is

whether the enterprise management can rely on the invisible hand to keep the balance of

growth between the two SBU’s or does it have to closely monitor the balance and exert

controls to assure the balance. For instance, a situation can occur where there is an

inadequate service staffing or infrastructure to assure SBU1’s products are properly serviced;

as a consequence SBU1 might grow faster than SBU2 can adjust to close the gap between the

symbiotic rates of growth. An imbalance is dangerous since the satisfaction level will be

reduced, and at some critical point the vicious spiral downward will start. The danger of a

downward spiral generates the need to maintain metrics for enterprise and SBU management

as well as worker tracking to keep key controllable variables within tolerances or above

thresholds. These key balances are cash availability, SBU1and SBU2 growth rates, customer

satisfaction, infrastructure growth rate, risk aversion, order win rate, and diversified portfolio

theory adherence.


7.2. Cash

As is typical in growth scenarios, cash is required to keep up with the pace of growth. The rate

of positive net cash flow generation must equal or exceed the rate of cash needed for new product

development and other organizational needs. If cash becomes the constraint, then growth will be

slowed. Management can choose to obtain more cash by stock offering, but this has a dilutionary

effect on current shareholders and can be used only if the result is that the discounted future

stream of earnings per share will be higher. Another option is to use debt. The cost of debt has two

issues, the interest rate and the risk of not having enough cash later to pay the interest and

principle. These issues make debt a poor choice to fund corporate growth for a company, which

will need increasing amounts of cash in the future to fund the growth. A lack of cash crisis can

significantly limit growth and potentially start a downward spiral.

7.3. Growth rate of SBU1 vis-a-vis SBU2

The rate of SBU1 growth must be equal to or slightly less than SBU2 as measured by products

with service compared to products without service (SBU1 Available Products with

Service > SBU1 Available Products without Service). Because of the time to develop SBU2

services and product transition rate, SBU2 must keep metrics indicating the rate of product

development, with time allowed to assure services are available ideally just before the product/

market requires the services. This is because SBU1 must have services to test and install their

products, and SBU2 must at the same time maintain the installed base of systems to keep order

fulfillment on time and the customer satisfaction level high.

7.4. Customer satisfaction

The faster that product development and production can bring products to market in a high

quality manner, the better customer satisfaction level. Operational effectiveness of equipment is a

current terminology being used to emphasize uptime utilization of equipment. The greater the

uptime the higher the satisfaction, especially in bottleneck zones which determine the rate of

throughput.

There is clear evidence in the model that the time to develop products has a dramatic effect on

the business. A policy must be targeted to reduce time-to-market. Time also has an almost 1:1

correlation with cost although not modeled here, so cost will likewise be reduced if the time-to-

market is quicker. The average product per RFP is highly elastic with the SBU stocks as

expected—the greater the ability to produce products from RFP’s the better. The RFP per

customer factor is very important to the behavior of the SBU stocks. Naturally, companies want

to increase the number of proposals submitted to the customers and increase the winning

percentage to create a thriving enterprise. The ‘‘SBU2 successful orders fulfilled %’’ is highly

related to quality. Frankly, lateness on orders cannot be tolerated due to the onerous effects on

increasing costs (not modeled) and degradation in the business. For project management, if

delays arise in a schedule, they must be diligently time recovered back to plan, such as by using

techniques as explained by Goldratt (1997).

Customer satisfaction is an absolutely critical dimension to take into consideration for the

success of the modern enterprise. It is essential to consolidate, analyze, and establish policies to

guarantee customer satisfaction. This policy-making process should occur throughout the

marketing, sales, and service stages.


Improved sociability will increase customer satisfaction, which can be accomplished by direct

contacts, newsletters, advertising, newsgroups, and ‘‘chat’’ rooms related to the products and

services. If sociability is low, then even high order fulfillment will not increase the satisfaction

level. The model indicates that customer satisfaction should be approximately 95% to sustain

upward virtuous spiral. Even a 90% rate will not significantly raise the positive loops.

7.5. Infrastructure

Another potential source of imbalance is the rate of infrastructural development to assure

successful order fulfillment. Infrastructure development can be quite complex. Support process

systems and tools take time to design, debug, train and become accepted in use. Infrastructure is

also a process to control the organization and prevent downward spirals. Providing the process

improvements for the organizational structure to support and encourage further growth is a

difficult management challenge. In addition, synchronization of the supply chain infrastructure to

the needs of the organization in the form of timing, quantity, quality, logistics, and price is vital

for successful order fulfillment and customer satisfaction.

7.6. Human resources

Talents can be difficult to recruit and retain in the competitive market. Training takes time and

is expensive:

� Having a good ratio of ‘‘Services’ Staff/Customer’’ usually means good customer service. This

typically increases the number of customers (via ‘‘word of mouth’’).

� More investments in the service organization (SBU2) cause more investments in service staff

and more development of services. Investments in service staff increase the recruitment and

training of the service force, which takes time. Professional development of an employee is

very important. Recruitment and training of the service force takes more resources than the

development of services.

� More services, more customers, and more products increase the sales and therefore profit in the

organization.

� The attrition of employees with ‘‘early’’ retirement packages usually affects the professional

development of the new ‘‘recruited’’ employees.

� It is important to invest in the services and product organizations. However, it is essential to

keep a balance. Services are very dependent on ‘‘good’’ products.

7.7. Supply chain infrastructure

The infrastructure and the organization strategy to synchronize the supply chain are very

important. The enterprise needs the required resources from suppliers (i.e., right time, right

quantity, right quality, right place, and right cost) to deliver products/services to customers

(again, right time, right quantity, right quality, right place, and right price).

7.8. Risk and change aversion

Growth limitation and potential decline is well described by Christensen (1997), which is the

failure of the organization to adjust to and accept change both from the outside and within the


organization. This reluctance to leave the comfort zone of the organization and put on blinders to

possible technological substitution is a serious problem frustrating the sustainability of the firm.

The model deals with this problem by emphasizing product development, which is change.

However, will the products be merely incremental or will they be truly breakthrough? A possible

policy could be instituted to engage in a product development when it has at least a 50% cost and

time-to-market (TTM) improvement, as well as features superior by a 50% margin over the

competition. This could be called the 50% rule. If the new product development candidate is 50%

lower cost (at least to build; one could sell at a higher market price if possible), 50% faster TTM,

and 50% better, then the project will be undertaken. Products must also be diversified into

different market segments. The organizational structure must foster this product/market

approach.

Risk aversion is captured in the model by the variables ‘‘SBU1 response rate to customer

requests’’ and ‘‘average product per RFP.’’ If SBU1 is overly averse to risk, it will respond to only

those few customer quotes that the SBU is confident it has a good chance of winning. The

response rate would be lower for risk adverse firms versus risk taking firms. The firm’s reputation

(word of mouth among customers) is modeled as a function of the ‘‘relative winning proposals,’’

which includes those RFP’s that the firm is asked to quote but refuses to quote. Therefore the

model penalizes firms that are adverse to risk. At the same time the firm needs to be able to deliver

on the quotations that it accepts by winning a good percentage of those quotes. For firms that are

risk takers, they should be able to develop new products (and new markets/customers) by

developing a larger number of products for each RFP that they win. Finally, the firm must be able

to come through on these products in terms of both quality and delivery. These factors are shown

in the model as contributors toward the customer satisfaction variables. In the future the model

could be expanded to tie some of these variables together by modeling and capturing the optimum

level of risk (i.e., a good balance between a high number of quotations accepted, a high number of

new products per RFP, and acceptable quality and delivery for these riskier projects).

7.9. Order win rate

The model is very sensitive to winning percentage. A slight change sends it spiraling upward

or downward very quickly since most of the loops are positive. Customers can be very sensitive—

they can be very cold and unforgiving, or they can be very tolerant depending a great deal on how

they are treated. This sensitivity variable cannot be ignored. Customers must be responded to

accordingly.

7.10. Adherence to diversified portfolio theory

SBU1 must incentivize to not only pursue growth from within the current product families but

create new product families in new market segments as well. Typical behavior is that people/

companies like to stay in their zone of comfort. Even in a product development area, companies

tend to improve on the known past in the form of incrementalism rather than take the risks of

going into new breakthrough territory. Usually the incentive systems in corporations inhibit or

penalize those who take a daring risk and lose. The sanctions imposed serve witness to others in

the organization to not take these risks. However, the model requires that SBU1 diversifies into

new market segments as well as grows new products within existing market segments. A risk and

reward system must be modeled which demonstrates the desired behavior of multi-market

segment penetration with new product families.


Top management must emphatically lead the way in funding system architectural teams in

the different market segments. It seems that the biggest gamble lies in accurately

understanding the wants of the perceived new market segment. An outsider to a market

has the disadvantage of not intimately knowing the customer drivers without a thorough

market research investigation. On the other hand, the outsider might be able to bring into the

market space fresh new ideas unencumbered by past paradigms. A policy for entering into the

new market space must account for the added risk, perhaps requiring again that the new

product line be at least 50% lower cost to build, 50% better in features, and 50% faster in

time-to-market, with an equal or preferably enhanced quality compared to existing

competition.

7.11. EVA measurement and incentive system

Rather than profitability, EVA has been chosen as the financial theoretical basis for

measurement of the firm’s successful contribution to society. The contribution margin policy

must be able to sustain a positive EVA along with a positive revenue growth.

8. Conclusions

Many processes, and the attentiveness of skilled workers to the processes, underlie the

success or failure of the interrelationship of SBU1 and SBU2. The model developed herein,

being a high level of abstraction, cannot cover every aspect of the operation of the enterprise.

However, it emphatically points out the sensitivity of successful order fulfillment in products

provided and necessary services support to the customer, as well as the balancing of

interrelationships of other variables. Any deviation from the equilibrium points affects

customer satisfaction and the positive reinforcing loops. On the one hand, the model

demonstrates in an alarming manner the precipitous downward spiral which is exhibited too

often in the life of an enterprise. On the other hand, the model shows the virtuous spiral of

growth that is sustainable forever if the inflection points are not crossed. Borrowing from

sports, the term momentum can be used to describe the upward spiral. By encouraging

managing above the inflection point thresholds, the model shows the way toward continued

corporate sustainability and enrichment of the company’s shareholders, customers served and

other stakeholders.

References

Bell, D., Leamon, A., 2000. Compaq Computer: Intel Inside? Harvard Business School Case Study.

Brandenburger, A., Nalebuff, B., 1996. Inside Intel. Harvard Business Review.

Casadesus-Masanell, R., Rukstad, M., 2001. Intel Corp. 1997–2000. Harvard Business School Case Study.

Christensen, C., 1997. The Innovator’s Dilemma. Harvard Business School Press.

Elberg, E., 2000. A System Dynamics Approach to the Software Development Business Problem. M.S. Thesis. MIT Sloan

School.

Farhoomand, A., Lovelock, P., Ng, P., 2000. Dell: Selling Directly, Global. Harvard Business School Press.

Fine, C., 1998. Clockspeed. Perseus Books.

Forrester, J., 1971. Principles of Systems. Pegasus Communications Inc. (reprint).

Garvin, D., Field, L., Simpson, J., 1991. Boeing 767: From Concept to Production. Harvard Business School Case Study.

Goldratt, E., 1997. Critical Chain. North River Press Publishing Corporation.

Grove, A., 1996. Only the Paranoid Survive. Bantam Doubleday Dell.

Imberman, W., 2001. Why Engineers Strike: The Boeing Story. Harvard Business School Case Study.


Ovans, A., 2001. Inside Boeing’s Big Move: A Conversation with John Warner. Harvard Business Review.

Rivkin, J., Porter, M., 1999. Matching Dell. Harvard Business School Case Study.

Senge, P., 1994. The Fifth Discipline. Currency Doubleday.

Sterman, J., 2000. Business Dynamics. McGraw-Hill.

Thomke, S., Krishnan, V., Nimgade, A., 1999. Product Development at Dell Computer Corp. Harvard Business School

Case Study.


sustaining growth in the modern enterprise

Documents