Sunteți pe pagina 1din 21

Accounting, Organizations and Society 35 (2010) 184204

Contents lists available at ScienceDirect

Accounting, Organizations and Society


journal homepage: www.elsevier.com/locate/aos

Accounting and strategising: A case study from new product development


Brian Jrgensen a,1, Martin Messner b,*
a
Department of Business Studies, Aarhus School of Business, University of Aarhus, Fuglsang All 4, 8210 Aarhus V, Denmark
b
Department of Accounting and Management Control, HEC School of Management, 1, Rue de la Libration, 78351 Jouy en Josas, France

a r t i c l e i n f o a b s t r a c t

This paper explores the relationship between accounting and strategy in a context that is
characterised by pluralistic demands and high uncertainty about outcomes. By way of an
ethnographic eld study in an R&D intensive company, we analyse new product develop-
ment (NPD) projects and the way in which decisions and practices concerning these pro-
jects are accounted for. Building upon a practice theory perspective, we nd that actors
account for the appropriateness of NPD practices not only or primarily on the basis of
accounting information, but also by strategising, i.e. by mobilising different strategic
objectives to which these practices are supposed to contribute. We argue that this has to
do with the ambiguous demands on NPD and the limits of calculability inherent in NPD
design decisions. At the same time, accounting information is not necessarily irrelevant
in such a case; it can enter the picture as a general understanding that guides actors
strategising efforts by reminding them of the ultimate importance of nancial numbers.
2009 Elsevier Ltd. All rights reserved.

Introduction recent developments in strategy literature, in which an in-


creased concern for the micro-dynamics of strategy-mak-
The way in which accounting information relates to ing can be observed (see Jarzabkowski, Balogun, & Seidl,
strategy is a question that has been repeatedly addressed 2007; Whittington, 2006). Instead of discussing strategy,
in accounting literature (see Langeld-Smith, 1997; Chap- these authors have adopted the notion of strategising as
man, 2005; Henri, 2006). The dominant approach to this is- a shortcut for the various activities through which rather
sue has been to take the contingency route, i.e. to examine abstract strategic ideas or objectives are interpreted and
the t between the type of accounting information and the enacted by organisational members who, in turn, shape
type of strategy (see Chapman, 1997). While contingency and develop these ideas.
studies conceptualise accounting and strategy as static Combining this concern for strategising with a concern
concepts, a case study approach offers the possibility to ad- for the everyday practice of accounting (see Tomkins &
dress the dynamic interactions between accounting and Groves, 1983) seems promising in many respects. A focus
strategy over time (Langeld-Smith, 1997). on strategy-accounting talk (Chua, 2007, p. 492) allows,
It has recently been suggested that a more detailed for example, discussion of how accounting is weaved into
examination of this dynamic relationship between strategic considerations and debates as well as how
accounting and strategy is needed (Chapman, 2005; Chua, accounting concepts, such as prot or cost, are mobi-
2007). Chua (2007, p. 493), for example, calls for a redis- lised when crafting strategy. It also makes it possible to
covery of accounting and strategy as contingent, lived examine to which extent strategising is, in itself, a form
verbs rather than abstract nouns. This call resonates with of accounting (in the broad sense of this term; see Garn-
kel, 1967) that may complement, or even replace, the reli-
ance on accounting representations.
* Corresponding author.
In this paper, we follow this call and seek to bring to the
E-mail address: messner@hec.fr (M. Messner).
1
Centre for Industrial Production, Aalborg University, Fibigerstraede 16, fore some of the complexities and nuances in this relation-
9220 Aalborg, Denmark. ship between accounting and strategising by considering

0361-3682/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2009.04.001
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 185

an empirical setting that is characterised by pluralistic de- of several different strategic imperatives. The sometimes
mands and high uncertainty about outcomes. We examine competing strategic demands added to the complexity of
new product development (NPD) projects and the way in the NPD process and increased the ambiguity of evaluating
which decisions and practices concerning these projects the appropriateness of NPD solutions. This particular set-
are accounted for. NPD arguably constitutes an interesting ting not only motivates an inquiry into the role of account-
setting for the study of accounting and strategising be- ing information; it also creates visibility into the way
cause of the particular characteristics of NPD practices. strategic imperatives were mobilised within the NPD pro-
NPD practices tend to be rather complex in terms of their cess. As such, our case invites a reection upon both
demands and uncertain in terms of their outcomes. Com- accounting and strategising, and upon the ways in which
plexity results from pluralistic requirements within the these two sets of activities relate to each other in a setting
development process, and it increases with the number characterised by pluralistic demands and uncertain
and diversity of organisational members that contribute outcomes.
to this process in one way or another. Uncertainty results While our paper is not the rst case study to address the
from the limited controllability of the outcome-input rela- relationship between accounting and strategy (see, e.g.,
tionship and tends to increase with the timespace dis- Dent, 1991; Roberts, 1990), we contribute to the existing
tance between the actions taken in NPD and the literature by detailing the continuous process of strategis-
consequences of these actions. ing and the role of accounting therein. More specically,
As a consequence of the complex and uncertain setting, we highlight challenges for accounting and strategising
one can expect a relatively high degree of ambiguity in the that may emerge in a pluralistic and uncertain setting such
evaluation of NPD practices. Often, there is not one best as that of NPD, and we explain how our case company
way NPD should be performed, but rather a range of pos- dealt with these challenges. On the one hand, we demon-
sible options that are acceptable to the relevant parties. strate how accounting information can be used to frame
This, in turn, suggests that there may be limits to the use- the strategising process, both as specic rules that top
fulness and applicability of those forms of accounting that management can enact at certain critical points in time
neatly specify what is right and wrong. The alleged preci- and as a general understanding that helps actors from dif-
sion and objectivity of accounting numbers may be of little ferent functional practices to make competing or even con-
value in a setting where a plurality of interpretations is not icting ends meet. Accounting is not the only practice that
only possible but also warranted. Nevertheless, there is shapes the strategising process, but when it is enacted as
likely to be a need for some form of accounting in such a rules and general understanding (see Schatzki, 1996,
case. For, ultimately, decisions need to be made and ac- 2002), it becomes a powerful means that allows managers
tions must be coordinated. Rather than relying only or pri- to move forward despite high degrees of complexity and
marily on accounting representations, organisations may uncertainty. On the other hand, we also demonstrate
resort to other types of accounts which allow goals, deci- how strategising may become a way to complement reli-
sions and actions to be selected and justied (Ahrens, ance on accounting information. The representational lim-
1996; Garnkel, 1967). The enactment of strategic objec- its of accounting in a pluralistic and uncertain context are
tives is a case in point. Strategic objectives may be mobi- less likely to be contested if accounting information can be
lised together with accounting information to make sense combined with a more hands-on approach to control that
of particular design choices or action alternatives and to relies on mobilising strategic arguments. We show how
control the trajectory of the NPD process. How this hap- such processes of strategising are promoted by top man-
pens, and to which extent accounting information is impli- agement but rely upon the local knowledge of managers
cated in such processes of strategising, shall be the focus of and engineers and their efforts to coordinate themselves
this paper. horizontally.
To this end, we apply a practice theory perspective (Ah- By considering the role of accounting in NPD practices,
rens & Chapman, 2007; cf. Schatzki, Knorr Cetina, & von we also, and more specically, contribute to the existing
Savigny, 2001), which is helpful in picturing accounting accounting literature dedicated to this particular subset
practice as forming a nexus with a wider set of NPD prac- of organisational practices. Most of the existing literature
tices. According to such perspective, NPD practices can be in this area has taken the contingency route, i.e. it has stud-
regarded as spaces of intelligibility (Schatzki, 2005, p. ied the t between forms of accounting information and
470) against the background of which other practices (such characteristics of NPD or R&D practices (e.g. Rockness &
as accounting) become meaningful. We investigate how Shields, 1984; Abernethy & Brownell, 1997; Davila,
accounting is practised in the context of new product 2000). While these contingency studies have drawn atten-
development by mobilising ndings from an ethnographic tion to the complexity and uncertainty of NPD projects, as
eld study of an R&D intensive manufacturing company. outlined above, they have studied the use of accounting
This company had implemented a number of change pro- information in such a setting mainly from a distance.2 Using
jects, e.g. the SAP project, the Lean project and the a positivistic methodology (see Ahrens & Chapman, 2006),
Procurement project, in the recent past. At the time of they have focused on what kind of accounting information
our study, another major change initiative was underway:
a switch from integral to modular products. This move to 2
A case study by Nixon (1998) is a notable exception. Nixon argues that
modularity was a strategic reorientation that considerably accounting may be seen as the integrating vernacular of project
affected the companys NPD practices. What is of particular members. The details of how this is accomplished are left relatively
interest to us is that modularity was promoted in the name unexplored, however.
186 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

is used with what effects, while abstracting from the ques- to a large extent about the establishment of order among
tion of how accounting is actually practised in the context activities. The social practices that can be found in organi-
of new product development. Hence, while the ndings of sations are varied production practices, strategy prac-
these studies have moved the debate on accounting for tices, marketing practices, accounting practices, and even
new product development forward, there remains a worry- practices of having lunch, meeting with others, taking a
ing lack of knowledge of what it actually means to practise break, or checking e-mails (see Schatzki, 2005). This is re-
accounting in new product development and how other ected in a growing interest in practice theory approaches
types of accounts may complement accounting representa- within the eld of management and organisation studies
tions in such a setting. With our paper, we seek to partly ll (e.g. Ahrens & Chapman, 2007; Jarzabkowski, 2004; Orli-
this gap. kowski, 2007; Schatzki, 2005; Whittington, 2006). Practice
To develop our argumentation, we rst lay out the the- theory has been particularly well received in recent strat-
oretical concepts that we mobilise in our case discussion. egy research, in which the notion of strategising has be-
Thereafter, we explain the research design and methodol- come the emblem for a concern with the details of how
ogy. This is followed by the case ndings, which are subse- strategy is being practised in the everyday operations of
quently discussed. The paper concludes with some an organisation (for an overview, see Jarzabkowski et al.,
implications for further research. 2007).
While this body of research shares, on the one hand, a
common concern with practice, it exhibits, on the other
A practice theory approach to accounting hand, a wide variety of different theoretical approaches
to the topic. To be sure, practice approaches ultimately rely
Practice theory on an empirical inquiry into practising. But this empirical
inquiry is guided by, and its results are made intelligible
A concern with practice initially appears as an obvious on the basis of, the conceptual vocabulary of the practice
concern for any social theorist. For it is what people do, theory that is applied. It is the combination of an empiri-
why they do what they do, and what consequences their cally based micro-view on what people actually do with
doings have that social theorists always have tried to ex- a more abstract conceptual understanding of practices that
plain. But it is only with the practice turn in social theory makes practice theory an attractive vehicle to generalise
that the concern with practice has established itself as a across contexts, while at the same time acknowledging
more distinct stream of sociological inquiry (Reckwitz, the limits of such generalisations.
2002; Schatzki et al., 2001). This has to do with two differ- In our paper, we will specically rely on the practice
ent meanings of the notion of practice. On the one hand, theory of Schatzki (1996, 2002, 2005). Schatzkis approach
there is the more commonsense understanding of practice to practice comes in many respects close to the arguably
as what is being done, i.e. what people do in practice. more well-known variant by Giddens (1979, 1984), but dif-
Practice theorists share the concern with the details of fers in the conceptual vocabulary used to characterise
everyday practice with cultural theorists more generally practices. While Giddens distinguishes between different
(see Reckwitz, 2002) and, in this respect, often adopt a types of rules and resources as the dening elements of
methodological approach similar to, for example, ethno- practice, Schatzki characterises practices in terms of four
methodology (Garnkel, 1967) or symbolic interactionism main elements: rules, practical understandings, teleoaffec-
(Blumer, 1969). On the other hand, the notion of practice tive structures and general understandings. Rules comprise
can be used in the more specic sense of an array of activ- formulations including everything from statute laws to
ity (Schatzki, 2001, p. 2) or a routinized type of behav- rules of thumb pertinent to the practice. Practical under-
iour (Reckwitz, 2002, p. 249) that is constituted and standing refers to what is required by an actor in terms
kept together by several elements, such as human bodies, of his or her ability to identify, perform and respond to
things, rules, knowledge or emotions. It is their interest the events within a practice. Teleoaffective structures are
in social practices as distinct ontological entities that dis- sets of ends and projects towards which a practice is ori-
tinguishes practice theories from other types of cultural ented allied with the emotions that are commonly ex-
theories (Reckwitz, 2002). In a nutshell, practice theorists pressed in the pursuit of these ends. General
claim that social action and order are best explained with understandings, nally, are elements of practices (ends,
reference to the existence of social practices that span rules, emotions etc.) that are tied to the site of which the
across time and space.3 practice is a part; they are thus common to several prac-
A focus on such practices seems to be particularly rele- tices within this site.
vant for the study of organisations, given that organising is Schatzkis vocabulary seems particularly fruitful for the
study of organisational practices, insofar as it puts consid-
3 erable emphasis on the elements of intentionality that are
While actor-network theory (see Latour, 2005) is sometimes regarded
as a practice theory, it should be pointed out that most authors writing in inscribed in practices (see Ahrens & Chapman, 2007). Many
that tradition do not rely on the idea of practice as a distinct ontological business practices do have a strong teleological orienta-
category. Rather, they seek to explain order by looking at how entities tion, i.e. they have clear ends or projects inscribed into
(actants) connect to each other. In other words, they equate practice with them. Also, there are usually quite robust general under-
the practising of the actants. We would claim that core practice theories
go beyond this by combining the empirical interest in practising with a
standings that guide a set of organisational practices in a
conceptual concern of how practices can be theorized as ontological more indirect way, such as a concern with efciency or
phenomena. protability (Schatzki, 2002, p. 87).
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 187

Moreover, Schatzkis practice theory is appreciative of Hopwood & Miller, 1994) can be seen as the wider intellec-
the complexity of practices and of the uidity of their fron- tual tradition within which the concern with management
tiers. Practices, he contends, are open-ended with respect accounting practice has gained momentum over time. Only
to how they can be carried out (ibid., p. 83). This is so be- recently, however, has there been a more explicit formula-
cause practices usually allow for alternative ways of being tion of a practice theory approach to management account-
performed, in the sense that actors can, for example, choose ing which takes up many of the insights from the tradition
among different projects or ends that are conceivable with- of interpretive accounting research, while at the same time
in the practice. If, for instance, a business practice is sup- introducing new elements from practice theory as outlined
posed to be both customer-oriented and prot-oriented, above (Ahrens & Chapman, 2007). Ahrens and Chapman
then this enables actors to make trade-offs. To which extent (2007) draw upon Schatzkis approach to practices to ana-
such trade-offs are possible depends on how narrowly the lyse management accounting in a British restaurant chain
practice describes acceptable behaviour. Schatzki (1996, (see also Ahrens & Chapman, 2004). Management account-
p. 124) refers to this as practices having different degrees ing, in their case, emerges as the continuous effort of bring-
of totalitarianism. The more totalitarian the practice, the ing restaurant practices in line with the overall strategic
more it species correct behaviour, whereas in a less total- and nancial objectives of the organisation. Restaurant
itarian practice, a range of acceptable alternatives is estab- managers and their hierarchical superiors would discuss
lished (ibid.). What is acceptable in a practice is, of course, different alternatives for action and evaluate them by mak-
subject to discussion within the relevant social community ing sense of how these actions relate to the companys
(e.g. organisation). Therefore, disagreement about the rules strategic priorities and what nancial consequences they
to be followed or the ends to be pursued may lead to a ques- are likely to imply. Ahrens and Chapman describe, for
tioning of a practice (Schatzki, 2002, p. 84). example, the practice of menu design, and how this was
Organisational practices may be subject to such dis- inherently linked to considerations about the target food
agreement for various reasons. One important reason is gi- margin that every menu should more or less achieve. Ques-
ven by interdependencies between practices. If the output tions of protability were interwoven with operational
of one practice enters another practice, then practitioners decisions in the sense that the dish and its calculative
of the second practice have an interest in the specication presence in the spreadsheet became mutually constitutive
of the rst one. Sales people, for example, have an interest during the menu design process (ibid., p. 15). Financial
in production practices insofar as the latter can make it information on its own was not practically useful in all
more or less difcult for them to sell a product. Also, practi- but the simplest of settings (ibid., p. 21), and in most
tioners within the same practice may have different per- cases, it needed to be contextualised by relating it to stra-
spectives on a given practice because they engage at the tegic messages and operational activities. As operational
same time in other practices which provide them with com- activities were discussed together with nancial informa-
peting understandings (Schatzki, 2002, p. 150). Production tion and strategic messages, managers were engaging in
people may, for example, have a different perspective on both accounting and strategising at the same time.
management practice than accounting people, given their Managers local efforts to improve their restaurants were
different functional backgrounds, knowledge, and goals. thus informed by but also constructive of the nancial
The more complex an organisation is in terms of the and strategic imperatives advertised by senior manage-
interdependencies between practices, and the more plural- ment (ibid., p. 21). The practice of management account-
istic it is in terms of the different orientations of its prac- ing that emerged in the company was not only driven by
tices and practitioners, the higher the potential for the rules and objectives inscribed in accounting informa-
disagreement and debate (see Denis, Langley, & Rouleau, tion; it also related to the general understanding expressed
2007). Such disagreement may be solved in different ways, in the organisations strategic messages and the more prac-
such as by introducing a meta-rationality which translates tical understandings that managers had of the operational
different rationalities into each other or by negotiating a practices they were familiar with.
compromise between the different rationalities. The goal While Ahrens and Chapman (2007) explicitly build
of protability, for example, may serve as a meta-rational- upon a practice theory perspective, other authors have
ity that can solve conicts between sales and production provided analyses of management accounting practice
practices by translating their preferences into costs and which are broadly in line with such a perspective, insofar
revenues. But, likewise, sales and production people may as they have also looked at how actors skilfully carry out
nd a workable compromise without such an overarching practices in the pursuit of organisational objectives and
criterion (see Thvenot, 2001). the extent to which accounting information does or does
Management accounting practices can enable the reso- not shape these local efforts. Vaivio (1999), for example,
lution of (potential) disagreements because they introduce discusses the impact of accounting numbers on discourses
a common criterion for the evaluation of practices (such as about the customer in a UK chemicals company. In the case
protability) and/or because they provide a space that al- company, a rather informal and mostly ad hoc way of man-
lows a workable compromise to be developed. aging the customer was at one point in time challenged by
an initiative to more strongly quantify the customer and
Management accounting as practice her needs. Existing modes of operational action were prob-
lematised as being too uncoordinated and unfocused. With
The rich body of interpretive accounting research that the introduction of a set of non-nancial performance indi-
has developed since the 1970s (see Baxter & Chua, 2003; cators, a new calculable space (Miller, 1994) was created
188 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

that allowed a new form of knowledge based on numeri- accounting-driven concern with costs was thus linked to
cal clarity to emerge (Vaivio, 1999, p. 697). Many rules of major strategic questions. What provided ammunition for
operational practice, in functional areas like eld engineer- the debate about indirect costs and eventually turned it
ing and the companys invoicing routines, were reshaped into a trial of strength (Latour, 1987) was the fact that
by this formal knowledge (Vaivio, 1999, p. 702). strategic concerns such as those with exibility had a
While, in this way, accountings inuence over func- rather complex relationship to accounting numbers. As
tional areas increased, sales managements inuence de- Mouritsen (1999, p. 39) explains:
creased. Over time, however, the pendulum swung back,
[I]t was difcult, if not in principle impossible, to
as the Quantied Customer was challenged by a compet-
produce an accounting calculation which would sort
ing knowledge that was rooted in sales managers local
out and evaluate the pros and cons of exibility.
interpretations of customer needs. The sales managers
Although accounting performance was drawn upon
perspective on the customer recognised the complexity
to produce and frame a debate concerning exibility,
of the business, the customers at times ambiguous de-
it could not dene the desirable level of exibility
mands, and their unstable preferences. This competing
and it merely enabled a debate about its pros and
knowledge mobilised the qualities of particular customers
cons (. . .). The debate connected indirect costs and
against their quantitative representation: the numbers
exibility but did little more than hint at possible
that represented The Customer became dubious proxies
consequences of different modes and levels of
for something more subtle that could not meaningfully
exibility.
be quantied (ibid., p. 705). To end or at least mitigate
the controversy between these two modes of management, If the costs and benets of exibility in production are dif-
the company eventually reduced the size of the calculable cult to pin down, would the same not be true to an even
space and decided to separate it more strongly from oper- bigger extent for the costs and benets of design decisions
ational management, resulting in a more passive and in new product development? And if, as Vaivio (1999) ar-
unintrusive form of monitoring and control (ibid., p. 709). gues, customer-related practices are often rooted in
Vaivios (1999) case study illustrates how different ap- knowledge which escapes extant quantication, should
proaches to managing the customer are informed by differ- the same not hold true to an even bigger extent in the case
ences in functional backgrounds and local practices. of NPD practices?
Mouritsen (1999) provides evidence on a very similar de- While nancial, strategic and operational arguments
bate between competing management control approaches. were shown to be highly compatible in Ahrens and Chap-
In his case, however, the debate was not originally moti- mans (2007) case of a restaurant chain, both Vaivio
vated by a concern with the customer, but rather by a con- (1999) and Mouritsen (1999) show that this is not auto-
cern with costs. BusinessPrint (the case company) faced a matically the case in every setting. Conicts between dif-
high amount of indirect production costs, and there was ferent types of knowledge claims may result in different
agreement among managers that this posed a threat to opinions on how to best account for organisational prac-
the companys overall protability. The CEO regarded indi- tice. And not always can this plurality of perspectives be
rect costs through the perspective of the existing contribu- translated into the common language of accounting, as
tion accounting system which focused on variable (direct) illustrated by Mouritsen (1999). Moreover, efforts to do
costs. For him, indirect costs represented an unexplained so may be challenged on the basis of accountings only
lump sum that resulted from the high degree of exibility imperfect representational qualities of the practices in
in production. He thus promoted an integrated information question, as shown by Vaivio (1999). In such cases, the eco-
system (based on the existing contribution accounting ap- nomics of calculation are replaced by a politics of expla-
proach) that would help increase visibility over the pro- nation (Mouritsen, 1999, p. 49), where different types of
duction process, thereby improving the possibility to accounts compete against each other in making sense of
plan, monitor, and stabilise production, which would even- the desirability of a particular set of practices. This is still
tually help reduce the indirect costs by reducing exibility. about accounting, but not in the narrow sense of calculat-
The production manager held a different view: while he ing with numbers, but in the wider sense of the production
also saw the need to manage indirect costs, this was not and negotiation of accountability.
to be done through increased integrated planning and Over time, the relative strength of particular types of ac-
monitoring but through local hands-on efforts of coordi- counts may result in a stabilisation of a certain pattern of
nation. Integration was thus to be achieved through talk accountability, such as shown by Vaivio (1999). Ahrens
and improvisation, maintaining a high degree of autonomy (1996) refers to such a pattern as a style of accountabil-
at the shop oor. ity. In his eld study on British and German breweries,
Again, as in Vaivios (1999) case, different conceptions Ahrens contrasts British and German management accoun-
of control are related to different functional backgrounds, tants perspectives on the relevance of accounting informa-
knowledge, and interests, and although the debate in Busi- tion. Commenting upon the specic case of a leaking roof
nessPrint was motivated by a concern with costs, it clearly in a British Brewery that had to be repaired, he states:
had wider ramications: Indirect costs represented more
than reduced protability as they also gave impetus to a The British view seemed to be that nancial perfor-
possible redenition of what exibility should be, what it mance objectives together with strategic ambitions
meant to be innovative, and consequently to what it meant to build brands are much more important than the
to serve the customer (Mouritsen, 1999, p. 38). The repair of a leaking roof. The Germans, on the other
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 189

hand, emphasized the integrity of operational pro- (1999), remains to be seen. It may well be that knowledge
cesses and expressed their indifference towards claims and rationalities are ultimately reconciled in a way
any accounting information which does not mirror similar to what could be observed in Ahrens and Chap-
what they regard as the underlying operational mans (2007) restaurant chain. If indeed they are recon-
economies of the organisation. It appears that the ciled, it will be interesting to see how this takes place,
British senior management accountant and his given the differences in settings.
nance director primarily hold themselves account-
able to managing their organisation such that it gen-
erates revenues now and in the future (. . .). The Research design and methodology
German manager and the senior management
accountant both seem to hold themselves account- The analysis in this paper is based upon a eld study
able to a concept of management which puts opera- conducted in a single organisation. We adopt an under-
tional integrity and economy before reported standing of eld study research according to which the
earnings (Ahrens, 1996, p. 152). main task of the researcher is to inquire into a eld of prac-
tices and to make sense of his or her observations by
Where this resonates with Vaivios (1999) and Mouritsens
abductive reasoning, i.e. by moving back and forth be-
(1999) ndings is in the illumination of competing concep-
tween data and theory (Ahrens & Chapman, 2006). This re-
tions of control and their relative importance within an
quires, rst and foremost, a close proximity to the eld
organisation; what it adds is the idea that such differences
(Garnkel, 1967; Jnsson & Macintosh, 1997). To identify
may be made intelligible in light of the wider institutional
and understand practice as it happens, argues Schatzki
and cultural setting of the organisation (see also Kurunm-
(2005, p. 476), requires considerable participant observa-
ki, 2004; Lounsbury, 2008).
tion: watching participants activities, interacting with
them (e.g. asking questions), and at least ideally
Research focus attempting to learn their practices. While we certainly
cannot claim to have learnt the complex set of practices in-
While Ahrens (1996) focused on comparing styles of volved in new product development, we feel that we have
accountability between two sets of organisations (German obtained a rather intimate understanding of what the peo-
and British breweries) and thus argued on the level of the ple in our case organisation do and why they do it. This
organisation, we are more interested in shedding light on was facilitated by the fact that the possibility to carry out
the style of accountability in one particular set of organisa- the research in the company was linked to our commit-
tional practices, i.e. NPD practices. Accordingly, while Ah- ment to provide feedback and recommendations at the
rens (1996) explained his ndings mainly on the basis of end of the observation period. This commitment was also
different organisational environments and different con- a commitment for the company to support us in sourcing
ceptions of the organisation, we link our ndings to the the information we required.
particular nature of NPD practices and to the ways in The presentation of our empirical ndings follows the
which these are organised in our case company. Our case narrative approach to writing (Czarniawska, 1999), which,
organisations NPD practices thus constitute the site in our view, is an adequate means to convey the richness of
(Schatzki, 2005) in which, and as part of which, accounting a qualitative eld study. While critics may think of narra-
practices take on a particular shape. tives as nothing but storytelling, [o]nce a narrative is
The particular site of NPD promises to bring to the fore understood as a template it is potentially transferable
aspects of management accounting practice that would across contexts and can be reinterpreted to become rele-
usually not emerge in the more commercial and arguably vant to other settings (Llewellyn, 1999, p. 225; see also
routine setting of a restaurant chain, as analysed by Ahrens Dyer & Wilkins, 1991). As such, it pleads for a contextual
and Chapman (2007). Decisions on menu design or mar- rhetoric of generalisation (Lukka & Kasanen, 1995; see also
keting efforts, which constitute an important part of man- Flyvbjerg, 2001). To do so requires profound theorising of
agement practices in a restaurant chain, have a more the empirical data, rather than leaving them as merely rich
immediate relation to accounting numbers than decisions accounts of the complexity of social life (Ahrens & Chap-
about NPD. In the latter, there is a wider timespace dis- man, 2006).
tance between decisions and outcomes and thus more The eld study was conducted over a period of 16
uncertainty about results. Also, there exist arguably more months, from September 2004 to December 2005. During
competing demands upon NPD practices in a technology this period, one of the researchers spent approximately
rm compared to menu design practices in a restaurant three days per week every second week at the company.
chain. While the high degree of plurality increases the po- Far from all days had programmed activities, but to ob-
tential for disagreement among competing rationalities, serve daily activities as they went on outside the articial
the uncertainty about results makes it difcult to translate interview situation and to increase acquaintance with
these rationalities into the language of accounting num- employees and managers and thereby increase access to
bers (Mouritsen, 1999). At least, efforts to do so may be re- potentially rich data, a large number of working days were
sisted on the basis of accountings only imperfect spent at the company. Formally, the research is based on a
representational qualities (Vaivio, 1999). Whether this combination of three types of data: interviews, archives
eventually materialises in a conict between different ac- and participant observation. Two formal rounds of inter-
tors, such as evidenced in Vaivio (1999) or Mouritsen views were conducted with a twelve-month interval (see
190 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

Fig. 1a). The rst round was rather wide in its scope both in Today, the company pioneers new product concepts in sev-
terms of the people interviewed and the topics covered to eral elds. In 1997, Company doubled both revenue and
obtain an overview (Ahrens & Dent, 1998), whereas the number of employees through acquisition. This put the
second was more focused and thus may be described as company in a position to be rst and best in analytical
theoretical sampling (Strauss & Corbin, 1998) to probe dee- solutions within segments of the food, agricultural, phar-
per into themes identied in the rst round and in general maceutical and chemical industries. From a small start
during the many days spent at the company. with a staff of three, Company has grown to incorporate
Interviews were recorded and transcribed. Notes were all aspects of development, assembly, marketing, sales
taken during the interviews and more detailed notes were and after-sales service, employing 1100 people in 21 com-
taken immediately after an interview while listening to the panies world-wide. Companys export share is 97%, and it
recordings not to interpret and code the data in a nal is nancially strong with a solvency ratio above 70% and
sense, but to make sense of what had been said and hence a steady pre-tax return on sales of 13%.
to continually revise the understanding of what was going Company has a niche focus, typically holding a market
on in the organisation. share of 80100% within its target markets. Companys
The archival data studies were primarily concerned products are often the most expensive, but they are also
with the authorisation documents for the NPD projects re- the most reliable and accurate on the market. They tend
searched. In the case company, each time a project had to have more functionalities than competitors products
developed a stage further it had to be authorised by the and are also differentiated in terms of design and support
Management Board (more about this later) to pass the gate service. Company uses the terms dedicated, user-
and move on to the next stage. Stages, gates and authorisa- friendly and tailor-made to describe its products. Also,
tion criteria were all described in the NPD process manual, Company uses the term solutions, rather than the term
which was also studied carefully. Two projects were fol- instruments, because it seeks to offer customers far more
lowed more closely: one in the late stages of development than technical equipment. Pre-sales consulting, ready-to-
(rst variants were released for sale in autumn 2005) and use calibrations, customised application development,
one in the early stages (followed during the concept and and industry discussion forums on the Company intranet
system design processes). However, strategic documents are supposed to create a value proposition that the cus-
about future plans and objectives and other top manage- tomers are willing to pay a premium for.
ment communication also were reviewed (see Fig. 1b). Division is the product development and production leg
Documents were typically accessed through the companys of Company (see Fig. 2). Division employs more than 500
intranet, handed out during interviews, or left on the people at two sites, one in Denmark and one in Sweden.
researchers desk. The two sites are run as one company (i.e. with one CEO,
Data from observation were retrieved in an unstruc- one R&D director etc.). The R&D Department within Divi-
tured and informal manner. A number of NPD project sion employs about 135 people. About 95 of these are em-
meetings and department meetings were attended. Out- ployed at the Danish site, where we conducted our
side of meetings, observation took place during lunch research. The remaining companies within the group are
breaks, coffee machine conversations, chats over the desk, sales and service subsidiaries.
and by overhearing and observing communications and ac-
tions of daily operations.
Case analysis

Case organisation Our analysis of NPD practices in Division is informed by


empirical material about the NPD process more generally
The case organisation, Division, is owned by Company, a as well as by data on two specic NPD projects, which
private company with annual revenue of about 170 million we hereafter will call Alpha and Beta, respectively. Al-
euros. According to Companys mission, it provides rapid, pha was a project aimed at developing a family of instru-
reliable and dedicated analytical solutions for routine con- ments for the analysis of liquids, as needed in the wine
trol of quality and processing of agricultural, food, pharma- and dairy industries. The project concerned an ambitious
ceutical and chemical products. Company was founded in revamp of earlier developed versions of these instruments.
the 1950s, and from the start the fundamental business Importantly, Alpha was the rst project that was based on
idea was to automate otherwise time-consuming analyti- the concept of modularity, as explained in more detail be-
cal methods and to offer a faster, more accurate, reliable low. Beta was the project immediately after Alpha. It also
and cost efcient method of analysis. The general principle concerned the development of an instrument for the anal-
behind Companys instruments is to illuminate the sub- ysis of wine, but it was less sophisticated than Alpha inso-
stance to be analysed and to measure the infrared or far as it was targeted not at industrial customers, but at
near-infrared refraction. The refraction is then compared farmers. On average, an NPD project at Division would last
to a reference database from which it can be deducted that for about 3 years. Alpha was more time-consuming due to
if you are measuring milk, for example, and this is your the whole discussion about modularity as explained below.
refraction, then your sample contains so much fat, so many It was initiated in 2000 and was in the nal stages when
bacteria etc. we conducted our empirical research (20042005). Beta
Agriculture was the main market in the beginning. La- was rst authorised in 2004, just before our research work
ter, the dairy industry became the largest customer group. started.
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 191

Round one Round two

Time December-January 2004/2005 November-December 2005


Interviewees 1 CEO 1 CEO
1 Finance director 1 R&D director
1 Production director 1 middle manager
1 R&D director 2 project managers
1 Group controller (from Company) 1 market analyst
11 middle managers across the 3 development engineers
organisation 9 interviews in total
1 R&D strategist
1 project coordinator
1 production planner
19 interviews in total

Fig. 1a. Interviews.

Document types Examples/Description


Authorisation documents Authorisation documents, prepared for the meetings at the
gates, from five different projects were reviewed. The
documents which related to Alpha and Beta were studied
in detail.
Ad hoc analyses The Test Platform
The costs of poor quality in Company
Modularity Means and Ends
The Ranking Report
Monthly management reports From the CEO of Company
From the CEO of Division
From the Quality manager
Other Strategy documents
Manual for the R&D Development process
Consulting reports

Fig. 1b. Documents reviewed.

Our analysis is organised in two sections. As accounting our research, Company made an important strategic deci-
plays a role both in the formulation and renement of the sion that would inuence all subsequent NPD projects. As
strategies that underlie NPD projects and in the day-to-day our analysis will show, this decision was to some extent
operations of these projects, we dedicate the rst part of based on nancial considerations and accounting num-
our analysis to strategic issues and the second part to the bers; it was also the starting point for a continuous practice
more operational tasks. That said, it will become clear of strategising during the course of which the nancial
throughout the text that the domains of strategy and dimension of NPD would repeatedly be enacted.
operations cannot be easily separated but must really
be understood as closely interwoven. The making of the strategy of modularisation
The making of the strategy of modularisation started in
Accounting for the strategy of new product development the late 1990s, when a customer with whom Division had a
very close relationship had purposely poured wine into an
Company is a very R&D intensive business. Its success instrument originally developed to measure levels of dif-
has been based on innovative products since its beginnings ferent constituents in milk to see what happened. The
in the 1950s. New product development is thus a critical instrument produced meaningful data by comparing the
area for Company. The particular practices carried out in measurement with some reference data. When wine was
NPD are inuenced by broader strategic and nancial con- used instead of milk as the object of analysis, the instru-
siderations regarding where the company wants to go and ment still produced output data, but these data had no
how it wants to go there. A few years prior to the start of immediate information value, as they had no relation to
192 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

Company

Division Sales & Service


Subsidiaries

Finance HRM

R&D Production Sales & Marketing

Team Project Support

Business Development Team Greenhouse Team Hardware Creation

Team Module Creation

Team Product Creation

Fig. 2. Organisation chart.

the reference data used for milk. However, this incident never even think about doing that in the old days
made the company inquire in more detail into the output because it would require a total reconstruction of
data produced in the case of wine, and the engineers rea- the instrument to install a new ow system. If we
lised that, ultimately, these data would have some infor- made it [modularisation] right, we could substitute
mation value if compared to a meaningful reference the existing ow system module with a specially
point. This suggested that the original instrument could designed ketchup module. A ketchup module may
eventually be sold to a completely new market segment. cost 125,000 euros, and then it may be protable
The company therefore embarked on a development pro- to sell only 50 or even 25 instruments. That is an
ject which resulted in a new instrument for the wine easy calculation! Is it protable to develop a module
industry, which was successfully sold to wine producers. for 125,000 if you can sell 25 instruments to Heinz?
After some time, several customers suggested that they That would never be protable in the old days. Senior
could make use of a similar instrument for the analysis of Manager, Team Greenhouse, when asked to tell the
the constituents of grape must. This led the company to ketchup story that everyone still remembered him
adapt the instrument again and to produce a third product carrying on about back in 2000.
variant of the instrument originally designed for milk. A
The nancial benet expected from modularity was that by
couple of years later, feedback from sales people and engi-
investing in platform development from which technology
neers suggested that it was about time to redesign the
and even physical modules could be reused, future variants
three instruments.
would be much cheaper to develop. These nancial bene-
It was at this point that a senior manager came up with
ts were not obvious, however, for it proved difcult to
the idea to design the three instruments on the basis of
translate the intuitive thoughts about the benets of the
reusable modules rather than in isolation from each other,
new strategy into hard numbers.
as the concept of modularity seemed to be of wider rele-
The pending project involving the three instruments for
vance for the companys products. The story that the man-
analysis was used to try to calculate whether modularisa-
ager promoted was that by basing products on combinable
tion would pay off. But there were two problems with the
modules, the company would be able to serve low-volume
comparison of the modular with the non-modular product
customer segments with special requirements:
variants of the three instruments. First, the existing
When you have really seen modularity, you are actu- spreadsheet model that was usually used for such prot-
ally in a position where you can enter markets that ability calculations has always been regarded with some
you would never have entered otherwise because it scepticism, for it was based on a generic NPD project
would be too costly. Lets pretend that Heinz came with particular characteristics. NPD projects differed quite
by and said that they would like to make some mea- considerably, however, and translating them into the
surements on ketchup and that it required a special terms of the model was often an act of creativity. In other
ow system to handle this substance. We would words, protability calculations at this early stage were
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 193

considered somewhat problematic in principle (see also tecture of Divisions instruments, but their entire business
our discussion below). Secondly, the existing spreadsheet model. It signied a major change in the strategic approach:
model was designed to handle only one product at a time, from being primarily explorative and market leading in
while in the case of modularity, there would be several terms of technology, Division had to become much more
products or product variants that had to be evaluated in exploitative and begin to reuse existing technology rather
tandem because they would share one or more modules. than invent something new. Internally, the R&D and Pro-
Normally, modules that can be shared among projects are duction Departments were reorganised to accommodate
more expensive in their development costs and variable the changes. Most importantly, engineers needed to change
production costs than stand-alone product parts. This their approach to design products, as the project authorisa-
was expected to be more than offset by the savings in tion report for Alpha made clear:
development costs over time and also savings on variable
The traditional strong focus on cost on the individual
costs due to economies of scale; however, these expected
instrument must be changed to a focus on the cost of
savings are not really visible in the accounting criteria used
the whole technology family.
in Division to evaluate projects if they are evaluated indi-
As an example:
vidually in the sense that all initial development costs
are fully allocated to the rst project.
 The developer who designs a unit without any con-
Throughout the debates, the Senior Manager was per-
siderations of modularity and optimises that unit to
sistent, and eventually he made a breakthrough:
an individual instrument and to a low cost gets no
Finally, after one of my numerous presentations to points in [Alpha].
the Management Board trying to get them to follow  The developer who designs a unit with consider-
my idea of modularisation, [the CEO] cut through the ations of modularity and optimises that unit as a
talk and said that he was fed up with this discussion. module to be used in many instruments, but to a
He believed in the idea [of modularisation] and the higher cost than in the previous example, gets a lot
project [Alpha] was hereby authorised! If he had of points in [Alpha].
not done that, the idea would have been scrapped. (Authorisation Report for Alpha, 2001, p. 8)
Senior Manager, Team Greenhouse.
And the report continued to acknowledge that this might
The decision to go modular was made in 2000. It was not a not be an easy change process: This is different from
direct result of the existing accounting model that was what we have asked our developers to do in previous
used to estimate the nancial protability of NPD pro- R&D projects and will for sure result in some confusion.
jects; rather, the CEO believed in the idea. Maybe he really The customer side was also up for revision since new
did, or maybe he had to because in the new ve-year market segments could be potentially targeted. Also, the
strategy plan, the CEO announced his double or trouble modular technology would make it possible to repair indi-
slogan signifying that he wanted revenues to double with- vidual modules that customers had returned and to sell
in ve years. The primary way in which Division could them as second-hand modules to customers willing to
contribute to this was by launching more new products buy used material. Thus, the management board saw mod-
faster. When Divisions management pondered how to ularisation as an opportunity not only to contribute specif-
realise this, the idea of modularisation was already there, ically to the growth strategy imposed on them from
waiting for a problem to solve. So, despite the problems of headquarters, but also to improve a set of strategic imper-
calculating the benets of modularisation or, perhaps, atives such as decreased lead time due to delayed product
because of these problems the CEO made the decision differentiation, increased customer satisfaction from im-
to go modular. Part of the rationalisation of this strategy proved testing abilities, efcient use of resources in the
was provided by the belief in the strategic importance of R&D, and increased rate of innovation due to development
modularity, which intuitively seemed to be a good way of technology platforms and product families based on
to increase the number of product launches and make modules.
product development less costly overall. Another reason The somewhat intuitive way in which the strategy of
to go forward without knowing the exact numbers was modularisation was decided upon was not particularly
the idea that the rst concrete NPD project that would exceptional for the company. Rather, it reected the com-
be based on modular design (the project Alpha) would panys practice of dealing with uncertainty in NPD projects
deliver the necessary answers to the yet open questions. more generally. The CEO and his management team knew
In light of these considerations, the decision to authorise that the spreadsheet model could not serve as an answer
Alpha was a logical next step. Although this decision machine (Burchell, Clubb, Hopwood, Hughes, & Nahapiet,
was formally separated (Alpha had to go through the 1980). But they had found a way to make decisions based
usual decision process for NPD projects), it was de facto al- on the model, even if the information in the model was
ready made when the decision to go modular was made not applied in a strict sense. Decisions were made by draw-
(as indicated in the above mentioned statement by the se- ing upon criteria which were not accounting criteria as
nior manager). such, but which nevertheless had a relationship to account-
There were perhaps other answers that could have been ing numbers. Considerations about functionalities, quality,
mobilised as a strategic move towards doubling revenue, production design and the like were not nancial consider-
but the CEO and his management team specically chose ations in the rst place. Implicitly, however, managers
this one. It was no small step. It involved not only the archi- would use them in their argumentation and sense-making
194 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

as shortcuts for the expected nancial impact of their deci- It think it is possible [to be successful with modular-
sions. Thinking in terms of more strategic or operational ity], but it is wrong to decide after just one round. It is
categories seemed to be a way to account for the rather like with projects. . .a team is only really good when it
limited explanatory power provided by the numbers as has had a couple of runs. Same with this we need to
such. go some rounds before we are really in gear before
Management had been quite aware that modularity we can run more smoothly without all these coordi-
was a concept that could bring about different types of nation meetings. Then it will be more a matter of rou-
benets for the company, as there were different strategic tine and we will just grab on to someone ad hoc
considerations linked to the use of modules. How exactly rather than through formalised meetings. Manager,
the company should exploit the possibilities of modularity Team Module Creation.
was not exactly clear at the time of deciding to go for mod-
Others, in contrast, took a more critical stance towards
ularity. However, the board expected that the rst NPD
the new strategic message:
project which would be based on modularity (Alpha)
would clarify the open issues:
I think it is a huge weakness that we have not
When we decided on the modularisation scheme, we dened explicitly what we want to achieve with
said that Alpha had to pave the way, and then we modularisation. We have a number of objectives:
would have to learn as we went along instead of faster production, testable modules, etc. But in mar-
doing the big theoretical groundwork from the out- keting they have a different idea they are thinking
set. Hell, we would not have got past that yet of modularisation as something that will enable
and the world would have passed us by. CEO. them to sell any conguration of the instrument
This process of learning as they went along turned out to highly driven by software conguration and there
be even more intense than management had envisaged, are many more. I think they are all very good, but
however. The decision to go modular was by far not the they only to some extent overlap and no one ever
end of the making of the strategy of modularity. really determined how much these ideas should
overlap. R&D strategist.

Accounting and strategising


Whether the given trajectory was approved did not
Once the decision about modularity had been made, the change the fact that decisions about the form and num-
questions surrounding the modular design were handed ber of modules had to be made in practice. How this was
over from the management board to the development engi- done can be illustrated with the example of Beta. Beta
neers. Bringing the strategic message to life in the day-to- was supposed to lead to the development of two variants
day operations of developing new products proved difcult, of a wine instrument and was in this sense a less com-
for what was a module? It turned out to be a major chal- plex project than Alpha. The accounting practice involved
lenge to develop self-contained subassemblies that would in this process was implicit and very qualitative. The
meet the requirements of the instrument variants to be development engineers in Team Greenhouse suggested
developed in a particular project and potentially also what they believed to be the best solution for the num-
requirements of instruments to be developed in future ber and size of modules, and described alongside it two
projects. extreme cases: one with very few modules and one with
In the case of Alpha, the many promising aspects of the largest number of modules that they could imagine
modularity identied in the strategic decision-making pro- the instrument being split into. With these three scenar-
cess became a hurdle. Not all potential benets were ios in hand, they met with representatives from all the
mutually sustainable, e.g. reuse and combinability called different areas of the company which they thought
for small and relatively simple modules, whereas improved would be affected by the degree of modularisation. In
quality and the development of an exchange system re- this meeting, the trade-offs between different strategic
quired what came to be known as intelligent modules that and operational goals were discussed. Fig. 3 summarises
were big and costly due to the necessary testing require- the conclusions made by the NPD team after the
ment for these specic purposes. meeting.
In line with managements view that Alpha would pro- The creation of the modular architecture of Beta is a good
vide a learning opportunity, some people within the devel- example of how local practices resulted in different
opment teams were not very concerned about these perspectives on the question of modularity, which were then
apparent difculties: negotiated in light of the overall strategic imperatives that
Division had set. The organisational members all knew
I am unsure whether we succeed in meeting the Divisions mission of developing high-quality analytical
objectives set up for the NPD side of things. It does instruments for particular industries, the strategy of modu-
not look like it, but we shouldnt stop believing in larisation and, in Betas case, the low cost approach. How
the idea just learn from Alpha. Alpha is not a go they each envisaged these aims, however, differed depend-
or no-go for modularisation. We will get some things ing on the local practices in which they participated. The Ser-
right and learn a lot, but it wont be a clear success. vice Engineer from Sales & Marketing was thus concerned
We will get something to develop on that is how I with what effect the proposed modular architecture would
will put it. Project coordinator, Production. have on customers cost of ownership, whereas the Project
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 195

Department Fewest modules Most likely Most modules


R&D Modules too complex The splitting Too time consuming
Too little flexibility corresponds to Very flexible
functionality, which
makes development
very manageable

Production Modules not Suitable size in terms High yield due to easy
manageable in of physical testing
assembly due to manageability in
physical size assembly
Danger of poor yield4 The splitting largely
due to quality issues corresponds to test
from problematic structure, which in
testing of such general is good for all
complex units sub-departments in
Production

Sales & Architectures


Marketing implications of
customers cost of
ownership are unclear
What is the service
strategy for the
instrument, and how
does the proposed
splitting fit with this?
Module size not so
important, as long as
it is easy to repair

Total Probably the poorest This was clearly the This solution received
solution. Several people most desired solution. some positive
expressed that they No direct negative comments, but it was
would have to divide effects were mentioned. generally believed that
modules into smaller Division did not yet
units to lift the local possess the modularity
tasks for which they maturity to reap the
were responsible. benefits of this scenario.

4
Yield = number of modules passing th e quality test in the first attempt.

Fig. 3. Conclusions from discussion of how to modularise beta.

Coordinator from the Production Department representing this presented a potential source of conict between differ-
the interests of Assembly had an interest in modules being ent actors:
physically manageable. The Quality Manager, of course, ar-
gued for an architecture that would increase yield4 I think that the production departments denition
and thereby make him achieve targets and full the role he of modules is what the development engineer thinks
had. of the most a self-sustained testable module,
There were no formal control systems through which all whereas at a more general level in the R&D depart-
these ends could be translated into an optimal solution. ment, we think more about the idea of reusability.
There were, however, skilful practitioners who would dis- However, what is being realised is not really as reus-
cuss with each other how an appropriate solution for this able as we hoped, and I think there are more sur-
particular project could look like (March & Olsen, 1989). prises waiting for us. It will probably be the
These practitioners shared a general understanding of the production department that will harvest the biggest
strategic objectives of Division; they differed in their expo- advantages. R&D strategist.
sure to local practices and in their practical understanding
In other words, strategic decisions taken at a certain point
of these practices.
would come up again at later stages in the process when
Such agreements about the approximate number and
they had to be broken down to a more operational level
design of modules, as in the case of Beta, were often not
on which discussions about trade-offs between strategic
the end of the story, however. Discussions would arise
goals (strategising) would be continued.
again later in the process, when the rather abstract deci-
This iterative process of accounting for the benets and
sions about the number of modules had to be translated
costs of different design solutions seemed, at least to some
into the concrete physical realisation of the parts. Again,
extent, inevitable, given the evolving nature of the NPD
projects. It was clearly impossible to decide upon every de-
4
Number of modules or instruments passing the test in the rst attempt. tail in advance. However, there was also dissatisfaction
196 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

with too much iteration, and part of this dissatisfaction 70 Hours spend to test one
was reected in the ranking report. This report was writ- 60
67 h [pre-Alpha instrument]
Max time
ten in November 2003 by staff members from the develop-
50
ment department of Division and was intended to rank the Average time
40
driving forces behind modularisation. In the introduction, Minimum time
the rationale for the report is made clear: 30
16 h Additional time for test
20 24 h of modules
This project has its origin in the development
Time to be saved
10 3h
department. The lack of ranking makes the early
5h
decision processes much harder, because it is not 0
Total test time pr. instrument
fully clear what is most prioritized and best for the
whole company in the long run. Increased costs in
the development phase do not necessarily lead to Benefit saved test time
decreased protability, because of the great savings
later in the value chain by for instance reduced man- 16 hours / instrument
ufacturing costs, reduced warranty costs and a more
119 DKK / hour
timely market introduction due to less rework at = 1.904 DKK / instrument
production ramp-up (Ranking Report, p. 4).
320 [Alpha] Instruments per year
The report was based upon interviews with employees of = 609.500 DKK / year
seven different departments as well as with representa-
tives from two customers. The results showed quite Estimated annual savings in [Division] are:
= 3.5 to 4.5 m DKK
remarkable differences in the perceived importance of
the various aspects of modularity. Hence, the authors of
the report concluded with the following statement:
Note: Absolute numbers were changed, but the same ratios apply.
This report could be a good basis of a further mod-
ularisation discussion within the management Fig. 4. Presentation slides used to make the case for the test platform.
group and hopefully facilitate future decision-mak-
ing. The recommendation for the next step is that
the idea of testing modules individually, or whether this
the management group meets and hammers out a
benet of modularisation should be given up. The issue ar-
strategy and priority list for the modularity imple-
ose during the project Alpha and was eventually solved
mentation in [Company]. It is important to declare
through a topdown decision by management, which
a common business focus and to diminish the dif-
decided that the test platform should be developed. Again,
ferences between the departments by, for instance,
this decision was to some extent based on calculations
meetings, discussions and education. To be able to
through which the costs and benets of the test platform
benet from all the advantages of modularisation,
were brought to the fore. Fig. 4 displays two slides taken
all departments and employees have to be
from a presentation in which the case for the test platform
involved, and it should be high priority to increase
was made. The rst slide shows the expected time savings
the knowledge and understanding of modularisa-
of an investment in the test platform; the second one
tion and customer needs. (Ranking Report, p. 34)
translates this into cost savings, based on an hourly full
What is notable here is that the report acknowledges cost.
the need for both a clear prioritisation by top management The representational limits of these calculations were
and for aligning the perspectives of different departments also acknowledged, however. Benets and costs were ulti-
in a more bottomup way through meetings or discus- mately not only or primarily discussed in terms of account-
sions. Strategic priorities had to be determined at the top, ing numbers. Rather, the decision about the test platform
but they equally needed to be continuously discussed was driven to a large extent by the question of what kind
and aligned within and between the departments. of strategic benets would be concerned by this decision:
Such a pattern also became evident in the question of
On the one hand it is obvious that this is what we
whether the company should invest in a test platform for
should do; on the other, it is very uffy. Do we
modules. One of the expected benets of modularity was
believe in it? There are a lot of in-built assumptions.
that each module would be individually testable rather
The upfront investments are enormous, so we had to
than only as part of the overall product. This individual
calculate it and had to understand how many of the
testing would provide more exibility and make problem
initial objectives related to modularisation were
analysis much more efcient. To test modules, a test plat-
actually going to be realised by investing in develop-
form was needed; unfortunately, the development of such
ing the test platform. Production Director.
a test platform would be a major investment for the com-
pany. Also, testing modules would increase the develop- But even though there was a decision to develop the test
ment costs of the modules themselves, since they would platform, the Production Director acknowledged a need to
have to be developed in a way that would make them test- consider anew in each case whether the modules should
able, which required some additional engineering. The be developed according to the needs of the testing platform
question was thus whether the company should stick to or whether a different design would be more appropriate:
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 197

Now we have accepted this in principle. It has been Three different teams were responsible for the project
all the way round the management board and the at different times during the development process (see
Group management. We know the assumptions, Fig. 5). Team Greenhouse did the fuzzy front end work
and we have decided to go ahead. But that does and then left it to Team Module Creation to physically rea-
not mean we should just do this [to use the test plat- lise the components of the new instrument. Team Product
form] in every project. We should use our brains and Creation was responsible for the integration of modules
evaluate whether the particular project is really sui- and other subassemblies into whole, functioning instru-
ted for the test platform. In general, yes, the test ments and for transfer of projects to production. During
platform is a good idea, and [Alpha] will demon- the research period the responsibility for the system de-
strate just that. But there may be special test needs sign stage was shifted from Team Greenhouse to Team
in future projects where we should choose a differ- Module Creation because the tasks involved were consid-
ent way. Production Director. ered to be too detailed to match the competencies held
by Team Greenhouse engineers.
Again, this may be seen as an example for the need to
At each gate, the project manager of the Team who cur-
strategise anew in each case, i.e. to dene the priorities
rently worked on the project had to prepare a report that
of how to design a particular product. This process of prior-
showed the key nancial indicators and other relevant
itising was not only a matter of deciding on the trade-off
information. The report was discussed in a meeting with
between strategic objectives; it also took place on a more
the management board, during which the board would
operational basis in the day-to-day work of NPD, in which
scrutinise the validity of the numbers presented. Since a
different possibilities to design and engineer a product had
small percentage of all details could be incorporated in
to be compared with respect to their impact on
the report or discussed in the meeting, the management
protability.
board developed a practice of raising sample questions.
These were questions that demonstrated to the board
Accounting in the everyday operations of NPD
whether the project manager was able to convincingly ar-
gue the case for continuing with the project. If the sample
Accounting at the gates
questions were answered satisfactorily, this indicated that
The decision process at each gate basically resembled
the project manager could have told the full story, if time
the process of initial authorisation of a new project. To
had allowed him to do so (see Jnsson, 2004).
become and continue to be an NPD project in Division,
The evaluation of the project during the meeting was
new ideas and projects in process had to show a certain
centred on the main accounting criteria. But it went be-
contribution ratio level and a payback ratio level found
yond accounting numbers insofar as the management
as the sum of ve years expected contribution margin
board tried to become clear about whether the project
divided by the development costs. These KPIs were
manager was aware of the various relevant dimensions of
calculated in the spreadsheet model mentioned above.
the NPD project:
Divisions stage-gate (Cooper, 1999) model (see Fig. 5)
had seventeen stages out of which seven were considered Its everything. Each of us [directors] represents spe-
so important that to pass the gate and go on to the next cic interests and possesses specic competences
stage required authorisation by the management board, and knowledge. It may be the costs: are they calcu-
which included all directors (CEO, R&D Director, Produc- lated the right way, is that thing right? I had a look
tion Director, Sales & Marketing Director, Finance Director, at [Alpha], and saw that they had allocated as much
HR Director). This authorisation was subject to meeting test time in assembly as in the old days. But that is
the KPIs and other less specied criteria related to the completely wrong, if our development of a test plat-
market and technology. form is to make any sense. So they had to go. And

Start Concept Design End System Design Market Release

Profitability Evaluation Start Product Creation Sales Release

Proof of Concept System Module Product Product & Follow- Respon-


Principle Design Design Creation Creation Production up sibility
(POP) Project Project Project Project Maturation Project transferred to
production

Transfer to Production Meeting

Team Greenhouse Team Module Creation Team Product Creation Production

Fig. 5. Divisions stage gate model.


198 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

evaluation of costs: are they really that expensive ferent concerns by different parties was important at any
considering what we know from earlier instru- stage during the NPD project to prevent development from
ments? Sales prices are just as interesting: have they moving into a wrong direction. In turn, it also fostered pro-
come up with unrealistically high prices or have they ject managers concerns with integrating these different
identied a realistic market level, or what is it? Then perspectives. As we will show in the next section, this is re-
there is the question of features can we really sell ected in the everyday practice through which project
this instrument with all these features or with only managers accounted for the progress of their project be-
those features? There is the positioning of the prod- tween the gates.
uct: what variants to market and what features
should they have relative to each other, and how Accounting during the stages of a project
should we price these features to cover the cost of Between the gates, technical issues about how to design
a reasonable market window? And then of course and engineer the products were at the centre of interest.
the project plan: is it realistic? Production Director. This does not mean, however, that the nancial conse-
quences of technical solutions were ignored. The fact that
When evaluating the projects at the gates, management
the project manager had to hand over the project at the
started out with the calculations as they were presented
next gate, and thereby present the report including key
to them by the project managers. In this respect, they man-
performance indicators to the board, impacted the engi-
aged by the numbers. However, they did not blindly rely on
neers and project managers sense-making during the
these calculations. They went beyond the numbers by crit-
stages. Although engineers were primarily concerned with
ically examining some of the assumptions behind the cal-
how to solve technical problems, they would try to relate
culations and by evaluating the project managers general
technical solutions to questions of revenues and costs. Pro-
understanding of the project. As part of this, they sought
ject managers, in turn, were particularly aware of the
to ensure that practices of information exchange and coor-
needs to make accounting numbers meet the specied
dination were in place, rather than trying to specify exactly
accounting criteria at the gates and to defend their project
how each and every practice should be carried out. Within
in front of the management board. Their job was mainly to
an organisation, individuals coexist through the intercon-
coordinate the work among the teams engineers as well as
nections among the practices that they carry out (Schatzki,
between the team and the other departments.
1996, p. 198). What is important for coordination is there-
fore to identify the relations among practices by virtue of In our department, Team Greenhouse, it is all about
which they form nexuses (Ibid., p. 199). In a sense, this is not controlling too much, but rather point out direc-
what the management board tried to insist on: tions in order not to restrict innovation and creativ-
ity. But then, when ideas start coming, we try to
The Management Board already knows the project
manage them in a meaningful direction. We do not
from earlier presentations. Team Greenhouse pre-
draw up a tight schedule from the start, but let it
sented the general idea to them this is what we
develop through a number of stages. Of course there
are going to develop so of course that is not up for
is a budget and a rough time estimate and some
discussion. Team Module Creation has developed
high-level targets for what we want to achieve, but
the modules, so that is also in place. So based on their
at this point in time we do not know what will work
work we check that the world has not changed too
which is why there must be some room for choosing
much. So at the Management Board meeting we start
new directions as we go along. Therefore, being a
out by agreeing that what we set out to do is in line
project manager is not so much about controlling. I
with what we are actually doing and also that this
mean, of course you need to control, but it is more
is what we still want to do. We have a much better
like coordinating what is going on by being in dia-
idea about the costs how much to expect for screws,
logue with all the parties involved in the project.
tools, packaging materials, etc. So we make an esti-
Project manager, Team Greenhouse.
mate of the total product costs for all product variants
and also an estimate about the resource costs not yet This suggests that NPD practices were organised in a non-
spent on the project. This is combined with the totalitarian manner (Schatzki, 1996, p. 124), i.e. that the
updated sales expectations from marketing in order understandings, rules, and teleoaffective structures within
to give a full view. [. . .] The Management Board is the practices were dened in such a way that they al-
particularly focused on the uncertainties in the pro- lowed for several alternatives of how to carry out the
jects in terms of costs, timing and sales budget. And practices. As explained above, during the meetings at the
of course they ask about the launch plans, and if the gates, the management board would challenge project
service organisation is ready for these new products. managers general understanding of the project in an at-
Project Manager, Team Product Creation. tempt to determine whether important issues from later
stages were properly reected in the accounting numbers
The questions were thus whether the project manager
presented in the reports. Project managers were thus ea-
linked the current development stage to the other prac-
ger to try and link to the perspectives of other
tices that the project would go through, and whether he
departments:
or she had a sufcient understanding of the rules and
objectives that would drive these other practices. This sort So you talk a lot, go to a lot of meetings and walk
of inquiry was based on the idea that an integration of dif- around to be on top of things and to nd out what
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 199

people are working with. Some of it is related to that engineers understood the eventual uses of the product
hardcore development issues, you know engineering they were developing, as engineers needed to get inside
stuff: does this technology work, what are the the customers practices to become acquainted with the ar-
parameters and things like that how do we develop ray of activities constituting these practices and their men-
this new instrument. But a large part of it is also to tal organisation.
make sure that what we do is in line with marketing, While being in favour of the development engineers
service and production. You need to have your rela- venturing into the eld, market analysts usually warned
tions in place and to call on the right people at the against using the impressions from such isolated events as
right time. Project manager, Team Greenhouse. generally valid market information:
Similarly, the R&D strategist explains that development You have to be careful, because then they made a
engineers in the Team Module Creation should adopt a few eld trips with the engineers, which I think is
wider perspective and be concerned not only with the a very positive thing, but when they had seen three
specications that they received from Team Greenhouse and now I am being a bit provocative but if they
but also with the wider implications of realising these had seen three, four or maybe ve producers, they
specications. thought they knew the business. That is dangerous
because they didnt, right? Sometimes we had argu-
For example, I heard about a power supply dimen-
ments like the customers want this and that, and
sioned for an [Alpha] variant. . ..that did not take into
then Id say: Yes, in the specic situation you expe-
account that the power input needed in [another
rienced in Spain, but. . .. Market Analyst, Business
instrument] was higher. It did not say explicitly in
Development.
the specication documents that this power supply
should account for both instruments, but it did say Market analysts, in turn, evaluated the market not on
that the project was supposed to develop both these the basis of one type of data but a combination of informa-
instruments. So we want the Team Module Creation tion from various sources:
engineers to take a wider approach, but not too
I do my market evaluations based on how many pro-
wide, of course it is the responsibility of the Team
ducers are there, how big are they, where are they
Greenhouse architect to develop the right specica-
located geographically and so on. So that is like hard
tion. R&D Strategist.
statistics that I can access on the Internet or buy
The way in which the integration of different perspec- from databases. [. . .] I know that we are not going
tives was achieved in practice can be illustrated with to sell to all of them: on the Internet I cannot nd
the example of Beta. Project Beta presented a challenge information on how interesting our idea is and
to the engineering teams in that it targeted a new type how much value it creates for the customers. There-
of customer segment. Beta was a low-cost product that fore I will call three or four of the most important
differed in its specication from the high-quality prod- markets and talk to our sales people: Listen, have
ucts on which Company had exclusively concentrated in any of your customers ever mentioned this? And
the past. In order to understand the users needs in terms he will respond: Yes, as a matter of fact they have.
of functionality and costs, development engineers were And then I will ask him to dig a bit deeper to nd out
sent on eld trips. There, they sought to understand what how much they would be willing to pay for such a
type of user they were up against (e.g. chemist or farm- solution, how they would use it, and how many of
er), how the users would handle the instruments (e.g. this type of customer he has, and then get back to
transportable or stationary), the environment in which me. And I also know consultants, major producers,
the future instruments would be used (e.g. levels of dust university people and others like that, and I choose
and humidity) and similar issues. You simply have to my sparring partner depending on the type of task
see it, you need to get your hands dirty; there is no at hand and its condentiality. Market Analyst, Busi-
way around it (Senior Manager, Team Greenhouse). Send- ness Development.
ing development engineers on eld trips was believed to
The knowledge acquired during these activities was then
be important for making the right design and engineering
translated into sales volumes, price points, instrument
decisions:
specications, etc., which was then used as inputs in the
It works much better when they have been out there NPD project in question. The result was a combination of
and have met these farmers that only speak Italian numbers and text through which it seemed that Division
and how do you think they are going to use an effectively bridged the latent conict between the quanti-
instrument with a lot of fancy functionalities? It is ed customer and the sales customer (Vaivio, 1999). Thus,
very effective. We are much more certain about engineers on eld trips acquired valuable knowledge about
what we are doing after having visited a producer, the particular use of the future product but not about the
whereas before where we fumbled a bit. Project entire market; conversely, the Market Analyst collected
Manager, Team Greenhouse. information about the market, but not about the product
and its use as such. However, when these understandings
The eld trips fullled a similar function as the internal were combined and reproduced in the product develop-
coordination with other departments that would take ment report (to be eventually presented at the gates) as
place later on in the project development. They ensured technical specications and accounting numbers, a strong
200 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

case emerged, the validity of which was discussed at pro- Exactly how this mesh of rules, ends and values was
ject meetings. At such meetings, negotiations about the accomplished was only partially explicable by the project
customers, technology, software, time plans, quality issues, manager and development engineers, underlining Schatz-
etc. were conducted in relation to the overall strategy, the kis (1996, p. 128) comment that the limits of intelligi-
strategy of the future product, and ultimately economic bility are broader than those of language. Certainly, it
elements of the business case, e.g. cannibalisation of exist- did not help that the physical sites of this practice were
ing products, sales volumes and contribution ratios. This dispersed in time and space: like in Prestons (1986)
required some efforts of translation. Knowledge about the arrangements of informing, development engineers met
sales customer was, for example, translated into the colleagues at the coffee machine or they incidentally
language of the quantied customer, which most organi- bumped into each other at lunch and used the opportu-
sational members could more easily relate to. Thus, when nity to exchange ideas and receive immediate feedback.
the sales people argued for a specic feature which they Seemingly, the textbook version of target cost manage-
believed was important for the sales customer, its contribu- ment refers to a much more clear-cut or totalitarian
tion was evaluated in relation to the quantied customer. practice compared to the actual activities involved in this
The process was reversed when sales people believed that process, as was found in Division (see also Mouritsen,
ideas borne by the quantied customer would harm the Hansen, & Hansen, 2001).
sales customer. However, since the language of the sales While, in this way, accountabilities were enacted in a
customer was less common compared to that of the rather informal manner and were shared among different
quantied customer, it required some perseverance to actors, there was at the same time some need for a more
make the case. Essentially, it required that the sales people formal denition and separation of accountabilities. This
were able to explain the customer practice to the other need was mainly created by the stage-gate-process de-
organisational members and thereby enable them to speak scribed above. Before our eldwork started, Company had
(or at least temporarily understand) the language of the introduced a change to this stage-gate-process. In the past,
sales customer. The exchange of perspectives and ratio- the same team of engineers would stay with a project
nales for action that could be observed in these meetings throughout all its stages. While this allowed the project
constituted these actors practice of accounting. Their local team to become owners of a particular project, it also
information allowed them to reduce uncertainty towards meant that engineers who would specialise in a particular
particular goals and discuss the trade-offs between differ- type of task had to cover a variety of different engineering
ent goals. activities, not all of them being of equal interest to every-
Back in the workshop, development engineers had to one. Moreover, when a project came to a gate and had to
come up with technological solutions that would meet be justied in terms of its viability, it was not always yet
the various specications that had been agreed on: the ready for the next stage. But since the same project team
target cost found as the expected sales price dened would continue with the project, it became an accepted
by Business Development, less the contribution ratio practice to have a certain overrun and to nish some of
required by top management. This solution had to be the activities later than initially planned. Mainly for those
achievable in a way that did not violate the other KPIs, two reasons, management reorganised the stage-gate pro-
meaning that ve years contribution margin had to cover cess by splitting each project into three main phases and
development costs by a factor X, nor could it violate the allocating three different teams to these phases. Not only
strategy and value statements of Division. The process of did this allow for more specialisation on the part of engi-
meeting target costs was not the outcome of a detailed neers; it also forced project managers to full all their du-
breakdown of functional requirements matched with ties before handing over the project to the next team. This
allowable costs. Trying to tease out this process revealed new practice therefore created a more clear-cut account-
a subtle array of activities organised by a mental causality ability of each project team and, given the sequential inter-
held by the engineers of what was necessary to know and dependencies of tasks, each project team had an interest in
do to move on given the situated case of technology, enforcement of the accountability of the others.
customer and strategy: In the case of Beta, this became obvious at the point
where Team Greenhouse had to hand over the project to
Eh, yes, well I did not really do it that way. I found the Team Module Creation. Team Greenhouse came up
out what they [the parts] were probably going to with the design of the architecture of the instrument and
cost. I mean, of course we have tried to split it up a outlined the various functionalities together with cost esti-
bit, but, really, I mean, [a key part] is not from our mates in the project report. While the project report thus
project team. Its from [one of our development engi- contained a representation of the project in accounting
neers]. And then there were some evaluations of terms, the receiving team was not entirely satised with
that. And I have tried to determine the manufactura- this representation. The numbers contained in the report
bility of his idea, and does it include the features were too aggregate for Team Module Creation to base their
necessary to make it functionally adequate. Sort of actions on. They requested a more detailed representation
backwards, I suppose you could say. [. . .] It is not like to see through the numbers and judge the viability of
I had a budget for a lot of functionalities, it is more realising the required functionalities within the indicated
sort of. . .here is the most boiled down solution we overall target costs. The receiving team simply did not
can imagine does it comply with the KPIs? Devel- want to run the risk of accepting unachievable target costs
opment Engineer, Team Greenhouse. and ultimately threatened to use its right not to sanction
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 201

the projects transfer to the next stage. That this became an Discussion
issue was a result of Alpha showing a signicant budget
overrun, which caused the manager from Team Module It is hard to say whether Division was more oriented to-
Creation and his people to feel the Senior Manager from wards nancial or strategic goals and whether it was ulti-
Team Product Creation and the MB breathing down their mately putting more focus on operational efciency or
necks. Team Module Creation did not want to end up being strategic effectiveness. However, the fact that there were
blamed again for something over which they had no con- several, at times competing, strategic objectives (a fact that
trol, as they felt had been the case with Alpha. the Ranking Report had criticised) indicates that there was
There was also another issue involved here, for besides at least no focus on one particular strategy. As a conse-
criticising the overall lack of detail in the documentation of quence of the variety of strategic objectives associated
costs in the project report, the receiving team also com- with modularity, there was considerable leeway regarding
plained that the report did not include any indirect vari- the interpretation and evaluation of NPD practices. NPD
able production costs (IPC). These costs became visible practices were in this respect less totalitarian than the
only once the Team Module Creation entered the bill of operational practices in Ahrens and Chapmans (2007) res-
materials (i.e. the list with all required materials) into taurant chain. Distinct practices in time and space clearly
the ERP system. The ERP system automatically adds a cer- existed and ensured continuity and order (Schatzki,
tain percentage to the direct variable costs in order to ac- 1996), but the way in which these practices were carried
count for these indirect costs, which are not entered out was subject to variation. Control of these practices
separately for a particular instrument. Apparently, Team did not work through specication of each and every de-
Greenhouse engineers were unaware of this automatic tail, but rather through high level targets and coordination.
allocation procedure and did not include it when preparing As a consequence, engineering and design choices could be
their cost estimates. For Team Module Creation, this meant accounted for and justied from different points of view
an additional cost item which increased actual costs and and this is in fact what happened in Division.
made it more difcult to stay within the target costs elab- But while there was competition between different
orated by Team Greenhouse. When this issue was men- strategic concerns, there was no open conict as evidenced
tioned at a project meeting with participants from all in Mouritsen (1999) or Vaivio (1999). Different strategic
relevant departments and teams, a task force was ap- objectives were also regarded as complementary in the
pointed to solve this problem. The result of the discussion sense that each of them was supposed to contribute in
was that the development engineer from Team Green- its own way to overall protability. It was clear that there
house who was responsible for outlining the target costs were trade-offs and that not all benets of modularity
would, from now on, have to include indirect production could be realised to the same extent. Ultimately, however,
costs in his cost estimate. there were no substantial political agendas to promote
The need to hand over the project at the gates reminded particular strategic objectives as ends in themselves or,
engineers and project managers that numbers matter at least, we did not observe any conicts that would have
and that representations of the status quo of a project stemmed from such agendas; rather, strategic concerns
had to be expressed in a way that allowed the separation were regarded as means to protability that had to be bal-
of accountabilities, but not always was everyone satised anced in a reasonable way. While the different strategic
with the available methods to accomplish this. Limits of objectives entered NPD practices as part of their teleoaffec-
calculability were acknowledged and better solutions tive structures (Schatzki, 2002), they were not the only
looked for, as the following two quotes illustrate: means through which the appropriateness of these prac-
tices was assessed. The overall concern with protability
A project manager in Team Greenhouse is not the
constituted the general understanding (Schatzki, 2002) of
best person to make profound nancial evaluations
functionality across the different practices and parties. It
in order to determine if a module should be of a cer-
was in the name of protability that strategic objectives
tain size or be produced of this or that component. It
had to be negotiated.
would be much better if we had someone with the
The stage-gate process was critical in this respect, as it
right skills that could focus on that. I mean, I can
created a formal structure of accountability which re-
throw numbers into a spreadsheet, but we really
minded project managers and engineers of the importance
need more than that. Project manager, Team
of the overall goal of protability. This created an incentive
Greenhouse.
to think about accounting numbers when debating design
I really need a tool to evaluate the nancial benet of choices and to discuss the trade-offs between strategic
reuse. What is the cost implication in the production objectives also in terms of their nancial impact, even if
department of having one product structure com- this was possible only to a limited extent. But the imper-
pared to another? I also miss a quantication of lead fect representational qualities of accounting were not con-
time implications what are the costs and benets tested, because strategic objectives were not translated into
of one week more or less? I think we know our var- numbers; they were rather mobilised in tandem with
iable costs, but what about all the activity-related nancial arguments. Accounting inscriptions were not re-
costs? I am unable to make a fact-driven argument garded as the single best way of representing what hap-
about all the cost types that are not direct. Project pens in the organisation or even as a road to
Coordinator, Production Department. commercial salvation, as in the case studied by Ezzamel,
202 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

Lilley, and Willmott (2004). Managers in Division did not there were no spokespersons (such as management
want to rely entirely on knowledge that was disembodied accountants) who would, on a continuous basis, try to en-
in inscriptions (Ibid., p. 808). This also became clear when act this form of accountability. However, the need to pro-
projects were to be evaluated at the gates, insofar as such vide accounting numbers at the gates motivated project
evaluation was not exclusively guided by nancial motives managers and others to enact nancial accountability also
either. On the one hand, the management board drew from during the stages.
their own practical expertise in order to see through This case shows that even in the absence of a strong
numbers that featured in the project reports. In this re- reliance on accounting numbers, a drifting of practices
spect, their inuence on project managers did not exclu- (Quattrone & Hopper, 2001) can be avoided because other
sively or primarily build upon the particular quality of forms of accountability are mobilised to ensure a coordina-
accounting inscriptions as mechanisms of action at a dis- tion of NPD practices. In our case, control assumed the
tance (Latour, 1987; Robson, 1992). On the other hand, form of horizontal information sharing, guided by strategic
the management board used the meetings at the gate to objectives and a general understanding of the need to be
ensure that there was local communication and coordina- protable. Interestingly, our case also shows that some
tion. Top managements control from the top was not sup- attention to accounting information is likely to be impor-
posed to replace the more hands-on practice of control in tant in uncertain and ambiguous settings such as NPD.
NPD, as in the case of BusinessPrint (Mouritsen, 1999); it The enforcement of nancial accountability at the gates
rather tried to promote it. In a way, this made sense given can be seen as a means to remind managers of the impor-
the plurality of strategic messages that modularisation tance of accounting numbers even if these are only
stood for. While the contribution that modularisation imperfect representations of the decisions and practices
could make to each of these strategic goals was clear in in NPD. Accordingly, accounting information was shown
principle, it was less evident how this could be achieved to have a rather indirect effect on organisational practices
in practice. Only by drawing upon diverse bodies of local in Division. It was not the case that different rationalities
knowledge and ongoing experiences in product develop- were translated into accounting; rather, accounting infor-
ment was it possible to understand the relationship be- mation was selectively enacted to amend the evaluation
tween module design and each of the strategic objectives, of these rationalities.
and only then was it possible to discuss how these objec- To be sure, the combination of the more formal, nan-
tives should be traded off against each other. In other cial accountability at the gates and the rather informal
words, the strategising process was both about uncertainty control mechanisms during the stages should not be read
reduction and goal congruence, and accounting informa- as neither an optimal nor a suboptimal solution. We
tion was mobilised in the name of both (see Davila, suggest simply seeing it as the companys reaction to the
2000). In pluralistic settings, these two functions will often particular setting in place. There were no management
intermingle, since the selection of a goal is at the same accountants who could have developed more sophisticated
time a reduction of uncertainty at the local level, and since costing systems that would have allowed translating the
local information is often needed to understand what dif- engineers logic more easily into an accounting logic. Per-
ferent goals actually mean and how they should be haps this was good, because it meant that engineers had
traded-off against each other. This is also in line with Ah- to come together and coordinate their work in a bottom-
rens and Chapmans (2007, p. 23) observation that means up way. Perhaps it was suboptimal, because things could
and ends are constructed simultaneously in practice (see have been done more efciently with a more sophisticated
also Lave, 1988). In contrast to their ndings from the res- formal structure in place. The fact that not everyone was
taurant chain, however, our case suggests a more pro- entirely happy with the amount and quality of nancial
nounced bottom-up contribution to the strategising information available was due to strategic imperatives
process. Local knowledge seems to play a stronger role being partly conicting and in need of trade-off. More
for the crafting of practices in our case of NPD than it does knowledge of the costs and benets of different design
in the commercial setting of Ahrens and Chapmans restau- choices could have helped make these trade-off decisions;
rant chain. This, we suppose, is linked to the high degrees however, it could also have created resistance against too
of uncertainty and complexity in our case or, rather, to the strong a quantication of what is regarded as a question
way in which organisational actors enacted these particu- of qualitative judgement.
lar circumstances. Whether optimal or not, the inuence of accounting
At the gates, nancial accountability was created information can only be understood when considered in
through a set of rules (formal criteria) dened by top man- interaction with other types of accounts or rationalities,
agement. Between the gates, it was more like a general such as strategic objectives. This has some important impli-
understanding that orchestrated behaviour. Strategic cations for research with which we conclude the paper.
objectives constituted part of the teleoaffective structure
of the different practices in NPD. The accountability toward Conclusion
them was enacted through the fact that different actors
quite naturally drew upon these objectives in the discus- Accounting information is enacted as part of an account-
sions, building upon the practical understanding of the ability relation (Ahrens, 1996). Accountabilities can be di-
practices they were familiar with (e.g. customer-related verse, especially in pluralistic settings (Denis et al., 2007;
practices). In contrast, this was not really the case for Sinclair, 1995). As a consequence, the role and relevance
nancial accountability. Apart from top management, of accounting information becomes intelligible only against
B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204 203

the background of its interaction with other types of behavior and personnel controls. Accounting, Organizations and
Society, 22(34), 233248.
accounts. In a sense, this is in line with contingency ap-
Ahrens, T. (1996). Styles of accountability. Accounting, Organizations and
proaches to NPD, which have tried to link the relevance of Society, 21(23), 139173.
accounting information to different types of strategy and Ahrens, T., & Chapman, C. (2004). Accounting for exibility and efciency:
uncertainty (Davila, 2000). However, as our case shows, it A eld study of management control systems in a restaurant chain.
Contemporary Accounting Research, 21(2), 271301.
may not (or at least not only) be the content of strategy that Ahrens, T., & Chapman, C. S. (2006). Doing qualitative research in
mediates the usefulness of accounting information; it may management accounting: Positioning data to contribute to theory.
also be the way in which, and the extent to which, strategic Accounting, Organizations and Society, 31, 819841.
Ahrens, T., & Chapman, C. (2007). Management accounting as practice.
ideas are mobilised that co-determine the relative impor- Accounting, Organizations and Society, 32(12), 127.
tance of accounting numbers. This can be an important Ahrens, T., & Dent, J. F. (1998). Accounting and organizations: Realizing
message for contingency approaches to the topic, which the richness of eld research. Journal of Management Accounting
Research, 10, 139.
tend to take strategy as a black box (Chua, 2007) rather Baxter, J., & Chua, W. F. (2003). Alternative management accounting
than considering the strategising efforts of actors. research whence and whither. Accounting, Organizations and Society,
As illustrated in this paper, practice theory offers a use- 28(23), 97126.
Blumer, H. (1969). Symbolic interactionism: Perspective and method.
ful vocabulary to illuminate these processes of strategising Berkeley: University of California Press.
and the way they relate to the use of accounting informa- Burchell, S., Clubb, C., Hopwood, A., Hughes, J., & Nahapiet, J. (1980). The
tion. While our paper has brought to the fore some of the roles of accounting in organizations and society. Accounting,
Organizations and Society, 5(1), 527.
complexities and nuances in the relationship between
Chapman, C. S. (2005). Controlling strategy: Management, accounting, and
accounting and strategising, there is clearly scope for more performance measurement. Oxford: Oxford University Press.
research in this area. It seems especially worthwhile to Chua, W. F. (2007). Accounting, measuring, reporting and strategizing
examine in more detail how accounting tools and concepts re-using verbs: A review essay. Accounting, Organizations and Society,
32, 484494.
contribute to the process of crafting strategies and to Cooper, R. G. (1999). Product leadership: Pathways to protable innovation.
which extent strategies are pre-congured by the exis- New York: Basic Books.
tence or non-existence of such tools. Similarly, the possible Czarniawska, B. (1999). Writing management: Organization theory as a
literary genre. Oxford: Oxford University Press.
competition between accounting representations and Davila, A. (2000). An empirical study on the drivers of management
strategic ideas is a promising area of research. Since it is control systems design in new product development. Accounting,
not always possible to unambiguously translate strategic Organizations and Society, 25(45), 383409.
Denis, J.-L., Langley, A., & Rouleau, L. (2007). Strategizing in pluralistic
objectives (relating, for example, to quality or customer contexts: Rethinking theoretical frames. Human Relations, 60(1),
satisfaction) into the language of accounting, the question 179215.
arises to which extent organisational actors refer to either Dent, J. F. (1991). Accounting and organizational cultures: A eld study of
the emergence of a new organizational reality. Accounting,
of them when evaluating alternative courses of action. Organizations and Society, 16(8), 705732.
Accounting is a powerful practice, but it is at the same Dyer, W. G., & Wilkins, A. L. (1991). Better stories, not better constructs, to
time only one among many activities within an organisa- generate better theory: A rejoinder to Eisenhardt. Academy of
Management Review, 16(3), 613619.
tion. To study accounting in the contexts in which operates
Ezzamel, M., Lilley, S., & Willmott, H. (2004). Accounting representation
(Hopwood, 1983) implies examining the ways in which and the road to commercial salvation. Accounting, Organizations and
accounting relates to, and intersects with, other types of Society, 29(8), 783813.
activities, such as strategising or marketing. An increased Flyvbjerg, B. (2001). Making social science matter. Cambridge: Cambridge
University Press.
examination of such intersections would not only be Garnkel, H. (1967). Studies in ethnomethodology. Cambridge: Polity Press.
worthwhile for accounting researchers but is also likely Giddens, A. (1979). Central problems in social theory. Action structure and
to increase the interest of other disciplines in the outputs contradiction in social analysis. Berkeley and Los Angeles: University of
California Press.
of accounting research. Giddens, A. (1984). The constitution of society. Cambridge: Polity Press.
Henri, J.-F. (2006). Management control systems and strategy: A resource-
based perspective. Accounting, Organizations and Society, 31(6),
Acknowledgements 529588.
Hopwood, A. G. (1983). On trying to study accounting in the contexts in
which it operates. Accounting, Organizations and Society, 8, 287305.
Comments from two anonymous reviewers as well as
Hopwood, A. G., & Miller, P. (Eds.). (1994). Accounting as social and
from Chris Chapman, David Cooper, Fiona Anderson-Gough, institutional practice. Cambridge: Cambridge University Press.
Kari Lukka, Martin Carlsson, Morten Jakobsen, Niels De- Jarzabkowski, P. (2004). Strategy as practice: Recursiveness, adaptation
chow, Paolo Quattrone, Taka Fujioka, and the participants and practices-in-use. Organization Studies, 25(4), 529560.
Jarzabkowski, P., Balogun, J., & Seidl, D. (2007). Strategizing: The
at the workshop on Field Work in Accounting Research at challenges of a practice perspective. Human Relations, 60(1), 527.
Warwick Business School (June 2006) have helped us im- Jnsson, S. (2004). Product development: Work for premium values.
prove this paper. Acknowledgements are also due to the Copenhagen: Liber and Copenhagen Business School Press.
Jnsson, S., & Macintosh, N. (1997). CATS, RATS, and EARS: Making the
accounting departments at Turku School of Economics case for ethnographic accounting research. Accounting, Organizations
and Business Administration and Sad Business School, Uni- and Society, 22(34), 367386.
versity of Oxford, for hosting the visits of one researcher Kurunmki, L. (2004). A hybrid profession the acquisition of
management accounting expertise by medical professionals.
during which parts of the paper were developed. Accounting, Organizations and Society, 29(34), 327347.
Langeld-Smith, K. (1997). Management control systems and strategy:
A critical review. Accounting, Organizations and Society, 22(2),
References 207232.
Latour, B. (1987). Science in action. Milton Keynes: Open University Press.
Abernethy, M. A., & Brownell, P. (1997). Management control systems in Latour, B. (2005). Reassembling the social. An introduction to actor-network-
research and development organizations: The role of accounting, theory. Oxford: Oxford University Press.
204 B. Jrgensen, M. Messner / Accounting, Organizations and Society 35 (2010) 184204

Lave, J. (1988). Cognition in practice. Cambridge: Cambridge University Robson, K. (1992). Accounting numbers as inscription: Action at a
Press. distance and the development of accounting. Accounting,
Llewellyn, S. (1999). Narratives in accounting and management research. Organizations and Society, 17(7), 685708.
Accounting, Auditing and Accountability Journal, 12(2), 220236. Rockness, H. O., & Shields, M. D. (1984). Organizational control systems in
Lounsbury, M. (2008). Institutional rationality and practice variation: research and development. Accounting, Organizations and Society, 9(2),
New directions in the institutional analysis of practice. Accounting, 165177.
Organizations and Society, 33(45), 349361. Schatzki, T. R. (1996). Social practices: A Wittgensteinian approach to
Lukka, K., & Kasanen, E. (1995). The problem of generalizability: human activity and the social. Cambridge: Cambridge University
Anecdotes and evidence in accounting research. Accounting, Auditing Press.
& Accountability Journal, 8(5), 7190. Schatzki, T. R. (2001). Introduction. Practice theory. In T. R. Schatzki, K.
March, J. G., & Olsen, J. P. (1989). Rediscovering institutions: The Knorr Cetina, & E. von Savigny (Eds.), The practice turn in contemporary
organizational basis of politics. New York: Free Press. theory (pp. 114). London and New York: Routledge.
Miller, P. (1994). Accounting and objectivity: The invention of calculating Schatzki, T. R. (2002). The site of the social: A philosophical account of the
selves and calculable spaces. In A. Megill (Ed.), Rethinking objectivity constitution of social life and change. University Park: Pennsylvania
(pp. 239264). London: Duke University Press. State University Press.
Mouritsen, J. (1999). The exible rm: Strategies for a subcontractors Schatzki, T. R. (2005). The sites of organizations. Organization Studies,
management control. Accounting, Organizations and Society, 24(1), 26(3), 465484.
3155. Schatzki, T. R., Knorr Cetina, K., & von Savigny, E. (2001). The practice turn
Mouritsen, J., Hansen, A., & Hansen, C. O. (2001). Inter-organizational in contemporary theory. London and New York: Routledge.
controls and organizational competencies: Episodes around target Sinclair, A. (1995). The chameleon of accountability: Forms and
cost management/functional analysis and open book accounting. discourses. Accounting, Organizations and Society, 20(23), 219
Management Accounting Research, 12, 221244. 237.
Orlikowski, W. J. (2007). Sociomaterial practices: Exploring technology at Strauss, A., & Corbin, J. (1998). Basics of qualitative research. London: Sage.
work. Organization Studies, 28(9), 14351448. Thvenot, L. (2001). Organized complexity: Conventions of coordination
Preston, A. (1986). Interactions and arrangements in the process of and the composition of economic arrangements. European Journal of
informing. Accounting, Organizations and Society, 11(6), 521540. Social Theory, 4(4), 405425.
Quattrone, P., & Hopper, T. (2001). What does organizational change Tomkins, C., & Groves, R. (1983). The everyday accountant and
mean? Speculations on a taken for granted category. Management researching his reality. Accounting, Organizations and Society, 8(4),
Accounting Research, 12, 403435. 361374.
Reckwitz, A. (2002). Toward a theory of social practices. A development in Vaivio, J. (1999). Examining The Quantied Customer. Accounting,
culturalist theorizing. European Journal of Social Theory, 5(2), 243263. Organizations and Society, 24(8), 689715.
Roberts, J. (1990). Strategy and accounting in a U.K. conglomerate. Whittington, R. (2006). Completing the practice turn in strategy research.
Accounting, Organizations and Society, 15(12), 107126. Organization Studies, 27(5), 613634.

S-ar putea să vă placă și