DISCOURSE REVISITED: DIMENSIONS AND ......SOTIRIOS PAROUTIS* and LOIZOS HERACLEOUS Warwick Business...

Strategic Management JournalStrat. Mgmt. J., 34: 935–956 (2013)

Published online EarlyView 6 March 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.2052

Received 2 June 2010 ; Final revision received 22 May 2012

DISCOURSE REVISITED: DIMENSIONS ANDEMPLOYMENT OF FIRST-ORDER STRATEGYDISCOURSE DURING INSTITUTIONAL ADOPTIONSOTIRIOS PAROUTIS* and LOIZOS HERACLEOUSWarwick Business School, University of Warwick, Coventry, U.K.

Despite decades of research on strategy, we still know little about what the concept of strategymeans to actual strategists and how they use it in practice. Working at the intersectionsof institutional and practice theories, we use exploratory interviews with strategy directorsand a longitudinal case study to uncover four dimensions of first-order strategy discourse:functional, contextual, identity, and metaphorical. We also reveal three phases in the interrelationbetween first-order strategy discourse and institutional work: shaping, settling, and selling anda differential emphasis (selective focusing) on dimensions of the first-order strategy discourseduring the institutional adoption process. We contribute to a deeper understanding of the conceptof strategy in practice, the process of institutional adoption, and of the role of discourse in thisprocess. Copyright 2013 John Wiley & Sons, Ltd.

INTRODUCTION

. . . if you presuppose language, you havealready presupposed institutions . . . Insteadof presupposing language and analyzinginstitutions, we have to analyze the role oflanguage in the constitution of institutions .J. R. Searle on what is an institution?(2005: 2)

We often debate the relevance of manage-ment theories to practitioners—yet let’s considerwhether we do enough to actually listen to them.This challenge is pertinent within strategic man-agement, which remains a fragmented and con-tested field (Bowman and Helfat, 2001; Hambrick,2004). Schendel (2006) poignantly notes about ourfield: “Ask yourself whether we are doing enoughwith application, with actually using ideas in

Keywords: discourse; institutional theory; strategicchange; strategy-as-practice; multibusiness firms*Correspondence to: Sotirios Paroutis, Warwick BusinessSchool, University of Warwick, Coventry CV4 7AL, UK.E-mail:[email protected]

Copyright 2013 John Wiley & Sons, Ltd.

practice” (2006: iv). However, while we exam-ine what the strategy concept means for our field(Nag, Hambrick, and Chen, 2007) and have callsurging us to explore strategists’ own conceptionsof strategy (Barry and Elmes, 1997), there is stilla gap in our understanding of what the strategyconcept means to strategy practitioners (or “strate-gists”) and how this might be employed in contextssuch as the process of institutional adoption. Thisis surprising since, as we will argue next, in addi-tion to the inherent advantages of understandingstrategists’ own perceptions, the concept of strat-egy has the potential to help us explain how therelation between managerial cognitions and man-agerial practices leads to organizational outcomesin the established strategic change literature. Toour knowledge, this is the first study that aims toidentify first-order meanings of strategy and howstrategists employ these meanings. By first-order,we refer to what strategists themselves mean bythe term “strategy”; as opposed to “second-order”meanings of what strategy might mean, as heldby others (Zahavi, 2002). As such, we apply theterm “first-order” in its phenomenological sense,

936 S. Paroutis and L. Heracleous

to denote meanings in use and how these mean-ings may change over time (Sandberg and Tsoukas,2011). As we will show, the way agents employfirst-order meanings has important implicationsfor institutional adoption. In theoretical terms, ourunderstanding of strategic discourse has advancedwith the emergence of the strategy-as-practiceperspective and neoinstitutional debates, particu-larly with reference to the notion of institutionalentrepreneurs. The strategy-as-practice approachviews strategy “as a socially accomplished, situ-ated activity arising from the actions and inter-actions of multiple level actors” (Jarzabkowski,2005: 6) and considers strategy not only as some-thing an organization has but something that itsmembers “do” (Jarzabkowski, Balogun, and Seidl,2007; Johnson, Melin, and Whittington, 2003;Vaara and Whittington, 2012; Whittington, 2007).Despite advancements in strategy-as-practice, ourunderstanding of the meanings of strategy as per-ceived by organizational actors “in practice” is stillfairly limited. Our study extends this approach byemploying a discursive lens to understand strat-egy practices back to a point more primary andfoundational than the strategy process, back to thevery concept of strategy as understood by prac-titioners. Such an investigation is important sincelanguage is a constituting element of daily practiceand strategy making (Tsoukas, 2010).

Institutional entrepreneurship involves the“activities of actors who have interest in partic-ular institutional arrangements and who leverageresources to create new institutions or to transformexisting ones” (Maguire, Hardy, and Lawrence,2004: 657) and highlights the ways in whichactors work toward their strategic objectives bydeliberately leveraging resources (Garud, Jain,and Kumaraswamy, 2002). Recently, studies havestarted exploring the intersections of practice andinstitutions (Jarzabkowski, Matthiesen and Vande Ven, 2009; Zietsma and Lawrence, 2010).At these intersections, we focus our attentionon a phenomenon that has received no priorexplicit theoretical or empirical attention: theprocess of strategy meaning making and howstrategists employ these meanings when adoptingan institution. In order to shed light on thisprocess, we examine the dimensions of strategydiscourse employed by those strategists seeking toadopt the institution of a strategy support functionin a multibusiness firm. While such functionscan be perceived as advanced versions of the

planning departments of the past, their adoptionin the context of large firms has four distinctivecharacteristics: (1) they are constituted by small,flexible teams (Paroutis and Pettigrew, 2007)working as a support function, (2) they are ledby the chief strategy officer who has close linkswith the CEO (Breene et al ., 2007), (3) instead ofbeing centrally located and resourced, they tendto be decentralized (Grant, 2003), and (4) theirrole spans a lot more activities beyond producingstrategy reports, such as organizing workshopsand away days relating to strategic planning(Whittington and Cailluet, 2008). Our approachaims to further our understanding of the conceptof strategy as employed by practitioners. Weshow that first-order strategy discourse, which wedefine as the group of statements about strategyemployed by strategists, is characterized by certaininstitutionalized central themes and understand-ings that, as structural features (Heracleous andBarrett, 2001) of first-order strategy discourse, areboth constraining as well as enabling (Giddens,1984) the practice of strategy. Further, we showthat the emphasis on strategy discourses differsover different periods in the institutionalizationprocess, and discuss the insights that can begained from such deeper understanding of the roleof language in this process. We find that attentionto what strategists do needs to include attentionto what they say, as the first-order strategy dis-course at the microlevel influences not only howstrategy is talked about more broadly, but is alsolinked to how these meanings are employed atlevels beyond the micro, to facilitate institutionaladoption. Further, we show that performativeoutcomes (Jarzabkowski, Lê and Feldman, 2012;Kornberger and Clegg, 2011; Whittington, 2007)related to issues such as organizational capacityand strategist legitimacy are important featuresof the institutional adoption process. We beginby discussing studies in strategy-as-practice andinstitutional entrepreneurship, to showcase howdiscourse can link these lines of enquiry.

FIRST-ORDER STRATEGYDISCOURSE: LINKING MANAGERIALCOGNITION AND PRACTICE TOACHIEVE INSTITUTIONAL ADOPTION

As identified earlier, there is a gap in our under-standing of the way the term strategy is employed

Copyright 2013 John Wiley & Sons, Ltd. Strat. Mgmt. J., 34: 935–956 (2013)DOI: 10.1002/smj

First-Order Strategy Discourse During Institutional Adoption 937

in practice by those actors involved in the adoptionof institutions. Institutional adoption representsone of the ways firms utilize to better align withtheir external environment; in other words, one ofthe ways firms employ to achieve strategic change(Dacin, Goodstein, and Scott, 2002; Heracleousand Barrett, 2001; Kostova and Roth, 2002; Munirand Phillips, 2005). Strategic change has been atopic of much debate, with one stream of studiesconceptualizing strategic change as a shift inmanagerial cognitions and sensemaking processes(e.g., Fiss and Zajac, 2006; Gioia and Chittipeddi,1991). Alongside this viewpoint, a practice viewis also emerging. This approach focuses on theactivities of strategists through the strategy-as-practice perspective (Jarzabkowski et al ., 2007;Johnson et al ., 2003; Whittington, 2007) anddevelopments in practice theory (Feldman andOrlikowski, 2011). While these streams have, thusfar, been developing separately, we argue thatthey share two important common threads: first,their potential to help us explain how strategicdiscourse can lead to particular institutionaloutcomes and, second, their foundational rela-tionship with strategic discourse (Tsoukas, 2010).Building on the second thread, we conceptualizestrategic discourse at the microlevel, throughthe notion of first-order strategy discourse.While most prior studies have highlighted thesubstantive importance of strategic change fororganizational survival, the processes by whichstrategists conceive of and employ the conceptof strategy in the context of institutional adoptionhave rarely been studied. However, discourse isfundamental and constitutive of strategic change,and therefore understanding this dimension can beenlightening to our understanding of change itself(Oswick et al ., 2010). Overall, we propose thata more comprehensive approach to the microlevelactivities of actors during strategic change needsto start with an understanding of the dimensionsof the first-order discourse these actors employ.

We define discourse as a group of “texts,whether oral or written, located within socialand organizational contexts that are patterned bycertain structural, inter-textual features that haveboth functional and constructive effects on theircontexts” (Heracleous, 2006: 2). In this context,by “first-order strategy discourse” we mean talk(oral text) about strategy uttered by strategists thatreveals their own perceived meanings of the nature

of strategy. Our view of first-order strategy dis-course as both a resource and constraint for inter-pretation and action by institutional entrepreneurscan be seen in terms of what Alvesson andKärreman (2000) call a meso-discourse approach:going beyond the text with the aim of identifyingpatterns and interrelations between the text andits context. Our approach therefore assumes atightly coupled relationship between discourse andmeaning, as well as discourse and social practices,where discourse constitutes actors’ meanings andinfluences their practices. This view is consistentwith what Heracleous and Barrett (2001) refer toas the interpretive stream of discourse research.Discourse, in its constitutive role, is not only aresource but also a constraint for agents. Agentsdo not just use discourse as they see fit. Infollowing established social practices shaped bybroader, societal-level dominant discourses, agentssubconsciously reaffirm, or act discourse out;rather than agents employing discourse, discoursein this sense employs agents in order to sustainand perpetuate particular meaning systems andpower arrangements. Foucault-inspired (Foucault,1972, 1977) studies exemplify this understandingof discourse (Knights and Willmott, 1989).Rather than following this Foucauldian approach,however, given the interpretive paradigm withinwhich this work was conducted, we instead retainour focus on the meso-level, tightly coupled,interpretive approach outlined above.

In order to visually represent our argumentabout the role of first-order strategy discourse,we developed a conceptual framework (refer toFigure 1) based on the actor-level aspects ofthe strategic change framework by Rajagopalanand Spreitzer (1996). These aspects are: man-agerial cognitions, managerial practices, and theirlinks to the content of strategy. Figure 1 portrayshow institutional outcomes are created, main-tained, and disrupted through cognition and prac-tice, which are in turn shaped by discourse.As such, it reveals three key relationships ofimportance to our microlevel understanding ofstrategic change: cognition-discourse, discourse-practice, and discourse-outcomes, which we dis-cuss below. In light of our research questions, inthis paper we focus on the last two, and, combinedwith focal points in strategy-as-practice and insti-tutional theory, we pose two unresolved puzzlesregarding our understanding of strategic changeconcerning (1) the nature of meanings of strategy



Managerialcognition

Institutionaloutcomes

Managerialpractice

Creates,maintains,disrupts

Creates,maintains,disrupts

Influences

First-orderstrategy

discourse

Shapes

Figure 1. The relationship between institutional out-comes, managerial cognition, managerial practice, and

first-order strategy discourse

employed by strategists during institutional adop-tion, and (2) how these meanings relate to partic-ular institutional work carried out by these actors.

Cognition–Discourse

Cognition has been operationalized through con-cepts such as “schema” in cognitive psychol-ogy (Condor and Antaki, 1997) or “cognitivemaps” in management studies (Eden, 1992). Thesemaps symbolically represent both broad knowl-edge about a domain as well as perceived inter-relationships among its parts, described by Weick(1977: 277) as “networks of causal sequences,”which shape processes of enactment (Weick, 1979)through agents’ practices. Discursive social inter-action is the primary means through which cog-nition is shaped as well as shared (van Dijk,1990). Discourse provides labels, typifications, orframes through which we understand and inter-pret the world around us (Berger and Luckmann,1966), at the same time drawing on and shap-ing our mental representations of the world. Instrategic management, the concept of dominantlogic, which is “stored via schemas” (Prahaladand Bettis, 1986: 490) and acts as an informa-tion filter—which influences what type of infor-mation organizations pay attention to and how(Bettis and Prahalad, 1995)—can be seen as anoperationalization of the broader idea of cognition.More recent developments examine material andmetaphorical aspects of managerial discourse (Her-acleous and Jacobs, 2008) as well as emotionaland nonconscious aspects of cognition (Hodgkin-son and Healy, 2011).

Practice–Discourse

Within strategy-as-practice, empirical studies haveset the seeds for such first-order examination. Jarz-abkowski and Sillince (2007) argue that top man-agers, employing internally consistent rhetoricallanguage, can construct a context for employeecommitment to multiple strategic goals. Man-tere and Vaara (2008) identify discourses thatseem to systematically impede or promote par-ticipation in the strategy process. While thisresearch demonstrates the potential that a focuson strategy discourse has for enhancing ourunderstanding of strategy practice, it has yetto offer extensive insights about the nature ofmeanings of strategy employed by practitionersduring strategic change in the form of institu-tional adoption. This is our first unresolved puz-zle. Our proposed solution is to approach theconcept of strategy by drawing on both thepractice perspective—through our focus on howagents employ the dimensions of the meaning ofstrategy—and on discursively informed work oninstitutional entrepreneurship—by exploring howagents’ use of first-order discourse relates to theinstitutional work they carry out.

Discourse–Outcomes

While discursive outcomes have been studiedthrough the lenses of strategic decisions (Hendry,2000), strategy implementation (Heracleous andBarrett, 2001), and strategic change processes(Hardy, Palmer, and Phillips, 2000), the mean-ing of strategy itself, from the perspective ofindividual strategists, has not. The strategists thatare the focus of our study are strategy directorsand their teams, seen as a new support function(Kaplan and Norton, 2005). Given strategy direc-tors’ key role in connecting strategy across levels(Angwin, Paroutis, and Miston, 2009), we con-ceptualize them as institutional entrepreneurs whohave the ability and skill-set to influence the socialorganization of firms and support the establish-ment of new practices (Garud et al ., 2002; Green-wood, Suddaby, and Hinings, 2002; Lounsbury,2002; Maguire et al ., 2004). Discourse is inte-gral to this institutional entrepreneurship processsince it is through discourse that stakeholders canbe engaged and new social realities diffused andlegitimated (Jarzabkowski and Sillince, 2007; Sud-daby and Greenwood, 2005), particularly when a



new institution is created (Tracey, Phillips, andJarvis, 2011). Yet despite these advancements,we are missing research particularly on the first-order aspects of the discourse used by institutionalentrepreneurs in the process of adopting an insti-tution. As we try to learn more about how strategyis carried out in practice, it becomes essential thatwe identify what strategy as a concept means tostrategists themselves, and how these meaningsrelate to strategists’ actions. This is our secondunresolved puzzle. Our proposed solution is toconceptualize institutional entrepreneurs as insti-tutionally embedded agents who utilize discoursein their institutional work. An important aspect ofinstitutional entrepreneurship studies is the out-comes of entrepreneurship activity. Lounsburyand Glynn (2001) proposed linking entrepreneurs’story-telling activities together with wealth cre-ation. Following a similar line of thought, theoutcome central to our study is the adoption ofa strategy-related institution in the form of a net-work of strategy professionals within a strategysupport function in a multibusiness firm. In addi-tion, we remained alert to performative outcomesfollowing recommendations posed for strategy-as-practice research (Johnson et al ., 2003; Tsoukas,2010, Whittington, 2007).

The purpose of the present study, therefore,was to build theory about how institutionalentrepreneurs use first-order strategy discoursewhen adopting an institution. Without such anunderstanding, we lack insights into the determi-nants of long-term discursive patterns and orga-nizational or industry-level changes that havereceived the majority of attention by previousstudies on strategic change. Given this gap inthe literature and rising calls for such microlevelwork in strategy-as-practice (Ketokivi and Man-tere, 2010; Tsoukas, 2010; Vaara and Whittington,2012) and institutional theory fields (Jarzabkowskiet al ., 2009; Lawrence and Suddaby, 2006; Louns-bury, 2008), we set out to examine how institu-tional entrepreneurs make sense of and employthe concept of strategy. As shown in Table 1,we also argue that discourse-oriented examina-tions can help us connect institutional and prac-tice theories. Of the few studies that highlightthese intersections, two focus on rituals (Dacin,Munir, and Tracey, 2010; Tilcsik, 2010) and oneon macrolevel discourses (Tracey et al ., 2011). Inthis context, our study contributes to a microlevelunderstanding of the strategy discourse employed

by institutional entrepreneurs when they adopt aninstitution from the organizational field of FTSE100 firms they are embedded in. Accordingly, wepose two questions: What are the different dimen-sions of first-order strategy discourse as perceivedby strategists, and how do strategists, as insti-tutional entrepreneurs, employ these dimensionsover time in the process of institutional adoption?

METHOD

Approach and data collection

We followed Grant (2003) and adopted anexploratory-oriented methodology for two rea-sons. First, in line with previous discourse stud-ies (Hendry, 2000; Heracleous and Barrett, 2001),our aim was to provide an inductive account offirst-order strategy discourse and how it changesover time. Second, there is little empirically basedtheory that would help us unpack the meaningsof strategy as employed by practitioners and toconstruct relevant hypotheses. We collected twodatasets from FTSE 100 firms: interviews withstrategy directors and an in-depth longitudinal casestudy at UtilCo (the actual names have been dis-guised). This dual dataset follows the recommen-dations by Sandberg and Tsoukas (2011) as it helpsus develop a general template of the nature of first-order strategy discourse (exploratory study) withthe situational specificity of revealing how suchdiscourse is employed during institutional adoption(longitudinal case study).

Exploratory study (strategy directors)

We conducted 15 interviews with strategy directorsin 11 FTSE 100 firms. The purpose of these inter-views was to explore the dimensions of first-orderstrategy discourse as perceived by actors with keystrategic roles. We chose FTSE 100 firms because,due to their size and complexity of operations, theytend to employ managers with dedicated strategyroles. As noted above, we use the term first-orderin its phenomenological sense, to denote the per-spective of the agents themselves as opposed tothe perspective of others. Deriving from EdmundHusserl and then elaborated by his student AlfredSchutz, the distinction between first- and second-order meanings has become a recognized method-ological distinction in interpretive research (Smith,Flowers, and Larkin, 2009).



Table 1. Key features and links of discourse, institutional, and practice theories

Discourse theory Institutional theory Practice theoryDiscourse-practice-

institutions links

Drivingresearchquestion

How do actors employdiscourse to influencetheir context?

How do institutions getcreated, maintained,and changed?

How do practices leadto particularoutcomes?

Discursive practicesboth constitute, andare shaped by,institutions.

Theoreticalaims

To understand theprocess of the socialconstruction of realitythrough organizationaldiscourse and the roleof agency in thisprocess

To theorize aboutinstitutional forms andthe processes throughwhich they are createdand changed

To theorize about theconstitutiveprocesses ofenactment that resultin particularoutcomes andperformances

Institutional changeoccurs throughspecific practices,includingdiscursive ones

Role of agency Agents, acting withinestablished discursivefields, utilize discourseto influence otheractors and legitimizeor challenge particularactions and outcomes

Agents can have acentral and decisiverole in changinginstitutions, acting asinstitutionalentrepreneurs

Through their actions,agents over timeinitiate, affirm, orchallenge particularpractices

Agents canaccomplishinstitutional workthrough theirdiscursive practices

Role ofprocesses

Discursive configurationsand their dynamicsshape organizationaloutcomes

Regulative, normative,and cognitive aspectsof institutions aremanifested in specificprocesses

Recurrent andrelational processesat the microlevelhelp explaineveryday practices

Discursive practicesand processesshape theregulative,normative, andcognitive aspects ofinstitutions

Outcomeinvestigated

The success or failure ofparticular strategies orinitiatives asmanifested in andinfluenced bydiscourse

The discrete institutionalforms created byactors, groups ofactors, and theirorganizations

The action-focusedrelationships andperformances thatproduce certainoutcomes

In what way candiscursive practicessupport theadoption of aninstitution?

Key studies Balogun (2003),Balogun,Jarzabkowski, andVaara, (2011),Heracleous and Barrett(2001), Mantere(2008), and Mantereand Vaara (2008)

Greenwood and Suddaby(2006), Greenwoodet al . (2002), Hardyand Maguire (2010),Maguire et al . (2004),and Rojas (2010)

Feldman andOrlikowski (2011),Jarzabkowski andBalogun (2009),Nicolini (2011),Regnér (2003),Sandberg andTsoukas (2011). andWhittington (2006)

Dacin et al . (2010),Tilcsik (2010), andTracey et al .(2011)

Main study (UtilCo)

UtilCo is an FTSE 100 utility firm from the sampleof 11 FTSE 100 firms in the first dataset. UtilCowas formed in 1997 when its parent company wassplit into two separate firms. By 2002, and after aperiod of major changes, it moved from being atraditional utility to a multiproduct, multibusiness

organization. We chose UtilCo as our main casefor three reasons. First, it represented a newlycreated organization that was searching for a strat-egy. Thus, the firm had to draw on the currentmanagers and their capabilities to develop its strat-egy. Second, the United Kingdom utility sectorchanged dramatically due to the deregulation and



liberalization processes, meaning that managers inthis sector had to make and execute strategy underconditions of increasing uncertainty. Accordingly,the way these managers perceived the notion ofstrategy in such context is of particular schol-arly and practical relevance. Third, as with otherFTSE 100 firms, UtilCo is comprised of mul-tiple business units facing varying competitiveforces. These conditions mean that managers withdistinctive strategy roles were hired to create andthen support the UtilCo strategy process. We fol-lowed a longitudinal-case-based approach (Petti-grew, 1992) focusing on the strategy process andpractice at the UtilCo corporate center and twobusiness units: TelUnit, a telecommunications unit,and GasUnit, a gas supply and distribution busi-ness. We conducted 36 semistructured interviewsasking questions about the strategic planning pro-cess, the actions of key actors, the initiativesimplemented, and the methods used to supportthe strategy process (Mantere and Vaara, 2008).Consistent with our research focus, each inter-viewee was also asked a specific question aboutthe meaning of strategy (e.g., “What does strat-egy mean to you?”). The interviews, which lastedfrom one to two and a half hours, were digitallyrecorded with the approval of each informant andfully transcribed. The total dataset from the inter-views at UtilCo amounted to 580 pages (314,602words) of textual data. In addition, archivaldata were gathered in the form of documentsrelated to the UtilCo strategy process: annualreports, strategy documents, presentations, andspeeches.

Data analysis

Drawing on Grant (2003), and Jarzabkowski et al .(2009), and Laamanen and Wallin (2009), wefollow a four-stage approach for our data analysis(refer to Table 2).

First stage (exploratory)

Instead of treating strategy as a concept with apredetermined meaning, we treat it as a contingentconcept, allowing it to emerge from the data. Thismethodological treatment enables us to get closeto how strategists use the term, so that we cankeep our second hermeneutic (our interpretationsof strategists’ interpretations) as close to thedata as possible (Giddens, 1984). Accordingly,

the 15 interviews with strategy directors wereanalyzed using inductive qualitative techniques(Glaser and Strauss, 1967; Strauss and Corbin,1990). Our analysis consisted of multiple, iterativereadings of the interview transcripts, and theidentification of dimensions of first-order strategydiscourse uttered by various actors. During coding,careful attention was paid to how our intervieweesdescribed and made sense of the concept ofstrategy, both through their response to our specificquestion about what strategy means to themas well as their related responses across thewhole interview narrative. Initial coding was verydetailed and included a number of codes aboutthe different dimensions of first-order strategydiscourse. Informed by our grounded approachand through a continual process of refinement,we grouped these initial codes into groups ofhigher-order concepts representing four principaldimensions of first-order discourse (Strauss andCorbin, 1990). Overall, the aim of this stage wasto identify the dimensions of first-order strategydiscourse as perceived by strategists. Codingreliability was checked by having two coders, inaddition to the original coder, code a sample of theinterviews (agreement was 94.8 percent). Overall,this analysis of our interviews with strategydirectors revealed four primary dimensions of first-order strategy discourse: functional, contextual,identity, and metaphorical.

Functional dimension

In this dimension, informants perceive the meaningof strategy to revolve around what strategy isintended to achieve and what strategists “do”to realize strategy: their particular activities thatare central in strategizing (executing, reflecting,coordinating, etc.), as well as the norms, dominantlogics, or procedures for doing so. Strategy here isall about facilitating decisions among competingalternatives and moving towards implementationthrough preparation of plans that guide actions andthe allocation of resources.

Contextual dimension

In this dimension, the meaning of strategy iscontingent on where strategy is carried out withina complex organization, as well as the location-related nuances around practicing strategy withinthe firm, for example, the level of centralization,



Table 2. The four research stages

Research stages Procedure and aim Data analyzed

First stage(exploratory) • Coding of the different dimensions offirst-order strategy discourse based on thestatements of interviewees about whatstrategy means to them.

• To identify relevant, broad dimensions offirst-order strategy discourse.

• Fifteen interviews with strategy directorsfrom 11 FTSE 100 companies in 2002

• Quality check: intercoder reliability checks.Second stage(case

history)• Production of a timeline summarizing the

evolution of UtilCo to enableorganization-level contextualization of thedimensions of first-order strategy discourseidentified in the first stage.

• Speeches, presentations, company annualreports, press releases, industry reportsabout the focal FTSE 100 company(UtilCo) covering the period 1997–2005.

• Quality check: feedback about the case from the UtilCo strategy director.Third stage(coding

analysis)• Thematic analysis of the key strategic

initiatives, actors, and their key actionsaround the annual strategic planning process.

• Coding of the interview data into distinctiveperiods in the firm’s history.

• To code first-order strategy discourses acrossthe four dimensions identified in the firststage.

• To code instances of institutional work basedon the typology of Lawrence and Suddaby(2006).

• Nineteen interviews with managers in theUtilCo corporate center.

• Seventeen interviews with managers intwo business units (TelUnit and GasUnit).

• Total: 36 interviews conducted in UtilCoduring the period 2002–2004.

Fourth stage(analysis offirst-order discourseand institutional work)

• Analysis of the dimensions of first-orderstrategy discourse over time using codingmatrixes.

• Cross-reference of dimensions of first-orderstrategy discourse with instances ofinstitutional work.

• To identify coding references counts andlevel of correlation.

• Quality checks: Intercoder reliability checks. Triangulation with archival data and casehistory. Follow-up meetings with strategy managers at UtilCo.

leadership style, history of strategy development,and the role of the strategists. Particularly in themultibusiness firm, business units (periphery) oftenhave different cycle times, markets, and localplanning processes across units, and between unitsand the corporate center (Martin and Eisenhardt,2010). This creates complex political dynamicsassociated with different and sometimes conflictingbeliefs about the meaning of strategy.

Identity dimension

A key aspect of the meaning of strategy involveswhat it means to “be” a strategist in a particularorganization. When strategy directors speak aboutstrategy, they typically refer to what makes astrategist, highlighting certain capabilities such as

challenging convention and understanding indus-try value drivers and their shifts. The strategist,therefore, is distinguished from the nonstrategistthrough their capabilities in a multiplicity of strate-gizing tasks, prowess in strategic thinking, andability to network with multiple strategic actors.

Metaphorical dimension

When discussing strategy, strategy directors com-monly display some key discursive elements,which include the normative aspect of aspirationstypically expressed through directional and mech-anistic metaphors. Often, the underlying metaphorhere is “strategy is a journey,” consistent with theprevalence of the journey metaphor in other typesof discourse (Lakoff and Johnson, 1980). Table 3



below outlines the coding results from this stage,which provided the template for the analysis of theUtilCo case study.

Second stage (case history)

We compiled a dataset of secondary archivaldata in the form of company annual reports,press releases, and industry reports about UtilCocovering the period 1997–2005. From this dataset,we developed a group level case (UtilCo corporatecenter), two embedded business unit cases (TelUnitand GasUnit), and a timeline of strategy-relatedevents. These case studies helped us contextualizethe dimensions of strategy identified in stage one.

Third stage (coding of interviews)

We followed three steps to develop the UtilCocase study. First, we conducted a thematic analysisof the key initiatives, actors, and their institutionalwork around the strategic planning process. Sec-ond, we coded the interview data into particularperiods in the firm’s history. Third, we codedthe instances of text related to first-order strategydiscourse and institutional work. For the first-orderstrategy discourse coding, we selected the fourdimensions (functional, contextual, identity, andmetaphorical). For the institutional work coding,we followed Lawrence and Suddaby (2006) andfocused on the dimensions of institutional workthat were more closely relevant to our casestudy. Overall, we found the following categoriesof institutional work: defining, constructingidentities, changing normative associations,theorizing, changing normative networks, andeducating (Lawrence and Suddaby, 2006).

Fourth stage (analysis of first-order strategydiscourse)

In this stage, we focused our attention on thecorpus of 380 interview segments relating tofirst-order strategy discourse. Having coded thedimensions of first-order strategy discourse, aswell as the instances of text referring to particularinitiatives, actors, periods, and institutional work,we related these two lines of enquiry by examiningthe extent to which the dimensions of first-orderstrategy discourse varied across different locations,actors, and over time. We also cross-referencedthese chronological changes in terms of emphasis

in the dimensions of first-order strategy discourse,with changes in institutional work associated withthe adoption of the institution at UtilCo. Thisprocess allowed us to identify three distinctiveconfigurations of first-order strategy dimensionsand associated institutional work in our dataset,which we represent across three phases: shaping,settling, and selling. In labeling these phases, wewere inspired by the three broad categories ofinstitutional work involved in creating institutionsoutlined by Lawrence and Suddaby (2006). Wealso performed quantitative analysis of the codingreferences of first-order strategy discourse tocomplement our initial qualitative appreciationof the change in emphasis on particular first-order dimensions over time. In order to limitthe researcher bias, we triangulated our interviewfindings with our archival data and our casehistories. Once more, we checked our codingreliability by having two coders code a sample ofUtilCo interviews (agreement was 96.4 percent).Finally, we also validated our findings in follow-up meetings with UtilCo managers.

FIRST-ORDER STRATEGY DISCOURSEDURING INSTITUTIONAL ADOPTION

We now turn to the findings of our case study.Following our research questions, we examine howpatterns in participants’ dimensions of strategicdiscourse over time relate to forms of institutionalwork they adopt. We find three distinctive phasesin the relation of first-order strategy discourse andassociated institutional work: shaping, settling, andselling. While these phases are continuous andoverlap, the configurations of first-order strategydiscourse and institutional work had a distinctiveimpact in the adoption of the strategy functionat UtilCo. In each phase, we take two steps.First, we describe the changing emphasis in thefour first-order discourse dimensions. Second, wedetail the institutional work performed duringeach phase. These two lines of analysis are thenconnected to examine how patterns in participants’dimensions of first-order discourse over time relateto forms of institutional work they adopt. Table 4provides supporting quantitative and qualitativeevidence for each phase with numbered referencesto particular quotations included in our findingsbelow. Figure 2 is a visual representation ofthe relationship between emphasis in first-order



Table 3. Dimensions of first-order strategy discourse from the exploratory study

Functional Contextual Identity Metaphorical

Definition • Strategy isgoal-oriented; itfacilitates decisionsamong competingalternatives and enablesimplementation throughpreparation of detailedplans that guide actionsand the allocation ofresources.

• Strategy is contingenton the organizationallocation of strategicactivities and thenorms of practicingstrategy in particularorganizations.

• Strategy is a resourcefor identityconstruction. Astrategist, for example,is able to challengeconvention andunderstand industryvalue drivers and theirshifts.

• Strategic discourse isaspirational, typicallyexpressed in terms ofdirectional, military,or mechanisticmetaphors, which areexplicitly employed.

Guidingquestion

• How do strategistsperceive the task andprocess of strategy?How does this influencewhat they do and howdo they do it?

• How does theorganizational contextof strategists affecthow they do strategy?

• Who are thestrategists? How dothey think ofthemselves and theirrole?

• How do strategistsrhetorically expresstheir view of strategy?

Indicativequotes

• We develop a fact base,we then developalternatives, thosealternatives are thensubmitted to a dialoguebetween the businessunit and the groupcenter. (director,BankCo)

• How are we going totake this groupforward, how weoperate as a group, sothat the role of eachbusiness is within thecontext of the group,where they have to beheld back, where theyhave to be allowed togo forward. (director,UtilBiz).

• A good strategist is apartner, is a soundingboard, he is achallenger, he isallowed to provideunsolicited criticism,as long as it isconstructive. (director,OilBiz)

• For me it means thedirection we point outfor the company. And,how to reach aninspirational target.(director, PharmaCo)

• There is a lot more thanpurely delivering theplan, and that is wherestrategy starts to comein, strategy starts toshow the strategicchoices that you intendto make in the future or,at least, the informationyou will take intoaccount to start makingthese choices. (director,OilBiz)

• We (group strategyteam) are altogetheraround 15 people. Andour role is to see tothat . . . there are nocontradictions withinthe different parts ofthe organization whenit comes to thesestrategies. (director,PharmaCo)

• A strategist needs tobe fully aware of theexternal world, thedevelopments there, tomake his assumptionsabout the future.(director, OilBiz)

• It is trying to formulatean aspiration, for abusiness or aninitiative or a country. . . It is an aspiration,it’s something youwould like to try toachieve in the longterm, and yourstrategy describes theroute . . . you can getthere. (director,OilBiz)

strategy discourse dimensions and the three phasesof institutional adoption.

Phase 1: Shaping—emphasis on identityand functional dimensions

Our case study examines the emergence of first-order strategy discourse over time by analyzingthe development of the UtilCo strategic planningprocess from 2000 to 2003. UtilCo was estab-lished as a utility provider in 1997 and rapidlyexpanded into new services and markets, changing

from a traditional focused utility to a diversifiedmultiutility. As the CEO notes: “By any stretch ofthe imagination the UtilCo story is an extraordi-nary one. We’ve lived through the most intenseperiod of upheaval and have changed beyondrecognition” (Conference speech, October 2001).During 1997–2000, the strategy-making processwas based on an annual budgetary process, whilethe principal strategic directions were taken atthe corporate center. The early strategic decisionsfocused on improving the economic performanceof the core business and acquiring firms to expand



Tabl

e4.

Dim

ensi

ons

offir

st-o

rder

stra

tegy

disc

ours

eem

phas

ized

and

asso

ciat

edfo

rms

ofin

stitu

tiona

lw

ork

acro

ssth

eth

ree

phas

esin

Util

Co

Phas

e1

Phas

e2

Phas

e3

Dim

ensi

onR

efer

ence

sPe

rcen

tage

(%)

Ref

eren

ces

Perc

enta

ge(%

)R

efer

ence

sPe

rcen

tage

(%)

Iden

tity

4639

.022

16.2

1814

.3Fu

nctio

nal

4437

.340

29.4

2822

.2C

onte

xtua

l15

12.7

4835

.342

33.3

Met

apho

rica

l13

11.0

2619

.138

30.2

Tota

l11

810

0.0

136

100.

012

610

0.0

Phas

e1:

Shap

ing

Dim

ensi

ons

offir

st-o

rder

stra

tegy

disc

ours

e(i

dent

ity,

func

tiona

l)Fo

rms

ofin

stitu

tiona

lw

ork

(defi

ning

,co

nstr

uctin

gid

entit

ies)

[1.1

]So

Im

ean,

you

know

,th

ere

was

n’t

stra

tegy

here

.It

was

n’t

just

me

that

reco

gniz

edth

is,

ofco

urse

,bu

t,yo

ukn

ow,

Iw

ason

eof

the

new

wav

eof

man

ager

sw

ithin

this

orga

niza

tion.

(Gas

Uni

t,di

rect

orof

mar

ketin

gan

dst

rate

gy)—

Iden

tity

[1.3

][T

hece

ntra

lst

rate

gyfu

nctio

n]th

enhe

lped

tofo

rmth

eto

ple

vel

stra

tegi

cim

pera

tives

whe

re’s

Util

Co

goin

g,an

dw

hat

dow

ene

edto

focu

son

.(G

asU

nit,

man

agin

gdi

rect

oran

dgr

oup

depu

tyC

EO

)—D

efini

ng[1

.2]

Ith

ink

that

we

wer

eve

rym

uch

arou

ndex

ecut

ion.

Wha

tth

ecu

stom

erw

ante

dw

asn’

tat

the

hear

tof

wha

tw

edi

d,an

dw

edi

dn’t

thin

kof

the

inte

rcon

nect

ivity

betw

een

the

vari

ous

stra

nds

ofou

rde

liver

y.(G

asU

nit,

dire

ctor

ofm

arke

ting

and

stra

tegy

)—Fu

nctio

nal

[1.4

]W

hat

the

cent

erth

enha

dw

asa

stra

tegy

func

tion

all

bein

gin

side

finan

ce,

soit

was

ast

rate

gyfu

nctio

nfin

ance

.B

ut,

Im

ean,

they

wer

ere

alst

rate

gist

sra

ther

than

finan

cepe

ople

who

turn

edth

eir

hats

tobe

ing

stra

tegi

sts.

(Gas

Uni

t,m

anag

ing

dire

ctor

and

grou

pde

puty

CE

O)—

Con

stru

ctin

gid

entit

ies

Phas

e2:

Settl

ing

Dim

ensi

ons

offir

st-o

rder

disc

ours

e(f

unct

iona

l,co

ntex

tual

)Fo

rms

ofin

stitu

tiona

lw

ork

(cha

ngin

gno

rmat

ive

asso

ciat

ions

,th

eori

zing

)[2

.1]

Ith

ink

for

me

now

it’s

am

uch

clea

rer

fram

ewor

k..

.th

ere

are

som

ecl

ear

corp

orat

ego

als

inw

hich

you

have

tow

ork.

And

ther

efor

ein

deve

lopi

ngyo

urow

nst

rate

gyyo

uca

nso

rtof

keep

goin

gth

roug

han

itera

tive

proc

ess

that

says

are

we

mee

ting

the

corp

orat

ego

als,

aren

’tw

e.So

inth

atse

nse

itw

orks

muc

hbe

tter.

(Tel

Uni

t,di

rect

orof

HR

)—Fu

nctio

nal

[2.3

]T

he“m

anag

ing-

for-

valu

e”th

inki

ngha

sgi

ven

them

[tho

sein

volv

edin

stra

tegy

]a

mor

eco

mm

onla

ngua

ge.

You

goin

toan

yof

the

busi

ness

esan

dhe

arpe

ople

talk

abou

tth

em

anag

emen

tag

enda

s,or

fact

base

orvo

lum

eto

valu

e..

.w

hich

has

help

eddi

alog

uebe

twee

nth

ein

divi

dual

busi

ness

es(T

elU

nit

hum

anre

sour

ces

dire

ctor

)—T

heor

izin

g[2

.2]

The

thin

gth

atm

ade

itsu

cces

sful

,I’

vere

ferr

edto

ital

read

y,w

asth

efa

ctth

atth

eG

roup

seem

edto

beco

mpl

etel

yal

igne

dw

ithan

dhe

lpin

gth

ebr

ands

[bus

ines

sun

its].

(Gas

Uni

t,m

anag

ing

dire

ctor

)—C

onte

xtua

l

[2.4

]T

hese

nior

finan

cete

amin

Gas

Uni

tis

new

toth

ebu

sine

ss.

And

you

can

see

that

patte

rnre

plic

ated

inot

her

func

tions

...

So,

new

thin

king

,ne

wpa

radi

gms,

new

way

sof

wor

king

,di

ffer

ent

valu

esan

dbe

havi

ors

into

the

orga

niza

tion.

(Gas

Uni

t,di

rect

orof

stra

tegy

and

plan

ning

)—C

hang

ing

norm

ativ

eas

soci

atio

ns

Phas

e3:

Selli

ng

Dim

ensi

ons

offir

st-o

rder

disc

ours

e(c

onte

xtua

l,m

etap

hori

cal)

Form

sof

inst

itutio

nal

wor

k(c

hang

ing

norm

ativ

ene

twor

ks,

educ

atin

g)[3

.1]

Wha

tI

see

toda

yis

the

bran

d[b

usin

ess

unit]

stra

tegi

esbe

com

ing

ever

mor

epo

wer

ful.

And

ther

efor

ein

form

ing

muc

hm

ore

wha

tth

eco

rpor

ate

stra

tegy

shou

ldlo

oklik

e.(G

asU

nit,

man

agin

gdi

rect

or)—

Con

text

ual

[3.2

]T

here

alch

alle

nge

for

usw

asac

tual

lyth

enge

tting

ever

ybod

yto

unde

rsta

ndw

hat

itm

eant

for

them

inth

e40

,000

peop

leth

atw

ork

inU

tilC

o.Y

oukn

ow,

wha

tpa

rtdo

Ipl

ay?

I’m

atin

ylit

tle,

you

know

,co

gin

this

huge

mac

hine

.W

hat

doI

have

todo

diff

eren

tly?

From

wha

tyo

u’ve

told

me,

this

isou

rst

rate

gyno

w,

but

wha

tdo

esit

mea

nfo

rm

e?(G

roup

mar

ketin

gdi

rect

or)—

Met

apho

rica

l[3

.3]

I’m

cons

tant

lyif

you

like

guid

ing

the

hors

ear

ound

the

race

trac

kw

ithth

eoc

casi

onal

whi

pbu

tm

ostly

anen

cour

agin

gw

ord.

(Gas

Uni

t,di

rect

orof

mar

ketin

gan

dst

rate

gy)—

Met

apho

rica

l

[3.4

]E

ach

busi

ness

has

ape

rfor

man

ceco

ntra

ct,

whi

chis

defin

edas

atw

o-w

ayag

reem

ent.

One

side

isag

reei

ngto

inve

stfu

nds

inth

isbu

sine

ssan

dsu

ppor

tth

est

rate

gy,

the

othe

rsi

deis

com

mitt

ing

tode

liver

ale

vel

ofpe

rfor

man

ceth

atth

eyha

veag

reed

onm

utua

lly.

(Gro

upst

rate

gym

anag

er)

—C

hang

ing

norm

ativ

ene

twor

ks[3

.5]

Soth

ere

isqu

itea

disc

iplin

edpr

oces

s,bu

tat

the

sam

etim

eas

doin

gth

at,

over

the

last

two

year

s,w

e’ve

been

real

lyon

ajo

urne

yto

teac

hm

ost

ofth

eor

gani

zatio

nw

hat

stra

tegy

actu

ally

is.

(Tel

Uni

t,m

arke

ting

dire

ctor

)—E

duca

ting



Phase 1shaping

Phase 2settling

Phase 3selling

Identity Functional Contextual Metaphorical

Dimensions of first-order strategy discourse

Ph

ases

of

inst

itu

tio

nal

ad

op

tio

n

High: above 25% of totalreferences in particular phase

Medium: between 16% and24% of total references inparticular phase

Low: below 15% of totalreferences in particular phase

Figure 2. Strength of relationship between phases of institutional adoption and dimensions of first-order strategydiscourse based on coding references

into new market segments. The newly acquiredbusinesses offered distinctive services and wereemploying distinctive strategy processes. At thesame time, there was pressure from investors toalign the firm more closely to practices found inother FTSE 100 firms. In the face of these pres-sures, the UtilCo executive team decided to rad-ically change the firm’s strategy process towardsa “managing-for-value” approach, and, as a result,project “Finland” was launched in 2001. To exe-cute this project, the corporate center strategy teamlaunched a number of initiatives to communicateacross the firm the principal aspects and require-ments of the new strategy process. This smallgroup of practitioners was acting as institutionalentrepreneurs—as powerful professionals who hadthe conviction that they possessed the capabilitiesand skill-set to support the new strategy process.As the data show, the principal dimensions ofthe first-order strategy discourse that were promi-nent in this phase were the identity and functionalones. The identity first-order strategy discoursewas central in constructing the identity of thenew central strategy team as experts in strategyand increased their organizational legitimacy at thelaunch of the new process [1.1]. This central strat-egy team was formed by hiring former consultantswho were seen as having the capabilities to sup-port the new process: “We brought in quite a lot offormer consultants, who were very bright people,into the organization, people from other blue-chiporganizations.”—Identity (Group HR manager).

Functional first-order strategy discourse wasalso emphasized during this period [1.2]. Strategyin this dimension is seen in a classical sense, as

goal-oriented, providing parameters for decisionmaking and taking action. As the group strategymanager notes, for example, the focus of strategyis about achieving particular value maximizingactions:

We have a single governing objective, andthe whole premise on which our wholeapproach to strategy is based, is that wewant to maximize value and value can bemanaged, it does not happen by accident. . . . .So different courses of action will yielddifferent economic value. Those courses ofaction, is what we call the strategy.—Functional (Group strategy manager)

Alongside first-order discourse, the institutionalwork emphasized in this phase was about definingand constructing identities. Defining involvedwork aimed at constructing a set of rules about theprecise expectations and deliverables of the newstrategy process [1.3]. The roles of those involvedin executing the new business model were created,which also created a number of expectations aboutthe distinctive contribution of the strategy directorand his strategy team across the group. Effort wasalso placed to become more innovative in respond-ing to new opportunities: “And during this period,I guess we started to become more innovative inthe way we approached our market opportunitiesand the challenges that we were facing, and par-ticularly in our core energy business”—Defining(Group marketing director). Constructing identi-ties was about describing the relationship between



strategists and the field in which they were calledto operate, both within the firm and the utilityindustry. For instance, despite the fact that thecentral strategy team was closely collaboratingwith the finance function, they were perceivedas having a separate identity [1.4]. Particularexpectations in terms of the analytical capabil-ities expected by strategy managers were alsoarticulated:

. . . if you’re a senior [strategy] manager. . . what you should be doing is determinewhich businesses should be part of theportfolio, to understand the influence of theregulatory environment, to set the strategicagenda . . . —Constructing identities (Groupmarketing director)

The performative outcomes in this phase wereabout developing organizational capacity andstrategist legitimacy to support institutional adop-tion. More specifically, the central strategy teambuilds their identity as the experts who havecome to install a new strategy process, and theyare also being called upon to build the capabil-ity of the senior team to discuss strategy-relatedissues. Overall, in this initial phase, we found thatfirst-order strategy discourse reflected efforts ofinstitutional entrepreneurs to define the rules, iden-tities, and boundaries around the adopted strategyfunction. It was a phase of initiation and experi-mentation, when the key features of the strategicplanning process were created.

Phase 2: Settling—emphasis on functional andcontextual dimensions

By the end of 2001, most of the UtilCobusinesses were delivering positive performanceresults. However, as UtilCo added more units to itsstructure, implementing particular strategic goalsbecame increasingly complicated. In these condi-tions, the UtilCo senior management supported thecreation of peripheral strategy teams located withineach business unit. These teams would help imple-ment project Finland and would be coordinatedby the central strategy team. In this phase, thefunctional and contextual dimensions of first-orderstrategy discourse were prominent. Regarding thefunctional dimension, the business unit strategyteams played an active role in clarifying the

managing-for-value framework [2.1]. For instance,they ran a number of training sessions in their unitsto help define the key “value-at-stake” activities:

Year two, the additions were that we had alot of effort into generating a fact-base aboutthe business, who our competitors were, whatwere the drivers for value in the area of cus-tomer contribution . . . we made more signif-icant efforts to define the key value-at-stakeactivities.—Functional (GasUnit, director ofmarketing and strategy)

The contextual first-order strategy discourse wasalso emphasized in this period. Strategy in thisdimension is contingent on the organizational loca-tion of strategic activities. Within UtilCo, theperipheral strategy teams supported the alignmentof the local strategy processes to the standardsrequired by the managing-for-value process [2.2].This active engagement of business unit strategymanagers led them to develop their own elaboratemeanings of strategy and work towards adjustingaspects of the planning process within their busi-nesses, emphasizing the contextual dimension ofstrategy discourse, presenting strategy as shapedby its organizational location: “It’s top-bottom-up . . . we don’t have a big group at the cen-ter trying to do all the strategic thinking. Strat-egy is coordinated at the center but the actualplanning and thinking and evaluation goes on inthe business units”—Contextual (Group strategydirector). Considering the actors involved in thestrategy process, from a single group of insti-tutional entrepreneurs (central strategy team) inphase 1, we found in phase 2 groups of institu-tional entrepreneurs (central and peripheral strat-egy teams) trying to make sense of strategy them-selves (for instance, during workshops) and at thesame time present strategy “at the level that peo-ple understand” in order “to get their buy-in.” Thismeans that in phase 2 these strategists, by sociallyconstructing (Berger and Luckmann, 1966) themeaning of strategy, were also socially construct-ing their identity within the organization as expertsin strategy who have the right skill-set to consultto top management and support middle-level man-agers during the strategy process.

The institutional work stressed in phase 2 wasabout theorizing and changing normative associa-tions. Theorizing referred to the development and



naming of concepts around the strategy process.Particular attention was placed on clarifying themanaging-for-value approach [2.3]. After the firstphase, when the managing-for-value concepts andframework were launched, in this phase, we have amore active discussion about what these conceptsactually mean, particularly in relation to the localstrategy processes in each business unit.

So it’s not enough to say that we’re creatingvalue, we have to “check” against other alter-natives to make sure that we are pursuing acourse of action that we think maximizes thevalue of the company.—Theorizing (Groupstrategy manager)

Changing normative associations referred tothe impact of the practices by strategists to theunderlying values and culture found in the strategyprocess. As such, the new strategy teams wereperceived as equals to other functions in UtilCo,and as a group representing new ways of actingbased on different values and mindsets [2.4]. Also,work carried out by the business unit strategyteams now helps to align the diverse processesfound in phase 1:

I think what has been the change, especiallyfor 2002, was a much greater focus onactually . . . bringing it all together inone clean process.—Changing normativeassociations (GasUnit, strategy manager)

The performative outcomes in phase 2 wereabout embedding new strategy concepts and build-ing strategy process capability to support institu-tional adoption. The new strategy concepts thatwere introduced in phase 1 (for instance, themanaging-for-value framework) were now dis-cussed more extensively, accompanied by effortsto build strategy process capability since strategypractitioners were starting to deal with issues thatemerged as problematic after phase 1. In order todeal with these issues, the central strategy teamcoordinated a series of interactions with the busi-ness unit strategy teams. Differences emerged interms of the way concepts, templates, tools, andmetrics of the strategy process had been perceivedby the various units. This, in turn, led the strategyteams to theorize about the managing-for-value

approach, which helped them clarify the key con-cepts of the new strategy process. Accordingly,we found that in this phase first-order strategy dis-course captured the work of strategists debatingabout the precise rules, belief systems, and bound-aries of the adopted institution. This was a phaseof reflection and coordination, when the features ofthe planning process were refined and discussed.

Phase 3: Selling—emphasis on contextual andmetaphorical dimensions

By 2003, the strategy teams at the corporatecenter and the business units of UtilCo had anincreased involvement in the strategy process.In this phase, execution through the managing-for-value model was at the top of the UtilComanagement agenda: “We’re now entering the nextphase of our development . . . our emphasis ison growing value from the business we’ve got.”(CEO, AGM speech, May 2003). The strategyprocess involved not only the strategy director andthe top management team, as in phase 1, but alsoa wider community of strategists, consisting ofthe central and peripheral strategy teams togetherwith middle-level managers. In order to achievethis level of engagement, a set of dialoguesbetween the center and the business units werelaunched. As more strategy teams and middle-levelmanagers were involved in the process, we foundthe contextual [3.1] and metaphorical [3.2] [3.3]dimensions of first-order strategy discourse to beemphasized. As a manager notes, strategy in phase3 was about delivering at the business unit levelwith middle-level managers who hold the localknowledge on performance indicators, and, as aresult, the contextual first-order strategy discoursewas predominant:

Making the UtilCo business model workis all down to translating it into realconcrete things that we do differentlyin our businesses . . . I think that thefocus of the 2003 strategy was turningwords into tangible commitments in ourbusinesses.—Contextual (Group strategymanager)

The emphasis on the metaphorical aspect ofthe first-order strategy discourse also becameincreasingly higher during this phase. Explicit



metaphors were employed that were at the sametime both aspirational and directional. In the quotebelow, the organization is seen as a ship:

We were afloat; we were not going to sink. . . And then, to continue my analogy,having now got the ship afloat and wethen said, well, hold on. We were goingto go to New York, that’s where the shipwas going. But now we’ve got a ship, youknow, that’s afloat, why don’t we go downto Rio de Janeiro.—Metaphorical (Groupmarketing director)

The institutional work emphasized in phase3 was about changing normative networks andeducating. Changing normative networks refersto the interactions across the network of topmanagement, strategy teams, and middle-levelmanagers to define a way to monitor and evaluatethe actions and deliverables of the strategy process.As a result of these interactions between the centerand the business units, performance contracts wereadopted and mutually agreed [3.4]. Through thesecontracts, the aim was to raise the standards ofdelivery within each business unit:

And then using tools we have such asvalue-driver trees . . . [to gain a] deepunderstanding of how value is driven inour business, this deep understanding wasabout building a set of performance commit-ments both at the business and the individ-ual level.—Changing normative associations(Group strategy manager)

Educating refers to the training of actors inskills and knowledge necessary to support theadopted strategy function. This training involvedthe central strategy team spending time in eachunit to enhance the capabilities of local managersaround the strategy process [3.5]:

Then each month they’ll [central strategyteam] come down and discuss with us thedrivers and the numbers. And I think thosemeetings are extremely valuable in educatingthe center in our business, in what arethe issues that affect our business, what isdriving our business.—Educating (TelUnit,finance manager)

The performative outcomes in this phase wereabout crystalizing the normative legitimacy ofthe strategy process and gaining the commitmentof multiple stakeholders to support institutionaladoption. Accordingly, we found an increase inthe perceived ability of business unit strategyteams and middle-level managers to (1) utilize theemerging strategy vocabulary around managingfor value, and (2) participate in an on-goingdialogue about adjustments in their local strategyplans with the corporate center strategy team.As such, the first-order strategy discourse in thisphase echoes work carried out across a network ofstrategy practitioners to formalize the rules, beliefsystems, and boundaries of the adopted institution.It was a phase of interaction and collaboration,when the key features of the strategic planningprocess were implemented and measurable per-formance outcomes were achieved. In summary,Table 5 presents the three identified phases,their definitions, the key actors, the first-orderstrategy discourse dimensions, and the associatedinstitutional work emphasized over time at UtilCo.

DISCUSSION AND CONCLUSION

Dimensions of first-order strategy discourse

Our study contributes to a deeper understandingof the concept of strategy in practice, of processesof institutionalization, and of the role of discoursein these processes. As indicated earlier, strategydirectors and their teams are now appearingin most large firms. The ways these “new”strategists understand, express, and employ theconcept of strategy as a discursive resourcereflects their efforts to legitimize their own actionsand identity in complex organizations. In ourtheory discussion, we suggested that strategydirectors and their teams can be perceived asinstitutional entrepreneurs whose role involvesthe fundamental rethinking and, if necessary, thechange of established ways of thinking and actingabout strategy. As they establish the parametersof an adopted institution (in our case, through theemphasis on the functional dimension), strategistssocially construct their identities within theircommunity (through emphasis on the identitydimension) and then project these identities acrosstheir organization (Bartunek, Balogun, and Boram,2011). This finding is consistent with Lounsbury



Table 5. Emphasized first-order strategy discourse dimensions and related forms of institutional work across threephases during institutional adoption

Phases (definition) Key actors

Dimensions offirst-orderstrategy

discourseRelated forms ofinstitutional work Nature of relationship

1. Shaping:Developingorganizationalcapacity andstrategist legitimacyto supportinstitutionaladoption

Central strategyteam, topmanagement

Identity andfunctional

Defining (the construction of rulesystems that confer status oridentity, define boundaries ofmembership, or create statushierarchies)

Constructing identities (definingthe relationship between anactor and the field in which thatactor operates)

First-order strategydiscourse reflectsefforts of institutionalentrepreneurs to definethe rules, identities,and boundaries aroundthe adopted institution.It’s a period ofinitiation andexperimentation

2. Settling:Embedding newstrategy conceptsand buildingstrategy processcapability tosupport institutionaladoption

Central strategyteam, businessunit strategyteams

Functionalandcontextual

Theorizing (the development andnaming of abstract categories,new concepts, and chains ofcause and effect)

Changing normative associations(remaking the connectionsbetween sets of practices andthe moral and culturalfoundations for those practices)

First-order strategydiscourse captures thework of strategy teamsthat debate about theprecise rules, beliefsystems, andboundaries of theadopted institution.It’s a period ofreflection andcoordination

3. Selling:Crystalizing thenormativelegitimacy of thestrategy processand gaining thecommitment ofmultiplestakeholders tosupport institutionaladoption

Central strategyteam, businessunit strategyteams,middle-levelmanagement

Contextualandmetaphori-cal

Changing normative networks(constructing of connectionsthrough which practices becomenormatively sanctioned andwhich form the relevant peergroup with respect tocompliance, monitoring, andevaluation)

Educating (training actors in skillsand knowledge necessary tosupport the institution)

First-order strategydiscourse echoes workcarried out across anetwork of actors toformalize the rules,belief systems, andboundaries of theadopted institution.It’s a period ofinteraction andcollaboration

and Glynn (2001) who propose that agents employdiscourse to fill a “cultural void” especially in thefounding moments of a new venture. In our studyof strategy as a concept in use, rather than asa term with a fixed, stable definition, we foundfour dimensions: functional, identity, contextual,and metaphorical. These dimensions can be seen asstructural features of first-order strategy discoursethat operate at the deeper semiotic or connotationallevel of strategic discourse (Heracleous, 2006).They act simultaneously as constraints and asresources (Hardy et al ., 2000) for interpretationand action, suggesting that strategy is a contingent

term intimately bound with strategy directors andmanagers who profess to do strategy and bestrategists.

Differential employment of dimensions offirst-order strategy discourse during phases ofinstitutional adoption

As shown in Table 5, we found patterns of differ-ential emphasis in the first-order strategy discourseacross the three phases. Specifically, we observedthat first-order strategy discourse evolves from



emphasizing predominantly functional and iden-tity dimensions in the first phase to emphasizingmainly metaphorical and contextual dimensions inthe third phase, with a settling phase in betweenwhere functional and contextual first-orderdiscourse is primary. Institutional theory canprovide resources for interpreting why thesedifferential emphases occurred. As the newinstitution in the form of a strategic functionwas beginning to take hold, it was importantto educate agents as to what strategy involvesand aims to achieve. An appropriate way to dothis was to draw from existing understandingsof strategy, itself a potent institution in terms ofits body of knowledge as well as its diffusionin organizations as a practice (Barry and Elmes,1997). This process can be seen as a form ofimitation (Galaskiewicz and Wasserman, 1989),associating the adopted institution of a strategicplanning function in UtilCo with an alreadyexisting form of knowledge and organizationalpractice in the broader institutional field of FTSE100 firms. Further, this emphasis on the functionaldimension was a form of definition (Lawrence andSuddaby, 2006: 222). The functional dimensionoutlined how the process of strategy would operatein UtilCo, in a way that distinguished strategists,the ones knowledgeable in the technology anddiscourse of strategy, from nonstrategists. Thedefining process involves close links betweeninstitutional logics and perceptions of identity(Rao, Monin, and Durant, 2003). Identity workduring the emergence of institutions is endemicand fundamental, as it defines the relationshipbetween agents and the institution (Lawrence andSuddaby, 2006; Oakes, Townley, and Cooper,1998). This sheds light on why the identitydimension of discourse was emphasized at theinitial stage to accompany the adoption of theinstitution. The following proposition capturesthis argument.

Proposition 1a: At the early stages ofinstitutional adoption, strategy practitionersutilize primarily functional and identitydimensions of first-order strategy discourseto develop organizational capacity andstrategist legitimacy to support institutionaladoption .

Our findings show that in period two, theintermediate phase of institutional adoption,

the functional and contextual dimensions wereemphasized. The central strategy team is coor-dinating (Jarzabkowski et al ., 2012) the workof peripheral strategy teams. Effectively, thislays the foundations for improved strategy pro-cess capability across the businesses. In theseconditions, a dialogue process (Tsoukas, 2009)starts across these strategy teams helping themto reflect on and clarify the rules, belief systems,and boundaries of the adopted institution. Thisdialogue process is then refined and extended toinclude more stakeholders in stage 3. Proposition1b encapsulates this process.

Proposition 1b: At the intermediate stage ofinstitutional adoption, strategy practitionersemphasize functional and contextual dimen-sions of first-order strategy discourse toembed new strategy concepts and build strat-egy process capability .

As our data show, in period three, the matu-rity phase of institutional adoption, the contextualand metaphorical dimensions were emphasized. AsLawrence and Suddaby (2006) note, embeddingand routinizing are central aspects of maintaininginstitutions, and attention to how generic strategyprocesses need to be adapted to fit specific orga-nizational contexts is essential to this embeddingand routinizing process. The institutionalizationof new practices is facilitated if they are consis-tent with existing practices. Such insight duringthis selling phase in the UtilCo multibusiness firmand across a network of strategy practitioners alsoconfirms recent findings about the discursive rela-tionship between parent-subsidiary in the multi-national firm (Balogun et al ., 2011). Further, atthis stage, where multiple stakeholders (central,peripheral strategy team, and middle-level man-agers) and contexts (corporate center and diversebusiness units) were interacting, the emphasis onthe metaphorical dimension is consistent with thepower of metaphor to appeal to multiple audiencesthrough its “plurality and openness of meaning”(Cornelissen, 2005: 753). Proposition 1c reflectsthis argument.

Proposition 1c: At the maturity stage ofinstitutional adoption, strategy practition-ers emphasize contextual and metaphoricaldimensions of first-order strategy discourse



to crystallize the normative legitimacy of thestrategy process and gain the commitment ofmultiple stakeholders .

This differential emphasis on dimensions offirst-order strategy discourse at different phasesof the institutionalization process contributes toa deeper understanding of the role of discoursein this process, an aspect that has not receivedsufficient research attention (Phillips, Lawrence,and Hardy, 2004). Such differential emphasis overtime, which we term “selective focusing,” allowsus to theoretically elaborate in more precise termsthe process of institutional adoption. Specifically,our findings demonstrate that selective focusingis connected with changes in institutional workcarried out by actors. This institutional work isabout defining, constructing identities, changingnormative associations and networks, theorizing,and educating (Lawrence and Suddaby, 2006). Forfuture studies, this realization opens the poten-tial to examine in more detail whether partic-ular trajectories of first-order strategy discoursecan be related to other kinds of institutional con-ditions beyond adoption, such as creation andmaintenance. Further, if we conceive groups asaggregations of individuals, and firms as aggre-gations of groups, the first-level strategy discourseat the microlevel influences how strategy is talkedabout and how these meanings are employed atthe group and firm levels. As such, the con-cept of first-order discourse provides a founda-tional view of the ways institutional entrepreneursinterpret and shape the strategy discourse, thusallowing researchers to capture more effectivelythe microdynamics of institutional entrepreneur-ship (Maguire et al ., 2004). The above argumentssuggest the following proposition.

Proposition 2: Strategy practitioners arelikely to utilize differentially multiple dimen-sions of first-order strategy discourse overtime and during their work associated withthe adoption of an institution .

We thus contribute to a deeper understand-ing of discourse in institutionalization processesby illustrating how practitioners emphasize differ-ent dimensions of first-order discourse while theyaccomplish aspects of institutional work. The fourdimensions we found are anchored close to the

principal actors producing the actual strategy dis-course, who employ them to challenge and influ-ence established ways of practicing strategy. Ourfindings also extend previous studies on stages ofinstitutionalization. While most prior such studieshave focused either on the broad issues of stability(DiMaggio and Powell, 1983) or change (Green-wood et al ., 2002), what we know less about ishow these stages are related and why (Zietsmaand Lawrence, 2010). Our study contributes tothis debate by showing that agents place differen-tial emphasis on discursive dimensions at partic-ular institutionalization stages, depending on thetype of institutional work they aim to achieve.We also provide an explanation on why these spe-cific links exist. In UtilCo, we find that the initialstage of institutional adoption (shaping), which ischaracteristic of change (Greenwood et al ., 2002),is followed by the intermediate stage of settling,which corresponds to a period of relative stability(DiMaggio and Powell, 1983), and the third, matu-rity stage is about selling, which is a return to aperiod of change. We argue that underlying theseshifts in these phases is a differential emphasis infirst-order strategy discourse or selective focusing.In other words, we suggest that selective focusingis one of the mechanisms that can help us explainthe relationships among institutional stages. Ourarguments so far are illustrated in Figure 3.

Performative outcomes in institutionaladoption

By examining first-order strategy discourse andthe associated institutional work, we reveal theimportance of considering the performative aspectsrelating to practitioners in their efforts to adoptan institution. Overall, we find that the organi-zational outcomes of institutional adoption (theadoption of a strategy function) are enmeshed withperformative outcomes (developing organizationalcapacity and strategist legitimacy, embedding newstrategy concepts and building strategy processcapability, crystalizing the normative legitimacyof the strategy process, and gaining the com-mitment of multiple stakeholders). Further, theseoutcomes are dependent on how agents employdiscourse to accomplish institutional work. Whileprevious studies of institutional entrepreneurshiphave focused on institutional outcomes at the inter-or intraorganizational level, our study addition-ally points to the importance of understanding the



Phase 1: Shaping Phase 2: Settling Phase 3: Selling

Constructingidentities

Time

First-orderstrategy

discourse

Forms ofinstitutional

work

Identity

Metaphorical

Contextual

Functional

Defining

Theorizing

Changing normativeassociations Educating

Changing normativenetworks

Contextual

Functional

Selectivefocusing

Selectivefocusing

Figure 3. First-order strategy discourse and forms of institutional work in UtilCo

microlevel, performative outcomes associated withthe adoption of institutions—such as the identityand credibility creation of practitioners involvedin the strategy process as knowledgeable experts.Such identification and legitimation processes takeplace at the individual and team level and, asposed by strategy-as-practice, are inherently socialaccomplishments (Chia and MacKay, 2007), withstrategists embarking on an on-going dialogue tosupport the strategy process (Tsoukas, 2009). Ourfindings also provide snapshots of performativeoutcomes during institutional adoption, specificallyrelated to the perceived improvement in the centraland peripheral actors’ capability to strategize. Pastinstitutional entrepreneurship research has exam-ined skills (Garud et al ., 2002) and points to theusefulness of focusing on the microlevel capa-bilities enabling (and hindering) the adoption ofinstitutions. As captured by Proposition 3, perfor-mative dimensions are an important feature of theinstitutional adoption process.

Proposition 3: Strategy practitioners employdimensions of first-order strategy discourseto accomplish performative outcomes, whichin turn facilitate the institutional adoptionbeing pursued .

Directions for future studies and practicalimplications

We propose directions for further studies by fol-lowing the same structure as our three domains

of contribution. Firstly, with respect to theidentification of four dimensions of first-orderstrategy discourse, it would be useful to investigatein more detail the links across the four dimensions(for instance, between functional and identity) andbetween each of the four dimensions and prac-tice (for instance, between identity and practice) toexplore whether, how, and why changes in thesedimensions could help us explain changes in strat-egy practices within and across firms. In addition,studies could focus on the extent of differences inthe perceptions of strategy between strategists andnonstrategists. With respect to our second domain,relating to the role of discourse in institutionaliza-tion processes, it would be useful to know whetherand to what extent the patterns we found of selec-tive focusing of discourse dimensions as relatingto the stages of adoption apply to other contexts,or whether different mechanisms (beyond selectivefocusing) or patterns can be observed. With respectto our third domain, the role of performative out-comes in institutionalization processes, it wouldbe useful to understand more about the types ofperformative outcomes most relevant to particu-lar settings. Finally, our study and findings have anumber of practical implications. As educators inour discussions with students, we can diverge fromstandard textbook definitions and focus on issuessuch as identity construction, legitimation, contex-tual embeddedness of strategy, and the variousmetaphors strategists employ. This would allow



us to educate students on the messy and contextu-ally contingent nature of strategy in practice, ratherthan to give the impression of strategy as a disen-gaged practice mainly involving strategic planningwith boxes and arrows. Further, strategists aspir-ing to change practices in their firms, or to influ-ence institutions, can gain sensitivity on how theyemploy strategic discourse, and what effects thismay have on their context.

ACKNOWLEDGEMENTS

We would like to thank Rich B

DISCOURSE REVISITED: DIMENSIONS AND ......SOTIRIOS PAROUTIS* and LOIZOS HERACLEOUS Warwick Business...

Documents

Transcript of DISCOURSE REVISITED: DIMENSIONS AND ......SOTIRIOS PAROUTIS* and LOIZOS HERACLEOUS Warwick Business...