Table of Contents
:
A Simulation-based approach to understanding the dynamics of
the transition from product manufacturer into service provider
Heiko Gebauer
Institute of Technology Management
Univesity of St Gallen
CH-9000 St. Gallen
Unterstrasse 22
Phone: 0041 71 228 24 02
Fax: 0041 71 228 24.20
Email: heiko gebaue@unisgch
Abstract
Despite the demonstrated benefits of an extended service business in driving conpetitive
advantage, most product manufacturers found it extremely difficult to manage the transition from
a product manufacturer to a service provider successfully. In this context, we observed different
phenomena. One phenomenon we term the “service jungle”. The “service jungle” describes the
phenomenon that the transition has often led to declining business because of increasing costs,
which could not be recovered with corresponding returns.
Unfortunately, existing theory does not explain sufficiently the challenges inherent in this
transition process. Transitioning from a product manufacturer to a services provider involves
interactions among service management theories, basic concepts for improvement programs and.
human decision making processes in the firm while existing frameworks tend to address one at
the expense of the other. To resolve the ,, service jungle’, we develop an integrative framework
combining existing service management theories, the basic concepts for improvement programs
with established theories on human cognition, learning, and organizational behavior. We then
translate that framework into a ral model and analyze it using computer simulation. The
analysis suggests a new construct the degree of integration among structural adoption, activity-
oriented adoption, behavioral adoption characterizing the dynamics of the transition The
construct provides an intemally consistent theory about the dynamics of the transition. It shows
howa lowdegree of integration can create the “ service jungle” phenomenon.
TI.1 Research method and data analysis .
II.2 Data collection...
IV. An integrative framework for the transition.
IV.1 Activity- oriented dimensions
IV.2 Structural dimensions for the transition.
IV.3 Behavioral dimensions for the transition
IV.3.1 Basic concept for adopting the outlined structural and activity- oriented dimensions 13
IV.3.2 Behavioral challenges and cognitive processes
IV.3.2.1 Managerial options
IV.3.2.2 Managerial bias against improvements and managerial service awareness.
1V.3.2.3 Employee pull eff 60t oe eesesssescsescnsscnseeceneerseceee
IV.3.2.4 Employee's bias against second order improvement
IV.3.3 Configuration of the behavioral dimensions
1. Introduction
The following paper offas a theoretical framework for undestanding the tansition from pure
jaoduct manufactres into savice provides. According to Chese (1981) ad Oliva ad
Kallenbay (2002), we assume that thee is a oontimmm fiom pue poduct memfacires to
service providers. Product manufactes ty to move along that line as they extend their savice
business. At the first exteme point of the continuum, we envisioned a product manufacturer that
yproduoes core products and services are just an Add-on to the product The pofits and revenue
are mainly generated through company’s core producs and the contribution of services in tems
of revenue, profit and customer satisfaction is quite low or even negative. At the second exteme
point, we envisioned, a savice organization for which its poducs ae just am Addon to the
services and represatt only a smell pat of the vdue poposition (figue 1). The ngjaity of
company’s value proposition comes from its service offerings.
Considerable research in the fidd of management has revealed the need for traditional product
manufactures to extend their savice business. The need is driven by the mimeous opportunities
that savices can provide financa opportmities (Mathe and Shain, 1993), makeing
opportunities and strategic opportunities The substantial revenue, the higher margins and the fact
that services are a more stable sourve for revenue represent the financial benefit (VDMA, 1998).
An extended savice business can genaate a incessing revewe and pofit, compensding
decreasing product margins (figwe 1). Marketing opportumities can be undestood as “better
sexvioes for sdling more products’ (Mathe and Shapiro, 1993 p. 33). Finally, thee ae straiegic
arguments like competitive strategy based on services or savices as a market enty batier
(Anderson and Nas, 1995; Evarsson, 1990; Oliva and Kallenbag, 2002). Savices by bang
more labor dependent ae nuch nor diffialt to imitae They ae becoming a sstainable
source of compditive advantage (Simon, 1993). For example, Lay & al (2002) have found out
that compared to seveal strategic options like fostering innovation / technology, product quality,
suitability to customer needs, cost leadaship and ddivey time compding through savices
enables companies to eam the highest margins.
The enty point to our theorizing is povided by our obsavation that most companies found it
extemdy difficult to manage the transition successfully. In this context we obsaved diffeent
Phenomena. One phenomenon we tem the “savice jungle’. The “savice jungle’ describes the
Phenomenon that the transition has often led to declining business because of increasing costs,
which could not be tecoveed with comesponding reums. Most companies stated savice
yograms to move dong the tansition line (figure 1). Unfortunately, companies have found it
difficult to sustain the initiated savice programs. The most sevice programs led to increasing
service offering and higher costs but did not lead to conesponding réums The contibution of
services in tam of revewe and pofit did not inceese in the intended wey (figue 1). The
companies left the transition line, gave up their savice programs and went into that we call the
“service jungle”.
Contribution of Services
“Service revenue
“Service profit A
‘Customer satisfaction
Suocessful Transition
High
‘Unsuccessful Transition
>
Product Mamufactie gee Timé
Service as add-on’
Figue 1: Transition line from product manufactures into savice provides and the _,savice
jungle’ phenomenon.
The reason for their difficulties stems from a lack of theoretical frameworks and management
guidance about how the transition should be implemented Scholars have recently answered the
question why product manufachres should undago the tansition info a savice povider
Unfortunately, they are surprisingly sparse on describing how the transition should be done.
II. Existing Theories
The raional for the temsition can be put forth dong three lines - financia, makeing ad
stategic aguments. But up to now, thee is no infegeaiive framewok eqlaning how the
transition should be implerented The existing literawe is sumprisingly sparse on descibing the
Challenges inherent in the tansition process (Oliva and Kdlenbag 2002). For example, Oliva
and Kallenbay (2002), Mathieu (2001b) and Homburg e al (2000) give some recommendations
on how the transition can be done. But, however, there ae two significant gaps in the literature
on the transition from product manufactures inio service providers.
Firs, a common thread mmning through = savice management framewods is ther focus m
modifying organizational stucures and fins activities For instance Oliva ad Kdlenbag
(2002) describe how to savice the installed based awl how to expand the savice offerings.
Honbug & a (2000) eqlain dso swea impovemnt atiivities that company hwe to
implement They explain how to stucure the existing savice offeings how to change the
communication how to devdop new savics ad so on But they co mt gve ay
recommendation on how to change the underlying behavior pattem and company’s culture.
Overall, the service management frameworks pay less attention to the concomitant behavior
changes required to modify organizational structures and finy’s activities. They put the necessary
behavioral changes at the implementation leva and do not offer some guidance on how the these
changes can be managed Nevethdess, the existing framewoks povide a excdiat
navigational chat for manages facing cellenges dung the modification of omenizaionl
structures and firm's activities.
In contest, organizational scholars have focused primarily on the behavioral aspects of change
(van de Ven and Pool, 1995, Huber and Glick 1998). However, whereas savice management
theories layely ignore the behavior of whose working within the omenizdion, omgenizaiona
theories genaally do not account for orgenizational stuchres and fim’s attivities We think
there is a dear need for an integrative framework that combines organizational struchmes and
fimis activities with undastanding of behavior chllengess to eqldn the tansitim from
product manufachwas info savice povidas) Mathie: (2001b) tied to oonbire the
arganizaional structure and ailtural perspective She descibess seve types of savice
manewes on the tarsition line Each savice manewer requires changes in ompanizdionl
stuctre and company’s culture But she does not explain the changes of the omanizaionl
stucture in detail as well as sevedl change management phenorema like resistance against
change and so on. Changing company’s culture always lead to such phenomena.
The second significant gap in the literae stars fiom the imdhility of existing fraveworks to
describe the transition as a dynamic process and is not surprising. The transition is by definition a
dynamic process. Each theory devdoped to eqdain the tansition defines ether eqiicily or
implicitly, a dynamic system of causa loop structures. When devdoping theories that explain the
dynamics of the transition, one has to check its logical consistency. Checking logical consistency
requires deemiining whether or not our integative framework is capable of genadiing the
behavior that it puports to explain Fstablishing whether our infegetive frameworks can
generate the dynamics of the transition, we have either to mentally solve the mubel renforing
links among the diffeent indicated factors or to smilde a system of diffeentia equations
Unfortunaiely, the rdiance of intuition to enfore logical consistency is not in line with wel-
established theories about limitations of human cognition Numacus sudies on the diility of
Tanages to contol a dynamic process or on researches tying to undestand the dynamic
consequences of their theories have shown that human dility to reliable infe or mentally
simulate the behavior of even low order dynamic systems is limited (Staman 198% and 198%,
Sestty 1997). Through translating our integrative framework into a fonmal modd, we ty to dose
that gap.
We use the existing theories and frameworks and ty to fonmulde an integrtive framework that
combines omganizaionl stucures ax fim’s attivities with undestaning of bdwiod
challenges. The purpose of this paper is to devdop the begiming of sich an iniegadive
framework. Three management profiles build the coe of our integeiive framework They
illustate the main success factors for tansitioning form poduct mamfacures into savice
moviders. Furthermore, these management profiles show which factors or dimensions have to be
adopted and how the different dimensions have to be configured in the sight way. Furthemor,
these profiles offer some guidance for managers seeking to design a successful transition.
Because explaining all implications of the tamsition (stating with the necessay adoption of
organizational structures and going further with the changes in finm's activities) goes behind the
soope of that paper we focus mainly on the behavior sie of the tansition Organization
structure and finm's activities are not explained in detail.
The paper is organized as follows. The next section desciibes our research methodology, data
collection and cefa analysis In section four we present an integrative framework that captures
the transition process from product manufactures into savice provides. We desaibe seved
structural and activity-oriented dimensions and propose a casal loop siucure to capure the
mein characteristics of the behavioral dimensions. In section five, we tanslde the integrative
framework in a fonmel modd, analyze the dynamic behavior of the transition and test the intemal
consistency. In section six implications for future research and practitioners are discussed.
TI. Research method, data collection and data analysis
IL.1 Research method and data analysis
Cese studies ae the main tool for theory devdopment We focused on Eumopean poduct
manufacturers whose products represent an high investment for the customers. The case studies
medical equipment tdecommumication industies dc). All investigated product mamfacures
have been looking for possibilities to enhance their pofitability through savices because the
yaoduds wee mainly in the matmity stage with deceasing mains and profitability. The results
and implications are limited to this area
Our research method oonsists of four phases. The first phase is based on hipolar typed case
studies (Eisenhardt, 1989). In the second phase, we devdoped a theory to explain the evolution
of the transition We used causal loop diagrams capturing the rich array of interdependencies and
feedback processes in the company and its ewionment Additionally, we used struction
literate (Giddens, 1984) to stress mutal, recusive caxd links among omganizational stuchie,
and the mental modds of organizational actors which guide their behavior. We used the system
dynamics method (Fonester 1961) to undastand the multiple feedback mechanisms that affect
the tensition.
We first began to identify the pattems of infaest Then we oontimed with am iteative
devdopment of categories, variables and causal links. The categories into which our observations
can be coded ae successful and unsuccessful transition processes (“savice jungle’). The
vatiables and causal links form the feedback processes that genedie the dynamics of the
transition, At the end, we injegred our findings in 2 unified framewok eqpaining both the
successful and unsuccessful transition processes. The result is a single set of feedback processes
capable to generate pattems of phenomena we observed.
As our causal loop structures enaged, we reviewed each link in the causal maps to asses
whether the rdationship wes supported by existing theory in the third phase Our framework
integrates the characteristics of savice management, the stuctues of impovenent pograms
and the behavioral side We draw on the basic precepts offered by service management (Simon,
1998, Oliva and Kallenbay, 2002, Mathien 200la and 2001b). By investigating the bewviord
dimension of the tansition, we dy on the ldest studies of impoverent pograms (Steman,
Repenning, Kofman, 1996, Stenvan, Repenning, 2002) and on basic concepts of humen decision
meking like judgement under unvatainty (Kabneran, Slovic and Tvesky, 1982), bounded
ralionality (Simon, 1957) and the valance-expectancy-theory (Vroom, 1964).
By combining the single set of feedback processes and the existing theory, we wee dle to
formulate an integrative framework to explain the transition The transition can be understood as
a goal-oriented adoption of company’s organizational structure, activities and behavion $= The
activities describe how a sequence of linked and interdependent working processes transform
inputs into the desired outcomes. Belwvior is defined as the undelying behavior pattems 90
deeply embedded and recurrent that they ae displayed by the most membas of the agenization
(Gavin 1995). The behavior pattems effect the decision making. It has no independent existence
apart from activities and organizational structure in which they appear They affect pofoundly
the character of activities and structure by shaping how the attivities ae canied out and how the
carganizational structure is going to work.
By combining the single set of feedback processes and the existing theory, we wee dile to
formulae an integrtive framework to explain the transition The integrative framework is textual
and diagrammatic. But it is not completely able to describe the transition as a dynamic process.
We translated our integrative framework into a formal model during the fourth phase. The formal
or simulation moda allows us the deailed amdysis of the dynamic behavior aeated by the
undelying assumptions of our framework Our amlysis of the fomel mod highlights and
Clarifies the interactions between the different elements used in our integrative framework.
IL.2 Data collection
As already mentioned, the research design focused more or less on polar types in our first phase.
First we began to identify companies that wer highly successful by the tansition Second we
identified companies that struggled to manage the tansition successfully. To distinguish between
sucoessfil and unsuccessful companies, we used indicators like savice revewe mimber of
service offeings, savice profit and offsetting of savices as wel as indicators like asomer
sdisfaction. The primary dala oollection methods wee samstucted inteviews (for the
successful practices) and action research among the companies that stuggled duing the
transition The subjects of the interviews and action resarch were chosen with the input of the
research team.
The interviews with each of the five successful practice companies lasted one day. The interview
subjects began by describing their background and history with the organization They wee then
asked to give a detailed account of their expaience with the transition Subjects were asked to
assess the key success factors and banies of the transition and to offer hypotheses about their
causes. The data analysis started with traditional methods for inductive fidd work Our research
team read all the infeview tansciipis. Based on the trnscipls, we wole five detailed case
studies describing the history of each transition Participants and the research team seviewed the
ceses, identifying gaps in the nandtive and suggesting ackitional dala collection The reviews
often led infeviewees to provide more daailed beckgound infomation By dlowing al the
participants to review their transition cases we oould offset some of the bias nonmally associated
‘with retrospective interviews.
Additionally, we tested our findings in action reseach among companies that struggled duting
the transition We conducted several projects and assisted more than 20 companies by doing the
tansition By reviewing our first finding in the fidd we found atkitiond issues that had not
been atisen at first We wee able to test our findings about how the tansition is influenced by
day-to-day work.
Similar to Oliva and Kallenbay (2002), all case studies were more or less selected according to
their perceived position along the continuum between product manufacturer and savice provider
The sampled case studies cover the whole transition line as wel as the phenomenon of the
“service jungle’. The hipolar case studies in combination with interviews and action research
hdped us to gourd or exaging casa loop smictue add inegaiive framewok on both
retrospective and current data.
Overall, the combination of the four phases helped us to ensure that our enaging framework is
gounded in the fidd cea, is consistent wih exising pimples in savice maagerst
Tmanagement of improvements, organization theory and liteare on human decision neking ad
is an intemal consistent theory able to generale the behavior it purports to explain
IV. An integrative framework for the transition
IV.1 Activity-oriented dimensions
During our intensive case studies, we indicated sveal stucura and activityoriented key
factors or dimensions that have a significant impact on the success of the tamsition These
dimensions emage dso from seveal causal loop stuctes. But we will not eqlain them in
detail. We just focus on the description of the dimensions and do not explain how they emerged.
Begiming with activityoiented dimensions, tensiioning fiom podut menfectres into
service providers requires the sdling of more and more services. Sdling more and more savices
implies that the astorer is willing to buy rew savios The astomes ae willing to by
services as long as customers benefit from these services.
To imovae savices addressing ofgin astomer neck, ore usely mess a@ Ouside-in
perspedive of the savice imovation pooss. Unforumatey, most poductt menfacures
devdop services from an Insideout perspective, because they ae mainly technology-driven
10
Services need more market pull activities The transition from product manufactures into savice
ymoviders requires a change in the perspective of the savice innovation process from an Inside
Out to an Outside-in view.
Based on an Outsidein oriented savice imovaiion process, the company is ale to imovae
marketdriven savices highly accepted by the astomeas. The Outside in savice imovedion
yprocess represents the first necessary adoption on the way to a successful transition process. It
leads to an extended service business through more and highly acoepted services.
According to Oliva and Kallenbery (2002) and Mathieu (2001b), the expansion of the savice
offerings changes the focus of the value proposition to the customer Mathieu describes it as the
change from satvices supporting the product to savices supporting the client As the company
moves dlong the dimension savvice offerings, the product becomes just a part of the offering and
is not opposed to be the center of the value proposition any more (Oliva and Kallenbay 2002).
Focusing service offerings on savices supporting the dient is equivalent to shifting from product
manufactures into “solution providers’. It can be also understood as the devdopment starting
with sdling products and give away savices, followed by sevedl steps like savioss ae sold
with hardware, savices ae sold integraied in the total offeing the total solution is sdd and
ending up with the total solutio is sold as a savice (Benchmarking Report 2001). Establishing
services supporting the client and changing the focus of the existing savice offeings has two
importantinpects.
At first it requires an adoption of the existing compdiitive strategy. Services supporting the
ywoduct mainly support the differentiation through products They do not emble a savicebesed
compditive strategy. Services supporting the dient (and moving to solution provides besed on
comprehensive or integrated services) need a savice-besed compditive strategy, because the
yoduct is no longer the core offering (Wise and Baumgartner 1999). It just can be seen as an
Addon to the savice offeings In case of introducing savices supporting the diet the
compdiitive strategy has to change from a product oriented to a service oriented strategy.
At second, an adoption of the marketing is also required. The customer interaction changes from
a transaction-besed to a rdationship-based approach Establishing a rdadionship-based customer
interaction is mainly besed on bundling existing transaction besed savices to rdationship-besed
services. For example, separate services around the product like repair, preventive maintenance
or inspection ae often bundled into maintenance contracts. The move towad a rddionship-
besed customer interaction is mainly driven by the desire to make better use of the existing
service capacities This argument explains how the savices provider benefits from :dationship-
besed customer interaction, there is mo compdling agurent on how astomes should benefit
11
from this kind of interaction The poblem is that rdaionshipbased customer interaction does
not add value direcly to the astomen The astomer benefit only when the rdafionship-based
sevice offaings ae transfered into an higher equipment availability and ae priced accordingly,
that the customer has the ability to quantify the value of the offeings (Oliva and Kallenbay
2002).
Another argument comes from the asymmetic infomation between the fim and its custome.
Savvioes have by definition more credence aitibutes than products (Meffat and Bruhn 2000).
That means, the customers have difficulties to judge the savice quality and benefit Unoatainty
and risk aversion are a besic feate of humen decision meking (Einhom and Hogath, 1985).
Therefore, customers prefer catain to uncatain oufoomes An established rdationship-besed
customer interaction can reduce the uncertainty and can be used as a reputation In that case, a
idationship-besed customer interaction can be understood as both the result of excellent savice
offerings and service quality as well as the necessary requirement for selling services.
By illustating our four activity-oriented dimensions or key factors for the transition, one ges the
following management profile The management profile (according to the St Galler Management
Moda) can be understood in the following way.
All four dimensions influence the success of the tansition If companies want to transition
themsdves fiom product memfacuwas to savice povidas al for ativityoiented
dimensions have to change in a consistent way. They have to be configured in such a way that
where is a ooalescence among them The coalescence is meant in a way that the onfiguation is
consistent complementary and mutual reinforcing. If companies want to move from product
manufactures to savice provides, the right configuration of the activityoriented dimensions in
tems of a coalescence will be on the outer citcle. The consistent configuration for companies
that do not want to move fiom product memufacies to savice provider would be the inner
aide An inconsistent configuration will not lead to a successful transition It leads direclly to
the “sarvice jungle”.
12
Activity-oriented dimensions
om
Outside.in Service-Oriented
——= Consistent configuration leading to a sustainable
and successful transition from product
manufacturers to service providers
=== Example for an inconsistent configuration leading to an
‘unsuccessful transition (service jungle) from product
manufacturers to sevice providers
Figure 2: Management profile for the adoption of the activity-oriented dimensions.
IV.2 Structural dimensions for the transition
Beside the activityoriented dimensions, we dso indicated four stuctueloriented dimensions
These structural dimensions explain necessary organizational changes on the way to a savice
provider They result from the diffeence betwen memifacuring and savice opadion.
Compared to manufactuing opadiions, savice opadions require usally a processorented
approach for the value acreation, low hiearchy, an high degre of de-oantalized responshility
and different multidimensional metrics to measure customer cdiisfaction, employee satisfaction
and the success of the service offerings (Meffet and Bruhn 2000, Grins 1997, Oliva and
Kallenbay 2002). The rational for such an adoption of the orgenizational stucure is nomelly
put forth along four lines. Service operations require an high empowenrent of the employees.
Reducing the levd of hieachy heaps companies to empower savice employees (Bowen and
Lawle, 1998). Second, because of the inseparability of savice delivery and savice outcome, the
service quality is mainly determined by the delivery process. A process-oriented approach for the
value cretion ensures high process quality and guarantees supaior sevice quality. Third, we
observed that the most successful finns in transitioning from product manufactures into sevice
13
provides wee those tht rm decentalized savice oagaizdios with pofit-artloss
responsibility. Fourth, these new de-centalized savice ayanizaions hae to nn ad get
coordinated with a different set of metrics. They have to mm like a nomel savice organization
using mettics needed by a savice oganizaion to messue astoner sdisfaction, employee
sdiisfaction and business suocess. We call these kind of metrics multidimensional, because they
consist of different kind of variables compared to meirics used in manufacturing companies. The
resulting management profile and the oonfiguation is illustaed in the following figue The
meaning of configuration is similar to the activity- oriented dimensions.
oriented Low hierarchy
i Multi-
Decent} 7 BI) ino
: : metrics
Consistent configuration leading to a sustainable
and successful transition from product
manufacturers to service providers
= == == Example foran inconsistent configuration leading to
an unsuccessful transition from product
manufacturers to service providers
Figure 3: Management profile for the adoption of the sudua dimensions (agenizaionl
structure).
IV.3 Behavioral dimensions for the transition
IV.3.1 Basic concept for adopting the outlined structural and activity-oriented
dimensions
To transition a pure product manufacturer into a service provider, one hes to change the outlined
stuctural and activityoiented dimensions in a oonsistent way. Most companies initide
improvement poganms to adopt these dimensions But this necessay adoption camot be
14
installed eesily. A successful adoption needs a sustainable and ocontinuos impovenent pogam
and is more like a process bat must be grown aranically. To do so, the poduct manufactures
ust grapple with several central issues of improvement programs during the transition.
We assume that the service management ties to initiate an improvement program, leading to the
adoptions of the structural and attivity-oriented dimensions Resources ae necessaly for the
adoption or improvements. In th next section, adoption and improvements ae synonyms. Savice
workes and savice manages have limited time, which must be dlocated ang the dily
business and impovereat ativities The inpoverent atiivities ae linked to the savice
delivery, because they interupt the savice production’ Furthemore to acidress the issue of
insepaability of savice ddivey and quality (Oliva 2001), service quality has been defined as a
function of the allocated time per order - a proxy for the degree of attention and care that service
workers ae providing. According to the Mill’s equation of savice quality to service productivity
(Mills, 1986), we assume that increased working effort leads to less time per order decreasing the
service quality.
By combining the issue of insepaability with interaction avising from finv's finite resources, one
gas a short and a long-team effect of adoptions or impoverent pogans For exanple
increesing resources for the improverents aus off resources for the daily business leading to
more working effort and less time per order ending savice quality and reducing the suocess and
swaindbiity of the tansition in short-tam If the improvements overcome the short tam effect,
in logteam am inceesing savice qelity and a improved savice imovaion process boost
sdes, savice ordas and market potential leading to more resources nesded to fulfil the daily
business, That makes fewer resources available for improvement programs in longtam Both
effects can Sow down improvement programs and do not lead to a suocessfl and sustainable
transition process (figure 4).
15
oe -\
ft Market -P Sales
a7” " see + |
? | \
va i *y,
ji ! <a
i ! “ss Resourves
! neededforthe —~
Market \ é z N
Dota \ daily business \
A \o+ + Timpr 4g - Working \
\ _ a sOOrder Effort By
\ “Ss Sewice :
\ Quality LL, + + \
i .
. Resources F v
. available for Resourves available
ee impovenens sor thedaily
Long term effectof service improvements
— Short term effect of service improvements
Figure 4: Short and Long tem effects.
Refening to the latest research in improvement programs, improvement programs can focus on
first or second order improvement. According to the savice or qaity movement (Dening
1986, Heskett & al, 19%), the first oder inproverent is fighting the sympioms. Second order
impoverens ae fighting stucura poblers leading to les pollens or sympions incessing
effidency and productivity. It is more like working smarter By concentaiing the impoverent
effort on the second order improvement, one can inaease effidency and productivity leading to
more resoutves available for improvement.
Adequate resources and concentaling on seoond oer impoverents lead to sulairable
improverent programs, strengthening the transition process. But what deenmines however the
described feedback loop operate? The answer is influence in large measure by the mental modes
of manages and employees The following section shows the impact of mangeia add
enmloyee's mental modds.
16
IV.3.2 Behavioral challenges and cognitive processes
1V.3.2.1 Managerial options
Aooording to the “tdedlogical” theory (van de Ven and Pool, 1995), the management sds gods
for an adequeie leva of ingroverent pogams / adoption of one dimension Combining that
with the aspiration and valance-expectancy theory (Vroom, 1964; Lant, 1992), manages assess
the adequacy of the current leva by comparing it to the desired level. In case of a gap, managers
must try to increese resounoes for the improvement program by using one or more options.
Tf ove looks at the managerial option to increase resources available for improverents, one gas
three different options. Each option initiates a balancing feedback loop (figwe 5). Fist to
existing resources or short cutting the resoues for the daily business Unforumaidy, effort
squeezing would initiate the fomer mentioned short-tam difect oveacoming Bl. Second,
Manages can extend the capacity by hiing more workes or puchesing atiitiod etera
capecity (B2). Expanding capecity involves substantial time ddays (taining building
capabilities and ...) and has just a longterm effect.
Desired Level of
the Adoption / Improvement
Figure 5: Managerial options for dosing the gap between the desired and current level of service
improvements (Based on Repenning and Stenman (2000)).
Third, the manages can free employees to participate in the improvement program That would
lead to the same side effect of fewer resources available for the daily business like working
17
harder (B1). The short tam effect would overcome the belancing effect of B3. This looks like
that it is quite hard to sustain for any improvement, because of the lack of resourves.
But nevetheless - according to ow fonmer cesciiption, one can recommend the following
resoutve strategy for the adoption of the different dimension. First, one has to free employees for
the impoverent activities. Simultaneously, new savice employees have to be acquired, because
service enployess need time to build the right capebilities Heang employess les to the
described effect of ending savice quality. To ovecome this effect, the anployess have to
concentrated on second order improvement, leading to an higher producivity. In mid-term these
impovements ovacome the qudity eosion This phenomenon is usielly called “worse before
better” effect (Keating and Oliva, 2000). In longtam the new savice enployess initiate B2 and
canpensate the long-term effect.
1V.3,2.2 Managerial bias against improvements and managerial service awareness
This is not the first atide to identify the tadeoff betwen impoveren ativities and cily
business and to explain the resource botileneck. They appear in several studies and have been the
subject of rational actor modes (Repaming and Steman 2002). The interesting question for
theory is why do meny product manufactures do not push such impoverent pogams The
answer is dermined in lage measure by mental modes of managers about the tight way to
allocate resourves.
Ye, thee ae several reasons, rooted mainly in cognitive processes, why improvement programs
do not sustain In choosing whether to pursue improverent or daily business, savice manages
must make a judgement about the causes of insuffident improvement programs or the lack of
adoption to the requirements of a service provider. If a gap between the current and desired leva
of improvement programs is thought to result fiom lack of worker effort or disdpling thn
manages will increase the production pressure (B1). If service managers believe the cause lies in
a lack of the adoption, they will focus their efforts on changing the cesaibed stuchal and
activity oriented dimensions. This initiates the balancing feedback loops (B2 and B3) tha ae
ae to dose the gap between the current and desired level of the adoption.
Unfortunately, atributing the gap to inaequate worker efor is consistent with Rosy
“fundamental attribution eor’ (Ross 1977). Ross reviewed that people use to attibute
uncksirable outcomes to people rather than to system stucue The “fundamental attibution
enor” gives the theoretical foundation of our findings that manages tend to aitibute a gap
between the cument and desired levd of the adoption to inadequate working effort and not to
stucturl or activityoriented poblers. While literate (Repaming 2002) and our fidd daa
18
support that linkages, two questions have to be answered. First, would not manager recognize the
existence of the feedback loops and take actions to avoid it? Second, if manages find themselves
stuck in the atiibution exor, wouldn't they leam to escape fiom it? Unfortmddy, seveal
theoretical studies of human decision making and our action reseath recommend that such
Jeaming is very farfrom reality and does not go automatically (Démer 1980, Domer 1983).
According to Senge (Senge 1995), by emcting the diffeent dimension, mamages reed a
reflection about how the adoption is working as well as an awareness that it is necessary. The
reflection represents the understanding of the illustrated feedback structures.
Similar to Oliva and Kdlenbay (2002), we observed that it is quite difficult for manager who
sal a multimillion Ewo mechine piece of equipment to get excited about a maintenance contract
worth€10,000. To initiate the right sevice awareness, one has to overcome the successive hurdle
in making manager aware about the value of services.
Tf manager think about savices as “value added’ activities leading to sstainable compditive
advantages and not as “non value aided” activities they will be highly motivded to push the
necessary adoptions The “value added’ savice awareness iefas to efforts to ponvte
impovemant efort or mandale paticipaion Promoting savice impovenenis and mancking
Pelticipation means that manager initiate B2 and B3 instead of B1. It can be undastood as the
Tmanagedidl commitment or management push.
1V.3.2.3 Employee pull effect
Freeing employees and adding new savice capacities (B2 and B3) can be undessiood as the
Tmanagement push or nomelive pessue to initide impoverent pogams ad to pone
changes in the explained stuchral and attivity-onented dimensions The managerent posh
often creates temporary excitement, but must be replaced by other sources of motivation like
enployee pull effect.
The employee pull effect leads to a changing savice awareness at the employee level. It boosts
the “value added” thinking add hep to oveome typficd alta lebts of pod
manufactures. Usually, savices ae thought of as ak-ons, and initid savioss (insalaion,
commissioning and so on, dc) ae often “given for free’ during the negotiations to sal products
in manufacturing firms. But whenever you give savices for free to customas, the astomer will
never understand the value of services. Genaating a “value ackded” thinking on the employee
Jevd, enforces the cultura transfonmation The product manufactres must leam how to value
their savices, how to sal the savice offeings how to convince astomas, how to deliver
superior services and how to hill them.
19
Employee pull and “value added” savice awareness aise when savice workes undestand the
benefits of savice impovenents and commit thamsdves to improvement efforts (figue 5. In
that cease, improvement effort is independent of management attitudes. The employee pull has the
same property like a reinvestrent cyde. It can function as a virtuous or vicious cyde As
illustrated in figure 6, we obsaved that a variely of factors could interfere with the employee pill
feedback raising the odds of a vicious cyde. In the next sections, we will explain several of these
factors.
Employee perception of benefit
Tn making the judgement that savice impoverent works, workes compae the inpoverent
results they observe to their expectations (Cyet and March, 1992). The expectation is influenced
by the objectives set by manages. The “value added” savice awareness raises if progress is high
dativdy to aspirations and falls whenever progress is disappointing. The described links ceate
a reinforcing feedback loop (R1). The loop, illustrated in figure 6 can opaaie as a vicious or
virtuous cyde The oljective manages se to the workess determine its behavior (Keating & al
1999). When objectives are set high, the expeditions can outstip the obsaved impovement
results and the savice awareness falls, Sowing down the employee pill feedback It eds ina
vicious cyde of goa q@usion and cyicsm While aggressive objectives can weaken the
enployee pull, adequate objectives can boost it When objectives ae se in the right way, the
impoverent results outpeform expectations and savice awareness increases, boosting the
enployee pull effect through R2.
Employees
Aspiration level
Adoptions /Adoptions
R2
\ Valued acted’ St)
Service
Awareness
ae Pereived
feasibility of
objectives
Figure 6. Employee pull effect boosting ,,value added” service awareness.
Adequacy of management support
Furthemore, the effectiveness of any improvement or adoption activity depends on the adequacy
of the chosn methodology or tool (R3). By foaxssng on more oonplex or soon oer
improvements, the acbquacy of methodologies and tools decreases drametically. We assume that
less adequacy of tools and methodology requires a higher management support That means, if
the impovement effort is focused on higher impoverent helf-lives the management hes to
support the use of the required tools much more It changes management's mole At first the
management hes to concentae on motivation and initiating the employee pill. Afterwacs, the
management hes to support the service workers by using the existing methodologies and tools In
cese of a lack of support, the employee pill is going to weeken lecding to @ impoverett
failure. Adequate support can boost the employ pull effect through B4.
21
Support
4
+ results +
+ Employee Perception Effort Allocated
Manegement Sy Motivation —> of Improvements / RI tolmprovements
Adoptions /Adoptions
NO / Nalued added’ S
Service +
Figure 7: Changing management's role.
Furthamore, the adequacy of management support has another important impact We assume
the management les to support multiple impovemat pogars (impovereat pogars for
arganizational adoptions, for new service innovation processes and so on). They mst povide
scarce resources that ae ciitical to the success of the impovenett pograms. Similar to the
enployees, the managerent must also form bdiefs as to the efficay of the impoveratt
program and judge the worth of expending their support in each improvement program.
We assume, that management's decision conocaming wher to dllocate their support The decision
is also influenced by the explained employee pull effect The feedback structure is illustated in
the next figure. The decision point is in the cater of the causal loop diagram, the fraction of
Tanagement support to improvement program one (in the figue ae only two dimersons
illustrated). We observed that the management makes this decision based on their bdiefs about
the relative efficacy of acbpiing or impoving dimension ome If the maregerert bdieve thet
dimension one works particulaly wel, they will increese their fraction of support to that awa As
management allocate more resourves to area one, all dse being equal, savice employees respond
by inceesing ther efot This aditioa efot incesss th rests ad rénfos
management's beliefs about the efficacy of dimension one. These links create a positive feedback
loop.
Management Commitment ‘Management Commitment
[ to Dimension one + 5 to Dimension two ‘\
+ +
Results in Dimension one Results in Dimension two
Fraction of Management
+ Support to Dimension one
ortwo ¥
Employees’ Effort Employees’ Effort
Dimension one Dimensiontwo
+
\ Management Support to + - ‘Management Support to A
Dimension one rs
AC yy
Management Resources
Figure 8: Management Resources.
In cese of figure 8, the two loops ae potentially coupled. If management resources are sufficient
to support both improvement programs in an adequate way, then both loops work in a virtuous
direction. If resource are scare, a decision to dlocae resource to improvement program are wil
be also a decision to allocate resources away from improvement program two (and vice versa).
The described link has a strong non-linearity. It shows that resouves play an important mle in
determining the dynamics of multiple improvement prograns.
The two links illustrated in figure 6 and the adequacy of management support in figure 7 cape
the influence of direct experience on the individual leva.
Diffusion
Another feedback occurs from observing other employee's behavior and consequences of it If
sevice awareness is a group level construct, then there must be an ackditional linkage between the
results and the “value added” service awareness. In the next figure, the influence of results on the
service awareness of employees who do not try to adopt the dimensions is captured by adding an
additional link between results and savice awareness with the intervening variable - obsavation
of service awareness-results linkage by othas The link captues the diffusion of savice
awareness throughout the group.
Observation of
Effort Results
4 Linkage by
Others
RB +
Desired Level of
Improvements / Improvement Ee
Adoptions a results
+ \ + Employee Pereption Effort Allocated]
Management +> = of Impovenetts / RI to Improvements
Adoptions /Adoptions
NO »Nalued added! 7
Service ie
Awareness
Figure 9: Diffusion process
IV.3.2.4 Employee's bias against second order improvement
The importance of the second order improvement and its impact on the success of the transition
‘wes already described in an eatlier section Even most enthusiastic manager can not monitor
evayone’s impoverent adivities in a lage ompanization The managerent sds gods for the
improvement and adoption, but does not monitor how they ge done Thee is no commandant
contol stucure after initiaing the adopiions and impovenet pogams Conmendand -
control structures are dependent on managerial supervision They ae unlikely to work in settings
whee employee's participation is difficult to monitor If the employee pill effect substitutes
initia simiation by management push, the commend-antoontol structure will be changed to a
jremise control (Peow 1986).
In cese of premise contol, mamages do not monitor th impoverett ativitis The
Tmanagement supports the impoverents, but does not necessaily supavise the kind of
improverents activities within the adoption of one dimension The employees decide themsdves
to enhance first or scoond oder impoveratt Similar to maregeren’s judgematt about
resource allocation, four basic cognitive processes limit the second order impoveras. The
besic cognitive processes refer to the “low hanging fruit” syndrome and overweighting of salient
and tangible features of the environment.
People have repeatedly shown to overweight salient and tangible feabres of the environment
First ordas improvements ae simply more sdlient and tangible than second onler impoverents.
24.
(Katmeman, Slovic, Tversky, 1982). For example an inadequate sevice concept is adapted to
customer needs by redefining it and is visible to al, wile the undalying savice imovdion
process is much harder to observe and diagnose.
Second, first order improvements and second order impoverents ae likey to work a different
speeds. Improving the savice imovation process and the extemal commumication takes time
identify the cment process, indicaie rot cases define asomer oontact points, inperat
sduions, train participants and so on The ddays between the stat of a second oer
impoverent progam for the savice imovaion process or a new extemal commuicaion
strategy ranges from months to years. First order improvements are easier to identify and quickly
solved. Therefore, employees under pressure to close a gap between the desired and coment levd
quidity ae likey to foous on first oder improvements even if they know that doing so does not
lead to fundamenta impoverents. At thid first order impoverents hae a moe catan
outcome than second order improvements. Convincing custome is easy and it is usally dear
that the customer buys the service In oatrest, changing etamal communication strategy is more
complex and its impact is more ambiguous. It is not clear whether and how a intended change in
extemal communication will in fact lead to more savice revente. Risk aversion is a basic feature
of humen dedsion making (Einhom and Hogath 1985). Savice employes pefer oatan
outcomes of first order improvements to the uncatain and ddayed outcome of a second order
impoverent
Fourth, concentaing improvement efforts on scoond onler impoverents, while it pees
fubre poblers, does not dininae existing poblers Existing savices with a lack of suitability
to customer needs represent substantial investments in labor and capital. According to the well
known sunk costs fallacy, decision makes often contime a pioject beyond the economicd
tational point when they already invest substantial resources. Thus, employees will foous on first
vather than second order improvements. For example, we often obsaved that savice workes
concentrate on the ackplation of existing sevice concepis to customer needs, because they
represent high investments. Unfortunately, the lack of adaptation of existing savice concepts
stems from ineffective savice imnoveiion processes, but companies do not concentrate on the
devdopment process. They are just fighting the symptoms and do not omcentaie on struchmal
poblens. Conduding the last findings diffeencss in infomation availdility, sdiexwe ad
ddiays hias employees against second order improvements.
To ovecome the bias aginst scoond onder impoverestt ore bes to initide a diffeett
understanding about impoverents. Savice enployess must undestand the effectiveness of
seoond improvements. They have to change thar improvement from first oder impoverenis to
2
second amder improvements enhancing fundamental changes leading to beter impovereat
results pushing the adoption of the different dimensions.
Improvement Fundamental
a eS changes
Employee Perception Effort Allocated
of Improvements / to Improvements
Adoptions /Adoptions
re ———, te) |:
+
F Second Order
Premise controll —_* I ‘i it
Figure 10: Second order improvements.
IV.3.3 Configuration of the behavioral dimensions
As pointed out in the last sections, tansitioning from poduct mamfacures into savice
gwovides has a strong behavioral dimension The challenges discussed result from an enor of
attribution made by managers in assessing the causes for a gap betwen the cment and the
desired leva of savice inproverents, from a changing managed ile fiom enployee’s bias in
choosing the kind of improvement activities and from employees’ paception of benefit All four
factors influence the suocess of the transition from product manufacturers into sevice provides.
They have to configured in such a way that where is a codlescence among them (figure 9). If
companies want to transition themsdves from product manufactures to savice povidas, dl
four behavioral dimensions have to change in a oonsistent way. The ooalescence is meant in a
way that the oonfiguation is consistent omperaiay ad nua renforing If companies
‘want to move from product manufacures to savice provides, the tight configuation of the
sucoess factors in tems of a coalescence will be also on the outer cide of the following figure.
Tn that case, the described feedback loops opaaie in the right direction initiating a sustainable
transition process. If one or more behavioral dimensions do not change, this feedback loop will
wok in a vicious way oveoming the positive effects of the other feedback structures leading to
26
an unsuocessful transition In case of an inconsistent configuration, the improvements would not
sustain and would leading to an unsuccessful transition.
Support Valueadded
Second order | | Value added
improvement
=
Enployee Level
= Consistent configuration leading to a sustainable and
suocessful transition from product manufactures to
service providers
==" Inconsistent configuration leading to an
unsuccessful transition (service jungle) from
product manufactures to service providers
Figure 11; Management profile for the adoption of the behavioral dimensions of the transition.
V. Formal Model
V.1 Specifications
Our integrative framework is textual and diagrammaiic. But it is not completly able to desaiibe
the transition as a dynamic process. For that reason, we translated our integrative framework into
a foo mod. Of couse ow mathematica mode is mor pecss bt moe resticive
embodiment of a verbal theory. The translation necessarily results in the loss of richness. There
is, however, one conesponding benefit The simulation enforces the intemal consistency of our
integrative framework, thus ensuing that behavior it purports to explain can in fact be generated
by its undelying assumptions. Our integrative framework suggests that the success of the
transition mainly depends on the degee of integaio arog th ilistaed neegeret
27
aofiles. The use the mathemetical modal to answer the following question - Does the degree of
integration has an impact on the transition?
Tn this section, our causal loop modd is tansaied into a metheneticd modd. To introduce the
modd fomulation, we stat with the link between impoverent effort and results. To devdop a
imathenstical representation, we are according to Schneidenman’s telf-live moda (Schneiderman
1988). He found out that any defect levd, subjected to legitimate improvements deceeses at a
constant (fractional) rate. We assume that any dimension can be improved in the same way. That
means every dimension can be improved through a constant (fractional rate). The improvement is
cdlailated through the following equation.
@ A) =A, enl-9,0.
A represents the degree of adoption for dimension n. A, is initial levd of the adoption of
dimension n. In case of a product manufacturer the initial leva of A.o is 1. Moving towads a
service provider means that A,(t) is convaging to zao. If A,(t) is aound zao, dimension n
will be compledy adopted to the requirements of savice provides. The degree of adoption can
be interpreted in the following way. It can be thought of as a leva of specific capabilities needed
to adopt dimension n The parameter ¢ meesues the intinsc and extinsic diffialty of
impoving a paticlar dimension Similar to Senvan’'s & a (1997) infeeation, inbinsic ad
extinsic difficilty lead to different inqrovement half-lives The mathematical link between the
improverent rae and the improvement half-live can be formulated in the following way.
Q t=W(2/o .
The half-life of one dimension represents the time required for one dimension to adopt by fifty
percent The improvement half-lives vay acioss the different activity-oiented and siuctral
dimensions. Greater inbinsic and exhinsic diffiailty dows down the impoverent rae due to
difficulties designing, conduding and implementing impoverent attivities We found out thet
impoverat half-lives gow with the intinsic and extinsic difficulty. Inbinsic diffiailty iefas
to the mmber ad type of people, from different organizational functions, required to cany out
the adoption of the dimensions or an effective improvement. Extinsic difficulty means that the
customer can have an impect on the savice impoverent Greater extinsic difficulty dows
down the improverent rate due to problems in eqlaining and convincing the astomer to make
us of the savice impoverens. The following figue illustates the intinsc ad extinsic
difficulty of each structural and activity- oriented dimension.
Increasing Half- Lives
Intrinsic «=A >
difficulty
A
: * Service offerings
High} * Configurationof value | . savice innovation
Creation + Communication
Increasing
Half-Lives
* Competitive Strategy
Low] * Hierarchy
* Coordination
difficulty
Figure 12: Dependence of improvement half-lives on intinsic and extiinsic diffiailty.
Translating equation (1) to a first-order differential equation, the time path of adoption can be
described through the following equation.
@ dA,/dt=-9,A,
Schneiderman’s modd is a typical specification of improvement in the opaiions liteabre It is
also used in latest studies of injrovement programs (Staman and Repenning 2002). It is a useful
stating point but has to be modified in two ways.
At first, because sustainable changes of the stuctmal and activity-oriented dimensions require a
dear strategy, the dimensions also decay to the initial levelA,) at the fractional rate 6 . The
fractional decay rate 6 is likely to be low for structural changes and higher for activity oriented
dimensions. The impact of 6 is influenced by the compditive stategy A. (t). If A, (t) is equal to
1 (groductoriented compditive strategy), the adoption of dimension n will decay to the initid
levd A, at the fractional rate 6 If the A,(t)is equal to zen (saviceorented staegy), the
adoption of dimension n will not decay to theinitial levd anymore.
Second, the helf-live model does not integrate the behavioral dynamics of the transition or
adoption. It treats it as an autonomous process that is solely a function of time. The following
modificetion relaxes these restrictions.
@ ha AB AAA.
29
The cment behavior, B, is defined as the fraction of the workforce curently showing a high
Jevd of service awareness and role understanding (““value-added” sarvice awareness and role
‘understanding in terms of business manager and second order improvement). Equation (4) links
behavior and results to each other The levd of adoption is now an explicit function of the
current commitment to the service improvement As mentioned in the former chapter, the right
behavior stems froma quite complex interaction between intinsic and extrinsic motivation The
mathematical representation for the evolution of behavior captures both sources. We rely on
established mathematical representation used in latest studies of improvement programs
(Staman &al 1997).
6 Sh .5 -B ywpa-p).
dt Ty
V.1.1 Managerial level
The first tam on the left-hand side of equation (5) represents managerent’s effort to caeate
necessaty behavior B’ is management's goal for the behavior in tams of savice awarness
and role undastanding = It depends on how much savice awareness and mle undastanding
Tmanagement devotes to promote savice impoverent programs Managenent’s god, B’ , is
endogenous. It is influenced by the company’s suocess. Company’s suocess is a construct defined
over the za to 1 interval, that measures management's willingness to promote the adoptions and
impoverents of each dimension The effect of companies success on management's god is
operationalized as a decreesing function with a second dejivetive thet is initidly positive ad
becomes negaiive at aggoximately the mid-point Small leveds of company’s success have litle
effect on management's goal, but as company’s success decreases, management becomes
inceesngly unwilling to pomote ad psh th tasitio This pheomenn wes idatified
through interviews with seveel manages duming our cese studies The indicated levd of
company’s suocess is a function of two measurements: market share and contribution of services.
The parameter 1/7=, oontols the speed at which employees’ behavior adjusts to management's
gods. That means, employees’ behavior approaches oataris paribus management's tayet with an
average dday of 1. According to Repanning (Repeming 2002), this delay aggregates three
important components. First senior manager necd time to devdop ad inperet ations
targeted at creating nonmetive pressure. Second, time is dso required for participants to react to
the management push Third, participants need time to acquire the appopide dlls ad to
30
modify the behavior These dements capuure both the tdeological actions of manages and the
different delays inherent in participants’ reaction.
V.1.2 Employee Pull
The first tem on the left-hand side aggregaies the enployee pull effect It represents the “pl”
effect generated by successful results and adequate management support The more people ae
involved in the in savice impoverents, the moe they wil commuicae thar enthsiam to
othas though wortofmouth w represents the wordofmouth It comeds the diffusion with
and the reinforcement process. w is not fixed parameter. It is a variable that represents the current
bdlief among those who actively try to improve the dimensions concaming their efficiency.
The wordof-mouth can have a positive or negative impact on the behavior We modd the sgn
and strength of the wordof-mouth effect as depending on the paveived savice impoverenis
and the adequacy of management support. Thus,
© w,=0,(f,fr}+f, fa),
where f,{} is the impect of results, r, on the wordof:mouthand f,{*}is the impact of abquacy
of management suppor a (Sten @ a 1997). w is linked to the paceved savice
improvements through the following three equations:
dA 1 _ GAB, +O4A,
a 2 or.
» GA, 1 _ OA
8 a AL A, =P,
P
(2) RS
Fost, the impoverent rate is first divided by the ament levd of impovemais to cdailate the
fractional rate of impovement The cdallation of the impovement rae on a fractional besis
ymovides a convenient nonrelization and is consistent with the pareption theory (Repaming
2002). The paeption theory suggests that people evaluate rates of changes on a proportional
leva rather than absolute besis (Plous 1993). The impovement rap P,, is then evaluated
rddive to menegerent’s god. In the smuilaion modd, P,” is predicted by the theortica
impoverent halflife It does not indude the behavior and the decay towards the initial level. P,
and P,, both ae used to cdalde the impovermnt rae as a fiaction of the objective This
fonmuldion is consistent with the aspiration concept of Cyeat and March (1992), whaeby
31
perfoomance is evaluated dative to an explicit goal or aspirdion The formulation is also used in
the latest studies of improvement programs (Repaming 2002; Staman, Repaming and Kofman
1997). Finally, p, is linked to wviathefunction f, {}.
w is linked to the akeqgecy of support through the following equetios Agin using the
aspiration formulation (Cyat and March 1992), a,, is cdlailded by dividing the avout of
resources currently allocated to savice impovement n, r,
required by that area,
(10) a, =1,(n)
is the amount of resouve amenily
(11) 4 =L,p,C,
The total resource requirements in area n, ris simply the product of the number of people in the
ara, L,, the improvement resource requirement per parson assuring full patipation 3,, and
the cunent believe in savice improvement. Finely, p,and a, are linked to w, via the functions
£.{} and £,{} and the parameter w. w is a constant that represents the frequency with which
uses interact The function f,{} captures how people upceie their behavior to the savice
improverent based on the results they obsave. Similar, the function f,{} captures how service
workers update their behavior based on the support they get We just want to discuss their
qualitative meanings A more fonrel derivation is not in the soope of this paper For a fomal
dativation, we acco to Repanning (2002).
Thee ae three major quadiitative assumptions embodied in the rdaionship, Fist all theories
suggest f,{} and f,{} Sopes upward - more results and more sipport lead to an increased
service awareness and better role understanding. Second, the lefthand limit is negaive implying
that the wordof-mouth is negative when service improvements show no results or when there is
a lack of management support. These act as a drain on employee's behavior. A negetive left-hand
f.{} implies that the actions of management are not sufficient to ceale 100 paroent behavior on
the enployee levd in absence of results A negative lefthand for f,{:} implies that results ae
not sufficient to generate 100 paroant behavior in absence of management support. Third, the s-
shape represents the assumption that near the left and right limits, smell changes in sesults and
adequacy of management support do not necessaily have an impact on the behavior That
means, if thee ae no results and no support, sell changes do mot charge the beliefs. Similaly,
if thee ae excdlent results and adequate support, smell changes have dlso little impact When
32
results and adequacy of support ae somewhere in middle relative to its goals, then changes have
a bigger impact on behavior.
The resource allocation decision of the service management is represented by the parameter x,,
the fraction of managements’ support devoted to area n. The amount of resources dlocaied to
dimension n, r,, is equal to the product of the total mumber of resources available R and the
allocation fraction We obsaved in our intensive case studies that the managerent uses two
pieces of infonveation in fonving their beliefs about the relative expectancies of success in the
compding adoptions of diffeent dimensions - the amet impoverent rate by ech savice
improvement and the cunent resouve requirement, q,. The raionde for using poduiivity
improverent is obvious management wanis to spend ther time in savice impoverans in
which they think the changes they introduce will be successful. Using resource requirements as
an input for management's decision nile capumes the obsaved phenomenon thet savice
manages look to sevice workess how wel things ae going. Resource requests represent a
simply proxy for the savice impoverent’s assessment of its own effort q, is calalated
‘through the following equation.
(122) 41, =x,R
N
a3 aq =n fl
. 4
(4) x, =Rral fe Peg
= o
To specify the decision nile we use an US(US+THEM) fomnidiion (Kalish and Lilien 1986).
This fomuldion is widely used in the ldet sudies of impoverett pogams The
“atractiveness’ of each dimension is detmmined by weighting both the fractional resouve
requirement and the rate of improvement by exponents, denoted by & and B. The resouve
allocation to savice impoverent n, x,, is the detemined by equation (14). The equation
calculates the attractiveness of dimension n as a fraction of the total attractiveness. If a =0 and
B =. then resources ae dlocated striclly according to need If a >0, then the dlocation will be
biased towards the dimension showing more rapid improvements. The oonvase is te if a <0.
Consistent with our reinforcement agument above as wel as with the commonly voiced
“successful change begins with results’ policy, a and B, ae chosen to represent a policy
allocating more support to dimensions showing beer results (a ,B >0). The commonly voiced
4
33
“suocessful change begins with results” policy is often found in the practitioner literabre (Kotter
1995, Schaffer and Thompson 1992).
V.2.3 Market and competitor
We assume that our simulated company faces a single aggregate compditor The compditors
supplies products and offers savices that compete direcly with the smulded company. The
market share depends on customer assessments of product attractiveness and perceived service
quality compared to the compditor The paceved savice quality is th maximum led of
service quality and commumicaion The savice qudity is deemined by company’s stucura
adoptions. Similar to perceived sarvice quality, service quality is defined as the maximum among
the adoptions of the four structural dimensions. The market shares is determined using standard
attractiveness or US/US+THEM fomulaiion This formulation implies that the tota market is
always split betwen the simulaed company add its compditos. The dtctiversss is
detemmined by summing product attractiveness and service quality.
The size of the potential market is detemined by mumber of savice offeings The mmber of
service offerings is increased by new savice introductions. The rae of new savice intoducions
depends on the adoption of the service innovation process.
V2 Analysis
With the specification complde, the fonmal mode can be used to undastand to dynamics of the
transition process. We are able to show the implications of the integrative framework oullined
above. The transition is started by introduce a step increase in B’. All the other parameter are in
the technical appendix. Figures 13 and 14 show the evolution of the degree of integration and
contribution of savices The degree of integation is opadionaized through the standad
vatiation among the outlined stuchral, activityoriented and behavioral dimensions A standard
variation of zero means the integration is on high levd. At the beginning, the degree integration
is quite high, because each dimension is equal to 1. Initially the degree of integration decreases
sharply and leads to a cecreesing oontibutio of savioss After passing the minimm the
increesing degree of integration causes an increasing contribution of savices. There is a stong
configuration between the diffaent dimensions g&4s more add moe oonsstet ad mid
reinforcing leading to a successful transition.
Degree of Integration
High \
x
\
Tow 5 50 100 150
Time
Contribution of Services
06
0.3
0.0 Nw
° 50 100 150
Time
Figure 13 and 14: Degree of integration and contribution of services.”
This bese case shows one possible behavior mode generaled by our integrative framework and
formal modal. The base case constitutes only one simulation nn with a specific parameter set.
To show that the degree of integration among organizational structure, company’s activities and
behavior plays a key role in determining the suocess of the transition, one has to nn a s&t of
simuldions.
Figure 15 shows a s& of simulation results in which the degre of integation among siucure,
activities and behavior is different This experiment can be intepeted as varying the degre of
integration by using different parameters for the simulation The results ae presented in the fom
of three dimensional response surfaces. The vatical axis represents the outcome vatiable of
interest (contribution of savices). The horizontal axis represents time and the third axis that
extends “into” the page captures the input variable bang directly menipilded through different
35
degrees of integration Reading fiom front to back dong the input vaiiddle axis, ay given line
shows how the value of the output variable - onntibution of savices, a one specific time
changes in response to changes in the input variable Viewing the resulting suface presents a
dynamic view of how the evolution of the contribution of savioes is influenced by changes in the
input variable (degree of integration). Figure 1 shows that the suooess of the transition depends
aitically on the degee of intgaio among sich, ativity-cieted ad bewiod
adoptions. An higher degree of integration reduces the “worse before better” effect and increases
the contibution of savices. An high degree of integration helps to sustain the transition and
leads to a successful transition process from product manufactures into savice provides. Our
analysis provides some insights into the dynamics of the tansition Fist it yidds a more precise
statement of the nature of the interaction among organizational stucure, activities and behavior
Second, the amlysis suggests that a pincipa fale moce in tensiioning from pode
manufactures into savice provides, one that is not mentioned eqlicily in existing theaics, is
the inability to engender an high degree of integration If the dynamics described here are central
to undastanding the transition, then the question of “why do most transitions fail?’ can be re
stated as “why do most transition processes fail to achieve an high degree of integration among
organizational structure, activities and behavior?”.
Contribution of
Services
0.8
-0.2 |
-0.4
High
Time 120 135 Low Degree of
150 Integration
Figure 15: Response to changes in the degree of integration.
36
VI. Implications for researcher and practitioners
Our sample of case studies was small, but covering several industries. The case studies cover the
whole tansition line as wel as the “service jungle’. We bdieve that our findings provide useful
guidance for manages seeking to design a successful transition prooss Our alysis and
condusion have an essential implication for both reseacher and menages. For savice
the transition from product manufachres to savice povides is stongly influenced by the
strucural, activity-oriented and behavioel dimenson A ocompee theny of the tarsitin
requires an inteisciplinary theory that integies savice management and lumen decision
meking. Our analysis suggests that future studies and research has to eqlicitly consider theses
factors which influence the transition. Of course, our ideas presented here offer a complementary
perspective to many of the existing ideas advocated by many practitioners.
Companies can push the tansition by activdy managing the feadback mechanism thet limit the
sucoess. They must carefully plan and rolbout savice impoverents. In shot, manages must
become adept in understanding their organizdiion as a dynamic system and in undestanding the
transition from product manufactures to service providers as a dynamic process.
The integrative framework fiom which our insight aises bes limitations of couse Most
significant, the framework we propose represents an abstraction from the detail of real sevice
and change management We aggregate seveal factors into single structural, activityoriented
and behavioral dimensions The myriad activities required to change the sucha and activity
aviented dimensions ae consideed smply “impoverents’. Nevatheless, while the mod is
exceedingly simple and capumes only a sell portion of the conplexity of any real tansition
yarocess, our findings capture an important set of dynamics that play a aitical role in deemining
the suooess of the transition.
The most intiguing finding however wes to realize that the sucoess of the transition is mainly
influenced by the degree of integration within each management profile as wel as between the
three management profiles A successful transition process ae characterized through the
consistert configurations in each management profile The consistent configuration is always on
the outer circle. In case of inconsistent configurations, the transition process leads to the “service
jungle’. Companies that do not want to move along the tansition line should ty to be on the
imer cide. Overall, our management profiles provide a beer navigational chat for manages
facing the different challenges inherent in the transition parooess.
To help manages, the matrix in figwe 2, 3 ad 6 can be used a8 a cecision support system The
decision support system helps to find the right configuration of the seveal dimensions in tems
37
of a Oodlescence The ooalescence guaranises sustainable improvement programs, leading to
successful and sustainable transition processes. It supports companies to leave the trap of the
“service jungle’ and strengthens company’s oompditive position through extended savice
offerings.
References
Andason, J.C. and Naus, J.A. (1995), ,Captuting the value of supplementary services“, Harvard
Business Review, Vol 73 No 3, pp. 75-83.
Arthur, B. A. (1993), “On designing economics agents that behave like human agents’, Jounal
of Evolutionary Eoonomics, Val 14, No 3, pp 361-383.
Benchmarking Report - Commazidizing incustial Savices (2001), Institute for Technology
Management
Bowe, D. add Lawer E, (1998), “Empowement im Diensidstungsbatich’, Meyes A.
(Bdito): Handbuch Dienstidstungs: Marketing, pp. 1031-104.
Chase, R. (1981), “The customer contact approach to services: Theorefical beses and practical
extensions’, Operations Research, Vol. 24 No 4, pp. 698-706.
Cyett, R. and March, J. (1992), “A behavioral theory of the fim’, MIT Press, Cambridge, MA.
Deming, W.E. (1986), “Out of the crisis’, MIT Press, Cambridge, MA.
Dormer, D. ( 1980), ,,On the difficulties people have in dealing with complexity”, Simulation and
Ganes, Vol 11 No 1, pp. 87-106.
Domes D, Kravig HW, Reith, F, Staudd, T. (1983), Lobhasex vom Unga nit
Unbestinnithat und Kompexitat, Ban.
Enhom, H. J. and Hogarth RM (1985), “Ambiguity and uncatainty in probebilistic inference’,
Psychological Review, Vol 92, pp. 433-461.
Einhom, HJ. and Hogarth, R. M. (1986), Behavioral Decision Theory: Process of Judgement and
Choice, Annual Review of Psychology, Vol 32, pp. 53-88
Eisnhadt KM (1989), “Building theories from case study reseath’, Acakny of
Management Review, Vol 14, No 4, pp. 532-550.
Evarsson, B. (1990), “Service production and savice marketing in manufacturing companies’,
Proceedings of the Ith Intemational Seminar in Savioe Management, IAE Aix-en-Provence,
FNEGE.
Fonrester, J.W. (1961), “Industrial Dynamics’, MIT Press, Cambridge, MA.
Gavin, D. (1995) Leveraging Processes for Strategic Advantage’, Harvard Business Review,
September-Oktober, 1995, S. 77-90.
Giddens, A. (1984), “The Constitution of socidy: Outline of the thay of Stuur’,
Univesity of Califomia Press, Berkeley, CA.
Grnroos, C (1997), ,, Keynote paper From marketing mix to rdationship marketing - towarcb a
paradigm shift in marketing“, Management Decision, Vol. 35 No. 5, pp. 322-339,
39
Heskett, J.L., Jones, T.O, Loveman, G.W., Sasser, W-E. and Schlesinger, L.A. (1994), “Putting
the seivice-profit chain to work”, Harvard Business Review, Vol 72 No 2, pp. 164-174.
Honbug, C, Gime; C. ad Fessecht M (200), “Wem Indstienteneom a
Diestiesem weden - Lenn von dn Besten’”, Insttit fir Marktorentiate
Hube; G. Glids W,, (1993) Organizational Chanhe and Redesign Ideas and Insights for
Improving performance, New Y ork, Oxford Press.
Kahneman, D,, Slovic, P. and Tvessky, A. (1982), “Judgement under uncattainty: heuristics and
biases’, Cambridge University Press, Cambridge MA.
Kalish S. and Lilli G. (1986), “Applications of Imovations Diffusion Modds in Marketing’,
Innovation Diffustion Modeds of New Product Acceptance, Mahajan V. and Wind Y. (eb),
Cambridge, MA: Ballinger
Keating EK, and Oliva R. (2000) “A Dynamic Theory for Sustaining Process Improvement
Teams in Product Devdopment’, Advances in Intedisciplinary Study of Work Teams, Vol. pp
245-281, Bayedein M. M. and Johnson D. A. (eds), Greenwich, Conn: JAI Press.
Keating E, Oliva R, Repaming N, Roda, S ad Same, J. (1999), “Ovavoming the
improverrent paradox”, European Management Joumal, Vol 17 No 2, pp. 120-134.
Kotte J (1995): ,,Leading Change Why Transformation Efforts Fail“, Harvard Business
Review, march anil, pp. 4-67.
Lan, T. (1992), ,Aspiraion levd adaption An empitical explation’, Management Scene,
Vol 38No 5, pp. 623-644.
Math, H and Shapiro, RD. (1993), “Integrating savice staegy in the manufacturing
company’, Chapren & Hall, London.
Mathieu, V. (2001a), “Service strategies within the manufacturing sector benefits, costs and
pertnership“, Intemational Journal of Service Industry Management, Vol. 12 No. 5,
pp. 451-475
Mathieu, V. (2001b), “Product services: froma service supporting the product to a service
supporting the dient’, Journal of Business & Industrial Marketing, Val 16, No 1, pp. 39-58
Meffert, H. and Bruln, M. (2000), Dienstieistngsmarketing, Wiesbaden.
Mills PK. (1996), “Managing service industries’, Ballinge; Publishing Company, Carbridge,
MA.
Oliva R and J. D. Stamen (2001), “Cutting Comas and Working Ovatime: Quality Erosion in
the Savice Industry”, Management Science, Vol 47, No 7, pp 894-914.
40
Oliva R. and Kallenbag R= (2002), “Managing the transition from products to savioes’, in
Steve T. andal., Quality in services: Crossing bounchties, Victoria, B.C., University of Victoria
Penow, C. (1986): “Complex orygnization’, New Y ork.
Plous, S. (1998): “The Psychology of Judgment and Decision Making’, New York,
Repeming N. and Steman J (forthcoming), “Capability taps and sdfconfimming dtibutio
eros in the dynamics of pooss impoverent Forthooming in Adninisiaive Scene
Quatady.
Repaming N= (2002), “A simildion-based apooach to undastanding the dynamics of
imoveaion imfeventation’, Orgenization Science, Vol 13 No 2, pp. 285-300.
Ross, L. (1977), “The intuitive psychologist and his shortcomings Distortions in the attribution
yaoces”, In Berkowitz (Ed), Advances in experimental social psychology, Academic Press, New
Yak, Val 10.
Schaffe; R, Thomson, H (1992), ,,Suocessfull change programs begin with results“, Harvard
Business Review, Vol 70, No 1, pp. 80-89.
Schneiderman, A. (1988), “Setting Quality Goals”, Quality Progress Apail, 55-57.
Senge P. (1996), “The fifth diszipline’, London.
Simon, H. (1957), “Modds of man’, NewYork.
Simon, H. (1998), ,,-Industridle Dienstleistungen”, Schaffer Poeschel, Stuttgart.
Stamen, J, N. Repeming and A. Kofman (1997), “Unanticipated side effects of successful
qality impovemets: Expaing a paadx of oaizdionl impoverst’, Mawgeresnt
Science, Vol 43 No 4, 508-521.
Van de Ve, A. add Pode MS. (1995), “Explaining devdopmett add change ad
crganizations’, Academy of Management Review, Vol 20, No 3, pp. 510-540.
VDMA, (1998), “Dienen und Vedienen’, VDMA Vedag, Frankfurt, Gamany.
Vroom, V. (1964), “Work and Motivation’, New York.
Wise, R. and Baumgatnes B. (1999), Go downsteam The new impadive in manfactuing,
Havard Business Review, Vol. 77 No. 5, pp. 133-141.
‘The most theorists assert that employees doing a job are the best informed expats and should be responsible for
identifying new service ideas or improvement opportunities for enhancing the service quality. This strategy leads to
‘two advantages. On the one hand, employees already understand their processes. That reduces the time for data
cdlection and implementation. On the other hand, employees have a strong interest in implementing improvement or
new services they developed themselves (Deming, 1986).
“The simulation is a shorter version of our formal modal. The complete simulation modd will be introduced, if the
‘paper is accepted.
Back to the Top
CC BY-NC-SA 4.0