ge history of org structure

29
Strategic Management Journal Strat. Mgmt. J., 33: 633–660 (2012) Published online EarlyView in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.1971 Received 15 January 2010; Final revision received 21 October 2011 ARCHITECTURE, ATTENTION, AND ADAPTATION IN THE MULTIBUSINESS FIRM: GENERAL ELECTRIC FROM 1951 TO 2001 JOHN JOSEPH 1 * and WILLIAM OCASIO 2 1 Duke University, Durham, North Carolina, U.S.A. 2 Northwestern University, Evanston, Illinois, U.S.A. In this study, we analyze how the organizational architecture of a multibusiness firm affects the adaptation of its constituent business units. Using an inductive analysis of GE’s governance system from 1951 to 2001, we examine how the integration of corporate and business unit attention occurs within and across the firm’s governance channels. Our theory identifies an unexplored aspect of the M-form’s architecture: collective vertical interactions between the corporate office and business units through cross-level channels. Overall, we articulate three types of channel integration—cross-level, cross-functional, and channel coupling—and examine their effects on responsiveness to threats and opportunities. We find that despite an elaborate organizational architecture, there were periods when GE’s governance system did not allow for coordination of corporate and business unit agendas. Our theory proposes that the temporal coupling of specialized, cross-level channels creates an organizational architecture that is both differentiated and integrated. This architecture integrates levels and issues simultaneously, yet focuses attention sequentially, providing more effective conditions for joint attention and coordination between the corporation office and the business unit and adaptive change at the business unit level. Copyright 2012 John Wiley & Sons, Ltd. INTRODUCTION This article examines a critical issue in strategic management: how the organizational architecture of a diversified multibusiness firm facilitates the strategic adaptation of its business units (Rumelt, Schendel, and Teece, 1994). In particular, our analysis focuses on the adaptive implications of corporate-business unit interactions. Beginning with Chandler (1962), the literature on diversi- fied firms focused on the superior information- processing and decision-making capacity of the multidivisional structure, with corporate office Keywords: attention; strategic adaptation; governance channels; cross-level; interactions Correspondence to: John Joseph, Duke University 100 Fuqua Dr., Durham, NC 27708, U.S.A. E-mail: [email protected] control of strategy formulation and business unit focus on implementation and operations (Chandler, 1962; Stinchcombe, 1990). Williamson (1975, 1983) expanded upon Chandler to develop the M- form hypotheses, where the corporate-level focus on strategy is accompanied by a financial con- trol system. The Chandler-Williamson perspective assumed a clear differentiation between business unit operations and corporate strategy, with lim- ited attention to integration beyond authority rela- tionships (cf. Lawrence and Lorsch, 1967; Gulati, Lawrence, and Puranam, 2005) and financial con- trol (Williamson, 1975). A large body of research has recognized that business unit adaptation is not merely a function of corporate fiat, but the consequence of a more elaborate relationship between corporate headquar- ters and business units (Gupta, 1987; Ghoshal Copyright 2012 John Wiley & Sons, Ltd.

Upload: artemisfowl01

Post on 16-Dec-2015

58 views

Category:

Documents


1 download

DESCRIPTION

Org structure of GE till 2001

TRANSCRIPT

  • Strategic Management JournalStrat. Mgmt. J., 33: 633660 (2012)

    Published online EarlyView in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.1971Received 15 January 2010; Final revision received 21 October 2011

    ARCHITECTURE, ATTENTION, AND ADAPTATIONIN THE MULTIBUSINESS FIRM: GENERAL ELECTRICFROM 1951 TO 2001JOHN JOSEPH1* and WILLIAM OCASIO21 Duke University, Durham, North Carolina, U.S.A.2 Northwestern University, Evanston, Illinois, U.S.A.

    In this study, we analyze how the organizational architecture of a multibusiness firm affects theadaptation of its constituent business units. Using an inductive analysis of GEs governancesystem from 1951 to 2001, we examine how the integration of corporate and business unitattention occurs within and across the firms governance channels. Our theory identifies anunexplored aspect of the M-forms architecture: collective vertical interactions between thecorporate office and business units through cross-level channels. Overall, we articulate threetypes of channel integrationcross-level, cross-functional, and channel couplingand examinetheir effects on responsiveness to threats and opportunities. We find that despite an elaborateorganizational architecture, there were periods when GEs governance system did not allow forcoordination of corporate and business unit agendas. Our theory proposes that the temporalcoupling of specialized, cross-level channels creates an organizational architecture that is bothdifferentiated and integrated. This architecture integrates levels and issues simultaneously,yet focuses attention sequentially, providing more effective conditions for joint attention andcoordination between the corporation office and the business unit and adaptive change at thebusiness unit level. Copyright 2012 John Wiley & Sons, Ltd.

    INTRODUCTION

    This article examines a critical issue in strategicmanagement: how the organizational architectureof a diversified multibusiness firm facilitates thestrategic adaptation of its business units (Rumelt,Schendel, and Teece, 1994). In particular, ouranalysis focuses on the adaptive implications ofcorporate-business unit interactions. Beginningwith Chandler (1962), the literature on diversi-fied firms focused on the superior information-processing and decision-making capacity of themultidivisional structure, with corporate office

    Keywords: attention; strategic adaptation; governancechannels; cross-level; interactions Correspondence to: John Joseph, Duke University 100 FuquaDr., Durham, NC 27708, U.S.A.E-mail: [email protected]

    control of strategy formulation and business unitfocus on implementation and operations (Chandler,1962; Stinchcombe, 1990). Williamson (1975,1983) expanded upon Chandler to develop the M-form hypotheses, where the corporate-level focuson strategy is accompanied by a financial con-trol system. The Chandler-Williamson perspectiveassumed a clear differentiation between businessunit operations and corporate strategy, with lim-ited attention to integration beyond authority rela-tionships (cf. Lawrence and Lorsch, 1967; Gulati,Lawrence, and Puranam, 2005) and financial con-trol (Williamson, 1975).

    A large body of research has recognized thatbusiness unit adaptation is not merely a functionof corporate fiat, but the consequence of a moreelaborate relationship between corporate headquar-ters and business units (Gupta, 1987; Ghoshal

    Copyright 2012 John Wiley & Sons, Ltd.

  • 634 J. Joseph and W. Ocasio

    and Nohria, 1989; Bartlett and Ghoshal, 1993;Volberda, Baden-Fuller, and van den Bosch, 2001).For example, horizontal linkages, such as cross-functional or cross-business unit teams (Galbraith,1974; Martin and Eisenhardt, 2010; Martin, 2011),provide business unit managers with opportuni-ties for adaptation through collaboration, as wellas economies of scope through greater knowledgeexchange and resource sharing. The corporate roleis one of providing the underlying support, such astotal quality management or new product develop-ment processes, for horizontal exchanges to takeplace and interdependencies to add value (Prahaladand Doz, 1987; Bartlett and Ghoshal, 1993).

    Another stream of research has examined therole of corporate executives in shaping the prod-uct markets in which business units compete. Inparticular, the literature on patching and recom-bination (Galunic and Eisenhardt, 1996, 2001;Rindova and Kotha, 2001; Birkinshaw and Ling-blad, 2005) finds distinct advantages of modularcorporate forms, with their capacity to add, divestand recombine businesses through charter changesin the firms business units. One particular insightof these studies is that while the corporate officemay influence charter decisions, much of the activ-ity concerning charter assignments reflects decen-tralized behavior. This challenges the idea thatstrategic adaptation is the province of the CEOand the corporate office. It is also consistent withFreeland (1996) who finds, in a direct examina-tion of Chandler and Williamson, that contrary tothe M-form hypotheses, business unit participa-tion in strategic formulation was not only prevalentat General Motors, but also yielded better perfor-mance than when it was solely under the purviewof the corporate office.

    A third stream of research emphasizes the cor-porate offices provision of decision premises formanagers at lower levels of the firm. These stud-ies focus on the cognitive control of the organi-zation through corporate goals, plans, frames, orrules that guide exploratory behavior (Burgelman,1983, Gilbert, 2006) and business unit initiatives,which often depart from the mainstream business(Birkinshaw, 1997; Birkinshaw, Hood, and John-son, 1998). Research in this category often focuseson the performance implications of top manage-ment attention. The idea of this research is that inorder for firms to commit to new activities, suffi-cient top management attention must be directedto them to gain the necessary political support,

    visibility, and resources to survive. For example,research has shown that the attention focus of topmanagement can increase firm receptivity to newtechnologies (Kaplan, 2008), shape the timing ofmarket entry (Burgelman, 1991), and amplify firmcapabilities (Eggers and Kaplan, 2009).

    In all, these studies have made great strides inupdating the Chandler-Williamson view of corpo-rate control over strategic adaptation. In the pro-cess, however, the literature has moved away fromthe foundational focus concerning the structuralrelationship between the corporate office and thebusiness units. Studies that have explicitly dealtwith the role of the corporate office (e.g., Goold,Alexander, and Campbell, 1994; Freeland, 1996)do not focus on issues of structure or considerthe role of organizational architecture. Similarly,research that highlights cognitive drivers of adap-tive behavior typically does so at the expense ofstructural mechanisms. For example, the idea thatattention mediates the relationship between topmanagement mental models and strategic actionhas been well established (Barr, Stimpert, andHuff, 1992; Cho and Hambrick, 2006), but its rolein the relationship between organizational architec-ture and adaptation remains an open question.

    With renewed interest in corporate-level effectson performance (Bowman and Helfat, 2001; Adnerand Helfat, 2003; Martin and Eisenhardt, 2010)and greater focus on organizational architecture(Gulati, Puranam, and Tushman, 2012, this issue)and corporate attention (Cho and Hambrick, 2006;Ambos and Birkinshaw, 2010), a closer reexami-nation of the corporate-business unit relationshipin diversified firms is needed. The current articlereturns to the structural foundations established inthe early literature and explores, with an attentionallens, how structural relations between corporateoffices and business units shape strategic businessunit adaptation. We complement prior focus onvertical authority relationships (Galbraith, 1973)and horizontal linkages (Gupta and Govindara-jan, 1991; Martin and Eisenhardt, 2010) and focuson exploring the effects of alternative cross-levelstructures, the attentional focus created, and theirrole in adaptation. Specifically, our study seeks toanswer two questions. First, given the importanceof both the corporate office and business units forstrategic adaptation, how does the organizationalarchitecture, defined as the structure of commu-nications, interactions, and authority relationshipsacross the organizational structure (c.f., Simon,

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 635

    1962), achieve successful alignment of the corpo-rate office and business unit for identifying com-petitive opportunities and threats? Second, howdoes the differentiation and integration of corpo-rate and business units through the organizationalarchitecture affect the coordination of responses tothreats and opportunities?

    Theoretically, we adopt an attentional perspec-tive on organizational architecture (March andSimon, 1958; March and Olsen, 1976; Ocasio andJoseph, 2005; Bouquet and Birkinshaw, 2008).This perspective suggests that the organizationalarchitecture structurally distributes managerialattention throughout the firm, with managers withinvarious subunits and organizational levels focusingattention on different aspects of the firms agenda.This distribution occurs because managerial atten-tion is situated within the firms governance chan-nels. We define governance channels as formalcollective interactions set up by the firm to con-trol, allocate and monitor organizational attentionand resources. The fact that attention is situated ingovernance channels implies that managers atten-tion is conditioned on the interactions betweenthem, both within channels and across the systemof channels. Since the interactions will vary fromchannel to channel, depending on factors such aswho is in attendance, their timing, frequency, andthe agenda, so too will the way in which attentionis engaged and whether or not joint attention isestablished.

    Because of the emphasis on situated attention,our approach to corporate-business unit interac-tions differs from information processing per-spectives (Galbraith, 1974; Tushman and Nadler,1978). From an attention-based view, the focusis not on information processing capacity, butwhether and how available information is attendedto in a particular time and place. Correspondingly,we do not presume the presence of ubiquitousdecision rules, goals, or culture (c.f. Burgelman,1991; Bourgeois and Brodwin, 1984), but that theparameters for directing attention and decisions arenot widely shared and are contingent on particularconfigurations of channels. Also, while much ofthe foundational work implicitly equates the merepresence of communication channels with integra-tion, we do not assume that channels necessarilyintegrate attention. Channels may also differentiateor focus attention.

    Our research approach is based on analyticalinduction (Hicks, 1994). We begin with a general

    theoretical frameworkthat organizational archi-tecture shapes adaptation through its effects onattentionand seek to develop specific theorybased on this general framework. Our longitudinalanalysis allows us to examine how organizationalarchitecture, which is relatively static, affects anumber of specific instances of business unit adap-tation (or lack thereof). For our inductive method,we selected a historical case analysis of GeneralElectrics (GEs) changing organizational archi-tecture from 1951 to 2001 and its impact on thestrategic adaptation of two business units: Comput-ers and Medical Systems. GE is an extreme caseof a highly diversified corporation with significantvariation in both organizational architecture andsuccessful business unit adaptation: the Computersunit was considered a failure and was ultimatelydivested; Medical Systems emerged from a smallpart of GEs portfolio to become a major sourceof revenue and profitability. In the 1951 to 2001period, GE experienced four significant shifts inits organizational architecture and multiple eventsthat reflect successful and unsuccessful attempts atstrategic adaptation in the two businesses studied.By utilizing an embedded design of key events,within a single, critical firm, we seek to rely onwithin-case analysis of GE and between-case anal-ysis of events (Eisenhardt, 1989; Yin, 2009) to helpdevelop new theory on the architecture, attention,and adaptation of diversified corporations.

    Our focus on corporate-business unit interac-tions parallels, yet is distinct from, state-of-the-artwork on collaboration and coordination in multi-business firms (e.g., Smith and Tushman, 2005;Martin and Eisnhardt, 2010; Martin, 2011). Mostprior studies have examined firms with a domi-nant business or related diversifiers (c.f. Burgel-man, 1983). However, we focus our analysis ona highly diversified conglomerate where the focusof attention is highly distributed and coordinationacross functions, levels and units may be especiallychallenging. And whereas these papers have pri-marily elaborated on the nature of horizontal inter-actions within the executive office and betweenbusiness units, we seek to highlight the relativelyless understood effects of collective vertical inter-actions across corporate headquarters and businessunits on attention within particular governancechannels. We consider not only the attention inte-gration and differentiation that results from interac-tions within specific governance channels, but alsothe linkages across governance channels and the

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 636 J. Joseph and W. Ocasio

    transmission of issues and initiatives from one toanother. Specifically, we identify and analyze threetypes of channel integration: cross-level (withinchannel integration of corporate and business unitagendas), cross-functional (within channel integra-tion of various functional areas), and channel cou-pling (integration across the system of channels).

    Our study complements research on headquar-ters role in the reconfiguration of resources byhighlighting the structural direction of manage-rial attention. While recent research has examineddifferentiation and integration in vertical supplier-buyer relationships (Gulati et al., 2005), theunderlying architectural (i.e., structural) integra-tion between the corporate office and its divisionsin multibusiness firms remain relatively unexam-ined. Miller (1986) is a partial exception, notingthe importance of formal committees via plans andbudgets as an integrating mechanism. The struc-ture and content of the integrating committees andtheir relationship to organizational adaptation isnot, however, examined.

    Accordingly, we also contribute to the liter-ature on organizational attention. Recent empiricalwork has shown the importance of top man-agement attention on strategic adaptation (Choand Hambrick, 2006; Kaplan, 2008; Nadkarni andBarr, 2008; Eggers and Kaplan, 2009; Ambos andBirkinshaw, 2010). This literature has not consid-ered, however, whether and how top managementattention is integrated with business unit attentionor how organizational architecture facilitates busi-ness unit adaptation. Through our study at GE,we demonstrate that the integration of managerialattention affects the firms ability to respond tothreats and opportunities and that lack of integra-tion within and across channels may have negativeconsequences for strategic adaptation.

    DIFFERENTIATION ANDINTEGRATION IN THEMULTIBUSINESS FIRM

    The organizational architecture is a critical deter-minant of managerial information processingcapacity and strategic adaptation (Lawrence andLorsch, 1967; Tushman and Nadler, 1978; Miller,1986; Siggelkow and Levinthal, 2003). FollowingChakravarthy (1982), we define strategic adapta-tion as the coordinated responses to competitivethreats as well as the proactive pursuit of new

    opportunities. Strategic adaptation may be con-strained by a number of factors such as the rigidityassociated with established routines and patterns ofresource allocation (Leonard-Barton, 1992; Nelsonand Winter, 1982; Gilbert, 2005) or the cognitiveimpediments that arise from divergent perceptionsof external shocks (Meyer, 1982), technologicalchange (Tripsas and Gavetti, 2000), or competitiveactivity (Reger and Palmer, 1996). Our focus is athird category, the information processing limits ofdecision makers as shaped by the organizationalarchitecture (March and Simon, 1958; Galbraith,1974; Nadler and Tushman, 1997).

    The original literature on M-form firms high-lighted how the segmentation of decisionresponsibility between strategy and implemen-tation deepens the information processing anddecision-making capacity of corporate executives(Chandler, 1962) and focuses line managers atten-tion on unique markets and facilitates the develop-ment of products that are tailored to meet marketneeds. However, given the limits of structural dif-ferentiation between strategy and operations forstrategic adaptation, multibusiness firms requirespecialized structures for integration across lev-els and functions. Lawrence and Lorsch (1967)established both differentiation and integration offirms information processing capacity as criti-cal determinants of performance. Differentiationwas defined as the state of segmentation of theorganizational system into subsystems, each ofwhich tends to develop particular attributes inrelation to the requirements posed by its rele-vant external environment (Lawrence and Lorsch,1967: 4). Integration was the quality of the stateof collaboration that exists among departmentsthat are required to achieve unity of effort bythe demands of the environment (Lawrence andLorsch, 1967: 11). Departmental (and divisional)specialization was identified as a key to differ-entiation. Integration may be achieved throughvarious mechanisms, such as dedicated roles,organizational goals, and interactions within cross-functional or cross-business unit teams. Cross-business unit teams, for example, may integratemultiple perspectives by fostering trust, improvinginformation sharing, and providing members withopportunities to navigate their differences (Martin,2011). Interactions within top management helpresolve conflicting agendas (Cho and Hambrick,2006) and balance exploration and exploitation(Smith and Tushman, 2005).

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 637

    Governance channels, collective verticalinteractions, and situated attention

    Our conception of integration in this article issimilar to, but distinct from, previous definitionsin several respects. In this study, integration refersto the alignment of managerial attention to com-mon issues, initiatives, and decisions (i.e.,agendas). The attention-based view of the firm(Ocasio, 1997; Ocasio and Joseph, 2005; Barnett,2008; Bouquet and Birkinshaw, 2008) provides thetheoretical framework for our approach and theunderlying mechanism of how organizationalarchitecture shapes the differentiation and integra-tion of managerial attention required for strate-gic adaptation. Attention is defined as the notic-ing, encoding, interpreting, and focusing of timeand effort by organizational decision makers onboth (1) issues: the available repertoire of cat-egories for making sense of the environment;and (2) answers: the available repertoire of actionalternatives (Ocasio, 1997: 189). The issues andanswers together constitute the corporationsagenda and are central to adaptation and change(Dutton, 1997).

    Within the context of distributed attention thatis characteristic of the multibusiness firm, organi-zational architecture becomes a particularly impor-tant mechanism to integrate the focus ofattention on the corporations strategic agenda.Integration, in this context, reflects the collectivevertical interactions between corporate and busi-ness unit managers that occur within governancechannels and the issues and initiatives that areaccordingly attended to as managers communicateand make discrete decisions that involve organi-zational resources (Dutton and Ashford, 1993).The concept of channels played an important rolein early administrative theory (March and Simon,1958; Simon, 1957; Thompson, 1967; Allison,1969) and is considered a key mechanism forprocessing information (Galbraith, 1973; Marchand Simon, 1958; Henderson and Clark, 1990;Gulati et al., 2005) and transferring and integratingknowledge (Miller, 1986; Henderson and Clark,1990; Puranam and Srikanth, 2007). They are alsocentral in focusing the attention of organizationaldecision makers (Ocasio, 1997). Governance chan-nels include strategic reviews, audits, personnelreviews, ad hoc committee meetings, and budgetforecasts as well as the Chief Executive Office.Governance channels reflect collective vertical

    interactions beyond vertical reporting relation-ships highlighted in transaction cost theory(Williamson, 1975) or dyadic exchange commonto social network perspectives (Hansen andHaas, 2001).

    An interaction lens recognizes that the atten-tion focus of decision makers is situated(Ocasio, 1997) and occurs at the level of the chan-nel, guided by the particular procedural, temporal,and spatial characteristics of the channel (Stinch-combe, 1968). Accordingly, the interactions thatoccur through attention-directing channels reflecta host of contextual factors such as the issueson the agenda (Dutton, 1997), standard operat-ing procedures and routines (Nelson and Win-ter, 1982), group norms, and the participation ofkey organizational actors (e.g., CEOs and CFOs),as well as their frequency (annually, monthly),sequence (early or late in the year), and location(at headquarters, in the field). In the multibusi-ness firm, interactions within governance channelsallows corporate managers to prioritize strategicissues and initiatives for the explicit purpose ofallocating resources and governing the corporation(Ocasio and Joseph, 2005), but also accommo-dates emergent issues and initiatives within thebusiness unit. It is the situated joint attention cre-ated within governance channels that generatesthe decision-making patterns necessary to identifyopportunities and respond to competitive threatsin a coordinated mannerand, therefore, servesas the basis for adaptive moves of the M-formfirm. Our research explores how the structure ofgovernance channels, a key characteristic of orga-nizational architecture, shapes the adaptation ofbusiness units, with particular focus on how chan-nels integrate and focus business unit and corporateagendas.

    METHODS AND DATA

    To examine the effects of the corporate architec-ture on attention and adaptation, we conducted anin-depth case study of GEs governance systemfrom 1951 to 2001. Case studies allow researchersto get close to the theoretical constructs and causalforces of interest (Siggelkow, 2007) and are effec-tive when the focus of the study is to understandor answer how questions (Eisenhardt and Graeb-ner, 2007), which are the focus of the presentstudy. As a setting for our research, GE offered

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 638 J. Joseph and W. Ocasio

    several advantages. First, GE is an important yetanomalous case for strategic management research.As a highly diversified conglomerate with highlevels of sustained profitability, it defies conven-tional wisdom on the limits of unrelated diversifi-cation (Rumelt, 1982). Second, GE is an extremecase, where the theoretical mechanisms that arethe focus of this investigation are especially trans-parent and well documented (Eisenhardt, 1989). Inaddition, GE offers significant longitudinal varia-tion in the organizational architecture and channelstructureincluding what GE has referred to atvarious times as its strategic planning system orits operating system (Ocasio and Joseph, 2008).Third, public information on the firm is avail-able over a long period to allow the study of keychannels in detail without having to assume theircontext or characteristics.

    To reconstruct GEs system of governance chan-nels and their effects on the firms adaptive behav-ior, we utilized a longitudinal embedded casedesign that allows for multiple levels of analy-sis (Eisenhardt, 1989; Yin, 2009). In our study,we identified four distinct organizational architec-tures at GE195163, 196371, 197186, and19862001allowing for longitudinal variationin our analysis. We selected two business unitsthat could be examined across periods, and weanalyzed attention focus at the channel level, ourprimary level of analysis. This allowed us to utilizea replication logic, whereby each subunit was usedto test emergent theoretical insights (Yin, 2009).Our basic unit of analysis is event based, consis-tent with other empirical studies of attention (e.g.,Hoffman and Ocasio, 2001). Events are embeddedin business units, which are themselves embeddedin one of four corporate architectures, all part ofa single firm. In our embedded design, the single-firm setting allowed for variation of the conditionsof interest, while controlling for firm factors suchas history, incentives, and top management. Weexamine, for each key event, the central channelfor strategy formulation as well as changes to thechannel configurations, and we observe variationsin attention and adaptation.

    Data for the study is drawn from several sources:public documents, including annual reports, proxystatements, and speeches; historical documentsfrom the GE corporate archives in Schenectady,New York, including board of directors minutes,executive correspondence, corporate newsletters,and organizational charts and directories; historical

    documents from archives on GEs computer busi-ness at the University of Minnesota; newspapers,magazines, and books published from 1951 to2001 that concern GE or one of its CEOs, sub-sidiaries, or operational units; and 16 interviewswith past and present GE employees.

    Data analysis

    First, we reviewed each period in detail andhighlighted key channels, the corporate agendasattended to within each channel, and primary link-ages to other channels. Utilizing Chandler (1977,1991) as a guide, we initially categorized the chan-nels as headquarters, line, or staff channels basedon the executives who participated in the chan-nel and the types of decisions made within thechannel. Headquarters channels are interactionswithin the top management team, for formulat-ing strategy, allocating resources, and governingthe corporation. Line channels are vertical interac-tions between managers with a formal authority orreporting relationship (c.f. Williamson, 1975). Staffchannels are those where corporate staff interactto address the functional aspects of the business.In addition, our analysis identified a fourth cat-egory: cross-level channels. Cross-level channelsinvolved interactions among corporate executive,business unit managers, and corporate staff. Bothline channels and cross-level channels are verticalgovernance channels. Cross-level channels differfrom line channels or vertical reporting relation-ships in that the latter involves dyadic interactionsbetween managers and subordinates where atten-tion is directed from the top down. Moreover,interactions within cross-level channels are col-lective, not dyadic, and involve both bottom-upand top-down influence on attention and decisions.Table 1 lists the channel types for each period.

    We also categorized the corporate agendas at-tended to within channels into the following fourfunctions: planning, finance, operations, andhuman resources. Channels that are specializedhave a focused agenda. Within these channels,managers attend primarily to the set of issuesof a particular function. Integrated channels havebroader agendas covering multiple functions, withmore than one foci of attention. Figures 14 con-tain the organizational architecture for each of theperiods. The transmission of issues and initiativesbetween channels determines whether the channels

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 639

    Table 1. Channel types

    Decentralization(195163)

    Corporate planning(196371)

    Strategic planning(197185)

    Operating system(19862001)

    HQ Office of the President Chief Exec. Office Chief Exec. Office Chief Exec. OfficeAdvisory Council Corp. Policy Board Corp. Policy Board

    Sector PlanningStaff Auditing Auditing Auditing Auditing

    Service Departments Corporate Exec. Staff Corporate Exec. StaffCorporate Adm. Staff Corporate Adm. StaffManpower ReviewGrowth Council SBU Planning

    Crotonville Crotonville Crotonville CrotonvilleShangri-La

    Corp. Operations. BoardCorp. Exec. Comm. Corp. Exec. Comm.

    Cross-level Financial reporting Financial reporting Financial reporting Financial reportingBellair BocaSession I Session I

    Annual Bus. Review Session II Session IISession C Session C

    PresidentGroup CEOGroup CEOSector CEOBusiness UnitLine GroupDivision GroupDivision SectorGroup

    DivisionDept DivisionDept GroupDivisionDivisionDept

    indicates cross-functional channels; indicates specialized channels.

    President

    Advisory Council*PLAN

    CRO

    SS-L

    EVEL

    HQ

    Department*PLAN, FIN, HR, OPS

    CROSS -LEVEL LINE STAFF

    Auditing*FIN

    Division*PLAN, FIN, HR, OPS

    Office of President*PLAN, FIN

    Annual Bus. Review*PLAN, FIN, HR, OPS

    Services*PLAN

    Crotonville*HR

    Group*PLAN, FIN

    Financial reporting*FIN

    Figure 1. Organizational architecture (19511952)Decentralization

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 640 J. Joseph and W. Ocasio

    President

    CRO

    SS-L

    EVEL

    HQ

    Department*PLAN, FIN, HR, OPS

    CROSS -LEVEL LINE STAFF

    Crotonville*HR

    Auditing*FIN

    Division*PLAN, FIN, HR, OPS

    Chief Exec. Office*PLAN, FIN

    Manpower Review*HR

    Growth Council*PLANGroup

    *PLAN, FIN, HR, OPS

    Corp. Exec. Staff *PLAN

    Corp. Adm. Staff*OPS

    Financial reporting*FIN

    Annual Bus. Review*PLAN, FIN, HR, OPS

    Corp. Policy Board *PLAN

    Shangri-La*PLAN

    Figure 2. Organizational architecture (19671968)Corporate planning

    CEO

    CRO

    SS-L

    EVEL

    HQ

    Department*PLAN, FIN, HR, OPS

    CROSS -LEVEL LINE STAFF

    Crotonville*HR

    Financial reporting*FIN

    Auditing*FIN

    Division*PLAN, FIN, HR, OPS

    Chief Exec. Office*PLAN, FIN

    Session II*PLAN, FIN, OPS

    Group*PLAN, FIN, HR, OPS

    Corp. Exec. Staff*PLAN

    Corp. Adm. Staff*OPS

    Belleaire*PLAN SBU Planning

    *PLAN

    Sector*PLAN, FIN, HR, OPS

    Session I*PLAN, FIN

    Session C*HR

    Corp. Policy Board*PLAN

    Figure 3. Organizational architecture (19771978)Strategic planningCopyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)

    DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 641

    CEO

    CRO

    SS-L

    EVEL

    HQ

    Department*OPS

    CROSS -LEVEL LINE STAFF

    Crotonville*HR

    Financial reporting*FIN

    Auditing*FIN

    Division*OPS

    Session II*PLAN, FIN, OPS

    Business Unit*PLAN, FIN, HR, OPS

    Session I*PLAN, FIN, OPS

    Session C*PLAN, FIN, OPS, HR

    Boca*PLAN

    Chief Exec. Office*PLAN, FIN

    Corp. Exec. Council*PLAN

    Figure 4. Organizational architecture (19951996)Operating system

    are linked in the corporate architecture. For exam-ple, in the first period, the Services channel wasfocused on planning, and the issues attended toin that channel served as an input to the AdvisoryCouncil and the Annual Business Review, creatinga link between Services and the Advisory Council,and Services and the Annual Business Review.

    In the next stage of research, we returned tothe data and identified two business units thatcould provide deeper insights into the integra-tion of channels. The businessesComputers andMedical Systemswere selected based on the fol-lowing criteria: (1) the business was widely rec-ognized through secondary sources as a majorone during GEs history; (2) the business spannedat least two distinct channel configurations; and(3) sufficient data on the business was availablefor analysis. The business units were within indus-tries at different stages of their life cycles, withan established incumbent in computers (IBM), butrelatively little competition in medical imaging(where GE was the dominant player in the U.S.).For each of the business units, we identified keyevents, our unit of analysis. Key events generatedopportunities or threats to the business unit and

    received significant attention by decision makers.We used a theoretical sampling approach (Eisen-hardt, 1989), which requires selecting events thatare theoretically useful and reflect relevant cate-gories. In this case, the categories were strategicopportunities and threats and the type of chan-nels involved in decision making (allowing forvariation across events in the success of strategicadaptation). A total of 10 events were identified:five for the Computer business and five for Med-ical Systems. In the third stage of our research,we conducted interviews with 16 current and for-mer GE employees. Twelve of the interviews wereconducted with different members of the GE Com-puter Department Alumni Association, and threeof the interviews were conducted with current andformer senior Medical Systems managers (nowGE Healthcare), including Walter Robb, the for-mer chief who oversaw the entry into CT scan-ners and MRI. We also conducted interviews withthe GE archivist in Schenectady, New York. Thesemistructured interviews lasted 60 to 120 min-utes and followed an interview guide directing thesubject to focus on the facts of each event. Thepurpose of these interviews was to review and

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 642 J. Joseph and W. Ocasio

    deepen our data analysis concerning the events andthe channels in which they were made, the contentof the interactions more generally, and the flow ofissues across channels.

    In what follows, we analyze the governancechannels prevalent during each of the key peri-ods and their impact on business unit adaptivemoves. We track key changes to the organiza-tional architecture and the channels, paying closeattention to the type of channels used for strat-egy formulation. We highlight the key events foreach period to illustrate how (and whether) thechannels integrated corporate and business unitattention focus. Finally, we examine how adaptivemoves were affected. Identification of successfulstrategic adaptation frequently depends on the timeframe considered (for example, the Computer ven-ture was initially successful, but unsuccessful inthe long term). Therefore, we focus our analysis onwhether adaptive moves were taken or not takenin response to attention patterns, rather than theiroverall impact on financial performance, thoughwe address the performance issue in the discussionsection. We use the historical analysis to gener-ate theoretical propositions on how organizationalarchitecture shapes business unit adaptation.

    ORGANIZATIONAL ARCHITECTURESAT GENERAL ELECTRIC

    The General Electric Company, incorporated in1892, was from its early days a diversified firmwith lines of business in lighting, power trans-formers, and a wide variety of electrical compo-nents for industrial, commercial, and residentialuse. From the period studied, GEs governancesystem was characterized by four distinct organi-zational architectures, each with its own structureof channels. We characterize the four architecturesas: (1) decentralization (195163), (2) corporateplanning (196371), (3) strategic planning (197186), and (4) operating system (19862001).

    DecentralizationIn 1951, Ralph J. Cordiner, president of GE, im-plemented a radical decentralization of thecorporation. Motivated by the need for greatercontrol of the sprawling organization, departmentsbecame the basic operating unit of the organiza-tionwith a full staff, a clear charter of business

    (and profit) responsibility, and limited horizontalcoordination. Departments reported to divisions,divisions reported to groups, and groups reportedto Cordiner.

    To govern the decentralized organizationalarchitecture, Cordiner established a series of gov-ernance channels to link operating units decisionsdirectly with headquarters. To introduce GEsmiddle manager ranks to GEs management phi-losophy and techniques, Cordiner created the GEManagement Development Institute, a staff chan-nel in Crotonville, New York. To closely monitoroperating performance, Cordiner utilized monthlyfinancial reports, prepared by department financemanagers, reviewed and approved by the depart-mental GM, and distributed to the division GM,group executive, comptroller, and president, mak-ing it a cross-level channel. Planning was carriedout, in part, by the Office of the President, aheadquarters channel that included Cordiner, groupVPs, and service (i.e., corporate staff) VPs. Sofar, the decentralized organizational architecturecorresponds to differentiation of strategy and oper-ations in the M-form, as originally described byChandler (1962, 1991) and Williamson (1971).But another critical channel, which facilitated inte-gration across corporate and business levelstheAnnual Business Reviewdiverged from thismodel. This cross-level channel brought the depart-ment staff, president, group executives, and serviceVPs together for one- or two-day meetingseitherat corporate headquarters at 570 Lexington Avenuein New Yorkor at the department headquarters,to review all functions of the business unit.

    Corporate planning

    Fred Borch, who succeeded Cordiner in 1963, car-ried out a major reorganization that more thandoubled the number of groups, divisions, anddepartments. He also altered the system of gov-ernance channels to accommodate a greater focuson corporate planning, with a greater separationof strategy at the corporate level from operationsat the business unit level. When Borch took overas CEO, GEs sales and profits had been rel-atively stagnant for five years. Senior manage-ment believed that future growth would requirespecialized and sustained corporate-level atten-tion to planning and diversifying new ventures.Correspondingly, there was a belief that venture

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 643

    planning should be separate from the regular oper-ations of the business.

    To separate planning from day-to-day opera-tions, Borch established numerous headquarterschannels, each designed to deal with specificstrategic issuesnone of which included opera-tional responsibility. He created the Chief Exec-utive Office (CXO), which focused attention onthe development and implementation of corporatepolicies. Cordiners Advisory Council evolved intothe Corporate Policy Board (CPB), which includedthe CXO and corporate staff VPs, but not thegroup VPs. The CPB was a headquarters chan-nel for reviewing and communicating staff mattersof broad corporate interest. Borch established twoadditional staff channels: the Corporate Adminis-trative Staff, charged with handling the day-to-daycorporate staff work (such as accounting), and theCorporate Executive Staff (CES) to focus on thelong-run strategic direction and new ventures ofthe company. The CES most notably included thehead of planning and R&D, but, like other head-quarters channels, had no direct participation fromgroup or division managers. His most significantaddition to the staff channels was the GrowthCouncil, chartered with directing the Companysgrowth and led by a newly created Manager ofCorporate Planning. Borchs channel structure dur-ing this period most clearly followed Williamsons(1975) model of vertical separation of strategyfrom operations.

    Strategic planningIn the late-1960s, to establish a more strategicfocus for planning, Borch engaged both the BostonConsulting Group and McKinsey in developingnew organizational architecture for what becameknown as a strategic planning system, includingthe strategic business units (SBUs). The SBU wasa parallel organizational structure superimposed onthe existing group-division-department organiza-tional reporting structure that focused attention onstrategic planning. SBUs were designations by theCXO regarding whether the specific department,division, or group was the correct business unitlevel at which competition took place. Diversifica-tion options were evaluated in terms of relationshipto the SBU core business and attractiveness of theindustry, rather than through problem solving atthe department level or separate growth planningat the corporate level.

    Reginald H. Jones, who followed Borch as CEOin 1971, refined the SBU planning organizationalarchitecture. To gain better control over the actionsof the divisions and business units, Jones moretightly coupled strategic planning to budgetingand operations. He made his three vice chairmenresponsible for the plans and financial results ofall units reporting to them (Vaghefi and Huellman-tel, 1998). Later, he would use the sector levelto fulfill the same function. A key developmenthere was the emphasis on the portfolio planningmodel, which included pruning as well as growingbusiness unit components. Sector and group headswere ultimately responsible for both plans and per-formance and exerted more weight in divisionaldecisions.

    Jones increased both staff and cross-level chan-nels in the organizational architecture. At head-quarters, decision making was coordinated throughthe CES, the Sector line of management, andthe CPB, a critical linkage between staff andline (as between sector heads themselves). Facedwith the segmentation of strategic planning, finan-cial reviews, and operational decision making,Jones created a new set of corporate-business unitcross-level channels, augmenting SBU planning.The Annual Business Review was reconstitutedas Executive Briefing Session I and II, yet stillserved as key cross-level channels. Session I inte-grated planning and budgeting, while Session IIwas primarily focused on financial objectives forthe upcoming year. Manpower reviews were con-stituted as Session C, a cross-level channel forplanning and evaluating managerial leaders. Thenew strategic reviews were distributed through-out the year and served to temporally link strate-gic planning with operational and human resourceplanning.

    Operating system

    In 1981, John F. Welch, Jr., became CEO andChairman of GE. In 1985, Welch centralizedthe corporate reporting structure, eliminating thesectors, groups, and SBUs, and he established15 businesses as the basic operating units reportingdirectly to him. Complex planning documents wereno longer required. Rather than bottom-up strategicplanning through the SBUs, planning was guidedby a corporate agenda comprised of key initiatives.Welch built on the prior system of cross-levelchannels to create what GE called its Operating

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 644 J. Joseph and W. Ocasio

    System (General Electric Annual Report, 1999)or social architecture (Welch, 2001). The Operat-ing System began with an annual retreat at BocaRaton, Florida, and was followed by a year-roundsystem of reviews. Welch noted that we poundedall our initiatives in January at Boca, at every quar-terly CEC meeting, at the human resources reviewin April (Session C), the planning sessions in July(Session I), the officers meeting in October, and theoperating plan meeting in November (Session II)(Welch, 2001: 299). Training to support initiativeswas done at Crotonville.

    ORGANIZATIONAL ARCHITECTUREAND STRATEGIC ADAPTATION

    We present our comparative analysis of 10 keyevents in Tables 25. Each of the tables reflectsdistinct features of the organizational architectureand their relationship to the process of strategicadaptation. The five events in the Computer busi-ness studied are: (C1) the financial and technicalsuccess of the Bank of America (BoA) computersystem and demands from banks and other poten-tial customers for similar products (1959); (C2)significant financial losses and delays in prod-uct development (1962); (C3) stagnant corporategrowth and profitability, which received renewedattention as Borch took over as CEO in 1963;(C4) strong growth and, for the first time, posi-tive net income in the business unit (1969); and(C5) the submission of the Advanced ProductLine (APL) to corporate headquarters and theU.S. governments restriction of GEs purchase ofanother computer manufacturer due to anti-trustconcerns (1969). In the Medical Systems business,the events studied are: (MS1) stagnant corporategrowth and profitability mentioned earlier (1963);(MS2) development of computerized tomography(CT) technology and entry of new competitors(1974); (MS3) increasing foreign competitiveness(197374); (MS4) development of magnetic res-onance imaging (MRI) technology (1980); and(MS5) quality problems and peer firm adoption ofSix Sigma (1995). Note that these events includeboth opportunities and threats to the Computerand Medical System businesses, but were oftenperceived somewhat differently by corporate head-quarters, as will be discussed.

    Cross-level channels

    Table 2 compares the primary channel throughwhich strategy formulation occurred and withinwhich perspectives on potential threats and oppor-tunities were debated. The Table includes the typeof channel, the attention to threats and opportu-nities within the channel, and how the interac-tions within the channel affected the alignment ofcorporate and business unit perspectives on mar-ket and technological opportunities. As Table 2indicates, strategy formulation was carried out incross-level channels in six of the 10 events stud-ied. In three events, staff channels were used, andone utilized a headquarters channel. Our induc-tive analysis found that within these cross-levelchannels, corporate and business unit managers,through open and frank dialogue, would establisha common focus of attention on particular threatsand opportunities and appropriate courses of actionto deal with them. The conversations at the AnnualBusiness Reviews were highly interactive in atalk out loud manner (Greenwood, 1982). Like-wise, Welch (2001: 193) described Session C asfun brawls: aggressive, gossipy, brutally honestmeetings that brought the CXO together with thebusiness unit heads. These meeting were effectivein aligning perspectives because the interactionsencouraged debate, negotiation (Martin, 2011) andefforts to create a collective perspective (Kaplan,2008) in order to come to a decision on next steps.The performance benefits of constant group com-munication are well documented (e.g., Brown andEisenhardt, 1997; Martin and Eisenhardt, 2010),and the interactions between decision makers hasbeen shown to shape cognition (Gilbert, 2006). Inour analysis, we find that senior managementsperspectives did not always dominate, leaving agreat deal of discretion to business units in shapingthe collective perspective and mobilizing action infavor of their desired outcome.

    At the 1960 Annual Business Review, for exam-ple, conflicting perspectives on the next stepsfor the Computer business came together (C1)headquarters strategy to stay away from compet-ing with IBM in the mainstream computer busi-ness, and due to the financial and technical suc-cess of the BoA computer system, the ComputerDepartments plan to enter the broader businesscomputing market. Interactions during the AnnualBusiness Review, a cross-level channel, helpedalign these divergent perspectives on the market

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 645

    Tabl

    e2.

    Cros

    s-le

    velc

    hann

    els

    and

    alig

    nmen

    tofc

    orp

    orat

    e-bu

    sines

    su

    nit

    pers

    pect

    ives

    oft

    hrea

    tsan

    do

    ppor

    tuni

    ties

    Even

    tN

    o.Ev

    ents

    Key

    chan

    nel

    Chan

    nel

    type

    Atte

    ntio

    nto

    thre

    ats/o

    ppor

    tuni

    ties

    with

    inch

    anne

    lR

    esul

    ting

    alig

    nmen

    tacr

    oss

    level

    sA

    lignm

    ent

    achi

    eved

    C1Fi

    nanc

    iala

    nd

    tech

    nica

    lsucc

    ess

    of

    BoA

    syste

    man

    dde

    man

    dfo

    rsim

    ilarp

    rodu

    cts

    (1959

    )

    Ann

    ualB

    usin

    ess

    Rev

    iew

    Cros

    s-le

    vel

    Corp

    orat

    eH

    Qfo

    cus

    on

    thre

    atofI

    BM

    asa

    com

    petit

    or,l

    oss

    ofI

    BM

    ascu

    stom

    eran

    dvie

    wth

    atG

    Eoffe

    red

    no

    tech

    nica

    lcontr

    ibut

    ion;

    busin

    ess

    unit

    focu

    son

    RCA

    asco

    mpe

    titor

    ,abi

    lity

    ofG

    Eto

    finan

    ceven

    ture

    and

    opp

    ortu

    nity

    todi

    rect

    lyen

    ter

    busin

    ess

    mac

    hine

    s.

    Corp

    orat

    eH

    Qsh

    fits

    focu

    sfro

    mIB

    Mto

    RCA

    as

    prim

    ary

    com

    petit

    oran

    dfro

    mte

    chni

    cal

    contr

    ibut

    ion

    tofin

    anci

    ngas

    key

    adva

    ntag

    e.Co

    rpor

    ate

    HQ

    and

    BU

    focu

    sal

    igne

    don

    busin

    ess

    mac

    hine

    sopp

    ortu

    nity

    .

    Yes

    C2Si

    gnifi

    cant

    finan

    cial

    loss

    esan

    dde

    lays

    inpr

    oduc

    tdev

    elop

    men

    t(19

    62)

    Ann

    ualB

    usin

    ess

    Rev

    iew

    Cros

    s-le

    vel

    Corp

    orat

    eH

    Qfo

    cus

    on

    failu

    reofb

    usin

    ess

    unit

    tom

    eet

    finan

    cial

    goal

    s;bu

    sines

    sunit

    focu

    son

    IBM

    thre

    atan

    dte

    chno

    logy

    prob

    lem

    s(pe

    riphe

    ralsa

    nd

    softw

    are).

    Corp

    orat

    eH

    Qan

    dB

    Ufo

    cus

    alig

    ned

    on

    impr

    ovin

    gfin

    anci

    alpe

    rform

    ance

    .Ye

    s

    C3St

    agna

    ntco

    rpor

    ate

    grow

    than

    dpr

    ofita

    bilit

    y(19

    63

    1964

    )G

    row

    thCo

    unci

    lSt

    aff

    Corp

    orat

    eH

    Qfo

    cus

    on

    grow

    thth

    roug

    hnew

    ven

    ture

    s.Co

    rpor

    ate

    HQ

    focu

    ses

    on

    com

    pute

    rsas

    one

    key

    grow

    thar

    eaan

    din

    tern

    atio

    nala

    cqui

    stion

    asm

    eans

    toin

    crea

    sepr

    esen

    ce;p

    ays

    little

    atte

    ntio

    nto

    grow

    ing

    tech

    nica

    l(pe

    riphe

    ral,s

    oftw

    are)

    or

    finan

    cial

    (leasi

    ng)i

    ssue

    sin

    BU

    .

    No

    C4R

    enew

    alofv

    olu

    me

    ship

    men

    tsof

    600

    serie

    san

    dpo

    sitiv

    enet

    inco

    me

    (1969

    )

    Shan

    gri-L

    aSt

    aff

    Corp

    orat

    ean

    dB

    Upl

    anni

    ngst

    afff

    ocus

    on

    tech

    nolo

    gica

    lopp

    ortu

    nity

    with

    new

    prod

    ucts

    tole

    apfro

    ngIB

    Ms

    pric

    e-pe

    rform

    ance

    adva

    ntag

    e.

    Plan

    ners

    atco

    rpor

    ate

    and

    BU

    level

    sfo

    cus

    on

    plan

    forA

    dvan

    ced

    Prod

    uctL

    ine

    (APL

    )at

    aco

    stof

    $500

    mill

    ion.

    Yes

    C5A

    PLpl

    ansu

    bmitt

    ed;g

    ovt

    anit-

    trust

    const

    rain

    ts(19

    69)

    Ventu

    res

    Task

    Forc

    eH

    QCo

    rpor

    ate

    HQ

    focu

    son

    thre

    atto

    shor

    t-ter

    mea

    rnin

    gsan

    dopp

    ortu

    nity

    tore

    duce

    pres

    sure

    on

    cash

    flow

    and

    opp

    ortu

    nity

    sell

    the

    busin

    ess.

    Corp

    orat

    eH

    Qvie

    ws

    step

    snec

    essa

    ryto

    atta

    inN

    o.2

    posit

    ion

    too

    cost

    lyan

    din

    ves

    tigat

    essa

    le;p

    ays

    little

    atte

    ntio

    nto

    BU

    plan

    toen

    hanc

    e60

    0se

    ries.

    No

    MS1

    Stag

    nant

    corp

    orat

    egr

    owth

    and

    profi

    tabi

    lity

    (1963

    19

    64)

    Gro

    wth

    Coun

    cil

    Staf

    fCo

    rpor

    ate

    HQ

    focu

    son

    grow

    thth

    roug

    hnew

    ven

    ture

    s.Co

    rpor

    ate

    staf

    ffoc

    uses

    on

    opp

    ortu

    nitie

    sw

    ithin

    med

    ical

    serv

    ices

    ;pay

    slit

    tleat

    tent

    ion

    toB

    Ude

    velo

    pmen

    tofa

    rtifi

    cial

    org

    ans,

    nucl

    ear

    med

    icin

    e,or

    prol

    ifera

    tion

    ofX

    -ra

    yeq

    uipm

    ent.

    No

    MS2

    Tech

    nolo

    gica

    ldisc

    ontin

    uity

    from

    CT(19

    74)

    Sess

    ions

    I&

    IICr

    oss-

    leve

    lCo

    rpor

    ate

    HQ

    focu

    son

    opp

    ortu

    nity

    togr

    owth

    roug

    hCT

    tech

    nolo

    gy;B

    Ufo

    cus

    on

    thre

    atfro

    mnew

    entr

    ants

    .

    Corp

    orat

    eH

    Qan

    dB

    Ufo

    cus

    alig

    ned

    on

    cras

    hde

    velo

    pmen

    tpro

    gram

    ofC

    Tte

    chno

    logy

    inCR

    D.

    Yes

    MS3

    Incr

    easin

    gag

    gres

    siven

    ess

    and

    succ

    ess

    off

    orei

    gnco

    mpe

    titor

    s(19

    74

    1975

    )

    Sess

    ion

    ICr

    oss-

    leve

    lCo

    rpor

    ate

    HQ

    focu

    son

    thre

    atofi

    nter

    natio

    nal

    com

    petit

    ion;

    BU

    focu

    son

    dom

    estic

    mar

    ket.

    Corp

    orat

    eH

    Qfo

    cuse

    sB

    Uon

    iden

    fiyin

    gpa

    rtne

    rin

    Japa

    n.Co

    rpor

    ate

    HQ

    and

    BU

    focu

    sal

    igne

    don

    glob

    alpe

    rform

    ance

    and

    stra

    tegy

    tode

    alw

    ithfo

    reig

    nco

    mpe

    titor

    s.

    Yes

    MS4

    Tech

    nolo

    gica

    ldisc

    ontin

    uity

    from

    MR

    I(19

    80)

    Sess

    ions

    I&

    IICr

    oss-

    leve

    lCo

    rpor

    ate

    HQ

    focu

    son

    opp

    ortu

    nity

    togr

    owth

    roug

    hM

    RI

    tech

    nolo

    gy;B

    Ufo

    cus

    on

    thre

    atfro

    mm

    ulti

    ple

    com

    petit

    ors

    incl

    udin

    gPi

    cker

    ,Sie

    men

    s,J&

    J,an

    dPh

    ilips

    .

    Corp

    orat

    eH

    Qan

    dB

    Ufo

    cus

    alig

    ned

    on

    cras

    hde

    velo

    pmen

    tpro

    gram

    ofM

    RI

    tech

    nolo

    gyin

    CRD

    .

    Yes

    MS5

    Gro

    win

    giss

    ues

    ofq

    ualit

    yan

    dad

    optio

    nofS

    ixSi

    gma

    bype

    er

    firm

    s(19

    95)

    Boc

    a,Se

    ssio

    nsI,

    IIan

    dC

    Cros

    s-le

    vel

    Corp

    orat

    eH

    Qfo

    cus

    on

    opp

    ortu

    nity

    toin

    crea

    sequ

    ality

    thro

    ugh

    Six

    Sigm

    a;B

    Ufo

    cus

    on

    thre

    atfro

    mEl

    scin

    tin

    CTs

    and

    prob

    lem

    sw

    ithX

    -ray

    tube

    man

    ufa

    ctur

    ing.

    Corp

    orat

    eH

    Qan

    dB

    Ufo

    cus

    alig

    ned

    on

    Six

    Sigm

    ain

    itiat

    ive

    and

    sele

    ctio

    noft

    rialp

    rojec

    tsin

    Med

    ical

    Syste

    ms

    on

    whi

    chto

    roll

    out.

    Yes

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 646 J. Joseph and W. Ocasio

    Tabl

    e3.

    Spec

    ializ

    edch

    anne

    lsan

    dad

    aptiv

    em

    ov

    es

    Even

    tN

    o.K

    eych

    anne

    l/typ

    eSp

    ecia

    lizat

    ion

    Chan

    nela

    gend

    aCo

    rpor

    ate

    and

    BU

    stra

    tegi

    cfo

    cus

    Res

    ultin

    gad

    aptiv

    em

    ov

    es

    C1A

    nnua

    lBusin

    ess

    Rev

    iew

    (cros

    s-lev

    el)N

    oFi

    nanc

    ialp

    erfo

    rman

    ce,k

    eypr

    obl

    ems,

    mar

    kets,

    prod

    uctl

    ines

    ,app

    licat

    ions

    ,new

    prod

    ucts,

    long

    -rang

    efin

    anci

    alfo

    reca

    st

    No

    Ada

    ptat

    ion

    thro

    ugh

    entr

    yin

    toco

    mm

    erci

    alan

    dbu

    sines

    sm

    achi

    nes

    mar

    ket(

    1960

    )

    C2A

    nnua

    lBusin

    ess

    Rev

    iew

    (cros

    s-lev

    el)N

    oFi

    nanc

    ialp

    erfo

    rman

    ce,k

    eypr

    obl

    ems,

    mar

    kets,

    prod

    uctl

    ines

    ,app

    licat

    ions

    ,new

    prod

    ucts,

    long

    -rang

    efin

    anci

    alfo

    reca

    st

    No

    No

    adap

    tatio

    n;no

    chan

    gein

    finan

    cial

    repo

    rting

    toal

    low

    for

    IBM

    leas

    ing

    mode

    l;de

    partm

    entG

    Mre

    plac

    edan

    dgi

    ven

    man

    date

    totu

    rnar

    ound

    perfo

    rman

    ce;i

    ncre

    men

    talp

    rodu

    ctch

    ange

    (1963

    )C3

    Gro

    wth

    Coun

    cil(

    staff)

    Yes

    (plan

    ning)

    Gro

    wth

    opp

    ortu

    nitie

    sfo

    rG

    Eth

    atar

    egr

    ow

    ing

    faste

    rth

    anG

    NP

    and

    inw

    hich

    GE

    has

    unqu

    eca

    pabi

    litie

    san

    dm

    aybe

    able

    toga

    inco

    mpe

    titiv

    ead

    vant

    age

    No

    Ada

    ptat

    ion

    thro

    ugh

    acqu

    isitio

    no

    fOliv

    etti

    (Italy

    )an

    dB

    ull

    Mac

    hine

    s(F

    rance

    )(19

    64)

    C4Sh

    angr

    i-La

    (staff

    )Ye

    s(pl

    annin

    g)Sp

    ecifi

    catio

    nsfo

    ran

    eww

    orld

    wid

    epr

    oduc

    tlin

    eto

    mak

    eG

    Ea

    clea

    rNo.

    2be

    hind

    IBM

    No

    No

    adap

    tatio

    n;A

    PLpl

    anre

    jected

    (1969

    )

    C5Ve

    ntu

    res

    Task

    Forc

    e(H

    Q)Ye

    s(pl

    anni

    ng)

    Stra

    tegi

    can

    dfin

    anci

    alev

    alua

    tion

    ofn

    ucl

    eare

    ner

    gy,

    com

    mer

    cial

    jeten

    gine

    s,an

    dco

    mpu

    ters

    Yes

    Ada

    ptat

    ion

    thro

    ugh

    dive

    stitu

    reo

    fco

    mpu

    teru

    nit

    (1970

    )M

    S1G

    row

    thCo

    unci

    l(sta

    ff)Ye

    s(pl

    annin

    g)G

    row

    thopp

    ortu

    nitie

    sfo

    rG

    Eth

    atar

    egr

    ow

    ing

    faste

    rth

    anG

    NP

    and

    inw

    hich

    GE

    has

    unqu

    eca

    pabi

    litie

    san

    dm

    aybe

    able

    toga

    inco

    mpe

    titiv

    ead

    vant

    age

    No

    No

    adap

    tatio

    n;pl

    anfo

    rmed

    ical

    info

    rmat

    ion

    syste

    ms

    and

    serv

    ices

    no

    tad

    opte

    d(19

    67)

    MS2

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    anni

    ng)

    Com

    petit

    ors,

    mar

    keto

    ppor

    tuni

    ties,

    new

    pro

    duct

    deve

    lopm

    ent,

    acqu

    isitio

    nta

    rget

    s,fin

    anci

    alpe

    rform

    ance

    and

    projec

    tions,

    alte

    rnat

    ive

    stra

    tegi

    es

    Yes

    Ada

    ptat

    ion

    thro

    ugh

    com

    mer

    cial

    izat

    ion

    ofC

    Tsc

    ann

    er(19

    76)

    Sess

    ion

    II(cr

    oss-l

    evel)

    Yes

    (budg

    etin

    g)O

    pera

    ting

    plan

    san

    dbu

    dget

    sN

    oM

    S3Se

    ssio

    nI

    (cro

    ss-le

    vel)

    Yes

    (plan

    ning

    )Co

    mpe

    titor

    s,m

    arke

    topp

    ortu

    nitie

    s,n

    ewpr

    odu

    ctde

    velo

    pmen

    t,ac

    quisi

    tion

    targ

    ets,

    finan

    cial

    perfo

    rman

    cean

    dpr

    ojec

    tions

    Yes

    Ada

    ptat

    ion

    thro

    ugh

    esta

    blish

    men

    tofi

    nte

    rnat

    iona

    ldi

    strib

    utio

    npa

    rtne

    rshi

    p(19

    77)

    MS4

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    anni

    ng)

    Com

    petit

    ors,

    mar

    keto

    ppor

    tuni

    ties,

    new

    pro

    duct

    deve

    lopm

    ent,

    acqu

    isitio

    nta

    rget

    s,fin

    anci

    alpe

    rform

    ance

    and

    projec

    tions

    Yes

    Ada

    ptat

    ion

    thro

    ugh

    com

    mer

    cial

    izat

    ion

    ofM

    RI

    (1984

    )

    Sess

    ion

    II(cr

    oss-l

    evel)

    Yes

    (budg

    etin

    g)O

    pera

    ting

    plan

    san

    dbu

    dget

    sN

    oM

    S5B

    oca

    (cross

    -leve

    l)Ye

    s(pl

    anni

    ng)

    Intro

    duct

    ion

    ofn

    ewin

    itiat

    ives

    ;rev

    iew

    ofp

    revio

    usin

    itiat

    ives

    succ

    esse

    sYe

    sA

    dapt

    atio

    nth

    roug

    hin

    corp

    orat

    ion

    ofS

    ixSi

    gma

    into

    X-r

    aytu

    bem

    anu

    fact

    urin

    gan

    dCT

    prod

    uct

    line

    (1996

    )Se

    ssio

    nC

    (cross

    -leve

    l)Ye

    s(H

    R)

    Eval

    uatio

    no

    fman

    ager

    sin

    supp

    orto

    finiti

    ativ

    es;

    gene

    rale

    val

    uatio

    no

    fman

    ager

    sN

    o

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    anni

    ng)

    Prog

    ress

    on

    initi

    ativ

    es,c

    om

    petit

    ors,

    mar

    ket

    opp

    ortu

    nitie

    s,n

    ewpr

    odu

    ctde

    vel

    opm

    ent,

    acqu

    isitio

    nta

    rget

    s,fin

    anci

    alpe

    rform

    ance

    and

    proje

    ction

    s

    Yes

    Sess

    ion

    II(cr

    oss-l

    evel)

    Yes

    (budg

    etin

    g)O

    pera

    ting

    plan

    san

    dbu

    dget

    s,pl

    ans

    fors

    peci

    ficin

    itiat

    ives

    No

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 647

    Tabl

    e4.

    Spec

    ializ

    edch

    anne

    ls,cr

    oss

    -func

    tiona

    lpar

    ticip

    atio

    n,an

    dfu

    nctio

    nalc

    oo

    rdin

    atio

    n

    Even

    tN

    o.K

    eych

    anne

    l/typ

    eSp

    ecia

    lizat

    ion

    Cros

    s-fu

    nctio

    nal

    parti

    cipa

    tion

    Atte

    ntio

    nall

    inka

    ges

    ofi

    ssu

    esan

    din

    itiat

    ives

    with

    inch

    anne

    lR

    esul

    ting

    func

    tiona

    lco

    ord

    inat

    ion

    with

    inch

    anne

    l

    C1A

    nnua

    lBus

    ines

    sR

    evie

    w(cr

    oss-l

    evel)

    No

    Yes

    Link

    sfin

    anci

    alre

    sou

    rces

    fors

    upp

    ortin

    gin

    itiat

    ive;

    limite

    dfo

    cus

    on

    oth

    erfu

    nctio

    nsCo

    ordi

    natio

    npr

    imar

    ilyin

    supp

    orto

    fsh

    ort-t

    erm

    finan

    cial

    perfo

    rman

    ceC2

    Ann

    ualB

    usin

    ess

    Rev

    iew

    (cros

    s-lev

    el)N

    oYe

    sLi

    nks

    plan

    ning

    with

    BU

    ope

    ratio

    nalc

    hang

    esnee

    ded

    (i.e.,

    new

    line

    man

    agem

    ent)

    Coor

    dina

    tion

    prim

    arily

    insu

    ppor

    tof

    shor

    t-ter

    mfin

    anci

    alpe

    rform

    ance

    C3G

    row

    thCo

    unci

    l(sta

    ff)Ye

    s(pl

    annin

    g)N

    oA

    cqui

    sitio

    nta

    rget

    sse

    lect

    edby

    HQ

    with

    no

    invo

    lvem

    ento

    fBU

    ope

    ratio

    nsan

    dfin

    ance

    pers

    onne

    l

    Lim

    ited

    cro

    ss-fu

    nctio

    nalc

    oo

    rdin

    atio

    n

    C4Sh

    angr

    i-La

    (staf

    f)Ye

    s(pl

    annin

    g)N

    oPr

    imar

    ilyco

    mpr

    ised

    ofp

    lann

    ing

    staf

    f,w

    ithlit

    tlein

    volv

    emen

    tofB

    Uo

    pera

    tions

    Lim

    ited

    cro

    ss-fu

    nctio

    nalc

    oo

    rdin

    atio

    n

    C5Ve

    ntu

    res

    Task

    Forc

    e(H

    Q)Ye

    s(pl

    annin

    g)Ye

    sLi

    nks

    betw

    een

    plan

    ning

    ,fina

    nce,

    and

    lega

    lat

    HQ

    leve

    lwith

    no

    invo

    lvem

    entf

    rom

    BU

    ope

    ratio

    nsLi

    mite

    dcr

    oss

    -func

    tiona

    lco

    ord

    inat

    ion

    MS1

    Gro

    wth

    Coun

    cil(

    staff)

    Yes

    (plan

    ning)

    Yes

    Prim

    arily

    com

    prise

    dofp

    lann

    ing

    staf

    f,w

    ithn

    olin

    kto

    BU

    ope

    ratio

    nsLi

    mite

    dcr

    oss

    -func

    tiona

    lco

    ord

    inat

    ion

    MS2

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    annin

    g)Ye

    sLi

    nks

    plan

    ning

    tode

    velo

    pmen

    teffo

    rtsin

    CRD

    and

    BU

    Stro

    ngcr

    oss

    -func

    tiona

    lco

    ord

    inat

    ion

    Sess

    ion

    II(cr

    oss-l

    evel)

    Yes

    (budg

    eting

    )Ye

    sLi

    nks

    plan

    ning

    tofin

    anci

    alre

    sou

    rces

    tofu

    ndCT

    deve

    lopm

    ent

    MS3

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    annin

    g)Ye

    sLi

    nks

    plan

    ning

    toin

    tern

    atio

    nalb

    usin

    ess

    dev

    elop

    men

    tan

    do

    pera

    tions

    Stro

    ngcr

    oss

    -func

    tiona

    lco

    ord

    inat

    ion

    MS4

    Sess

    ion

    I(cr

    oss

    -leve

    l)Ye

    s(pl

    annin

    g)Ye

    sLi

    nks

    tode

    velo

    pmen

    teffo

    rtsin

    CRD

    and

    BU

    Stro

    ngcr

    oss

    -func

    tiona

    lco

    ord

    inat

    ion

    Sess

    ion

    II(cr

    oss-l

    evel)

    Yes

    (budg

    eting

    )Ye

    sPr

    ov

    ision

    ofn

    eces

    sary

    finan

    cial

    flexi

    bilit

    yto

    fund

    MR

    Ide

    vel

    opm

    ent

    MS5

    Boc

    a(cr

    oss

    -leve

    l)Ye

    s(pl

    annin

    g)Ye

    sLi

    nks

    stra

    tegi

    cin

    itiat

    ives

    with

    allf

    unct

    ions

    ;ou

    tline

    sro

    lefo

    reac

    hfu

    nctio

    nan

    ddi

    visio

    nVe

    ryst

    rong

    cro

    ss-fu

    nctio

    nal

    coo

    rdin

    atio

    nSe

    ssio

    nC

    (cro

    ss-le

    vel)

    Yes

    (HR

    )Ye

    sLi

    nks

    Six

    Sigm

    ato

    lead

    ersh

    ipde

    velo

    pmen

    t,tr

    aini

    ngo

    fSix

    Sigm

    aB

    lack

    Bel

    ts,an

    dco

    mpe

    nsat

    ion

    for

    top

    perfo

    rmer

    sSe

    ssio

    nI

    (cro

    ss-le

    vel)

    Yes

    (plan

    ning)

    Yes

    Link

    sSi

    xSi

    gma

    tom

    arke

    topp

    ortu

    nitie

    s,n

    ewpr

    oduc

    tdev

    elop

    men

    t,an

    dfin

    anci

    alpe

    rform

    ance

    and

    proje

    ction

    sSe

    ssio

    nII

    (cros

    s-lev

    el)Ye

    s(bu

    dgeti

    ng)

    Yes

    Link

    sSi

    xSi

    gma

    too

    pera

    tiona

    lpla

    nsan

    dbu

    dget

    s

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 648 J. Joseph and W. Ocasio

    Tabl

    e5.

    Chan

    nelc

    ou

    plin

    gan

    dv

    ertic

    alan

    dho

    rizon

    talc

    oo

    rdin

    atio

    nac

    ross

    chan

    nels

    Even

    tN

    o.O

    rgan

    izat

    iona

    lar

    chite

    ctur

    eCo

    uplin

    go

    fgo

    vern

    ance

    syste

    mFo

    cus

    ofo

    rgan

    izat

    iona

    lar

    chite

    ctur

    eR

    esul

    ting

    horiz

    onta

    land/

    orver

    tical

    coord

    inat

    ion

    acro

    ssch

    anne

    ls

    C1D

    ecen

    traliz

    atio

    nLo

    ose

    coupl

    ing

    Ove

    rall

    arch

    itect

    ure

    loos

    ely

    coupl

    edth

    roug

    hfin

    anci

    aliss

    ues.

    Bot

    tom

    -up

    plan

    ning

    ,em

    phas

    ison

    finan

    cial

    perfo

    rman

    cew

    ithin

    ABR

    and

    perio

    dic

    finan

    cial

    repo

    rtsga

    veth

    eover

    alls

    yste

    ma

    finan

    cial

    focu

    s.

    Vert

    ical

    coord

    inat

    ion

    with

    Trea

    sure

    rs

    offi

    ceto

    ensu

    refu

    ndin

    gfo

    rini

    tiativ

    e,an

    dH

    Rto

    deve

    lop

    trai

    ning

    prog

    ram

    fors

    ales

    .

    C2O

    vera

    llar

    chite

    ctur

    ede

    coup

    led.

    Top

    dow

    nco

    rpor

    ate

    plan

    ning

    (e.g.,

    Gro

    wth

    Coun

    cil)

    deco

    uple

    dfro

    mbo

    ttom

    up

    busin

    ess

    unit

    plan

    ning

    and

    finan

    cial

    issue

    s.CX

    Ode

    coup

    led

    from

    Gro

    wth

    Coun

    cila

    nd

    busin

    ess

    unit

    plan

    ning

    and

    ope

    ratio

    ns.

    Lim

    ited

    ver

    tical

    coord

    inat

    ion

    oth

    erth

    anto

    rein

    forc

    efin

    anci

    alta

    rget

    s.H

    oriz

    onta

    lcoord

    inat

    ion

    toad

    opt

    Hea

    vyM

    ilita

    rys

    M23

    60co

    mpu

    terr

    athe

    rtha

    nde

    velo

    pin

    tegr

    ated

    prod

    uctl

    ine.

    C3La

    ckofv

    ertic

    alco

    ord

    inat

    ion

    with

    busin

    ess

    unit

    prod

    uctd

    evel

    opm

    enta

    ctiv

    ities

    .Lac

    kofh

    oriz

    onta

    lin

    tegr

    atio

    nbe

    twee

    nG

    row

    thCo

    unci

    lpla

    nnin

    gan

    dB

    Upr

    oduc

    tdev

    elop

    men

    t.La

    ckofh

    oriz

    onta

    lco

    ord

    inat

    ion

    and

    invo

    lvem

    ento

    fany

    oth

    erdi

    visio

    nsin

    Oliv

    etti

    or

    Bul

    lacq

    uisit

    ions

    .C4

    Corp

    orat

    epl

    anni

    ngD

    ecou

    plin

    g

    Lack

    ofv

    ertic

    alco

    ord

    inat

    ion

    with

    busin

    ess

    unit

    prod

    uctd

    evel

    opm

    enta

    ctiv

    ities

    .Lac

    kofh

    oriz

    onta

    lco

    ord

    inat

    ion

    with

    oth

    erke

    ybu

    sines

    sunits

    :In

    dustr

    ialG

    roup

    or

    Aer

    ospa

    ceG

    roup

    .C5

    Lack

    ofv

    ertic

    alco

    ord

    inat

    ion

    with

    busin

    ess

    unit

    prod

    uctd

    evel

    opm

    enta

    ctiv

    ities

    .Hor

    izon

    tal

    coord

    inat

    ion

    betw

    een

    plan

    ning

    ,fin

    ance

    and

    lega

    lin

    test

    ofp

    ortfo

    liom

    anag

    emen

    tpro

    cess

    .M

    S1La

    ckofv

    ertic

    alco

    ord

    inat

    ion

    betw

    een

    HQ

    eval

    uatio

    noft

    rans

    actio

    nan

    dB

    Uin

    vestm

    ents

    inar

    tici

    alhe

    arts.

    Lack

    ofv

    ertic

    alco

    ord

    inat

    ion

    betw

    een

    Gro

    wth

    Coun

    cilp

    lann

    ing

    and

    busin

    ess

    unit

    prod

    uctp

    lann

    ing.

    Lack

    ofh

    oriz

    onta

    lcoord

    inat

    ion

    betw

    een

    plan

    ning

    and

    ope

    ratio

    ns.

    MS2

    Ove

    rall

    arch

    itect

    ure

    tight

    lyco

    upl

    edth

    roug

    hbo

    ttom

    up

    SBU

    plan

    ning

    syste

    m;c

    hann

    els

    shar

    ediss

    ues

    yeth

    adsp

    ecia

    lized

    focu

    sofa

    ttent

    ion:

    Sess

    ion

    C(H

    R),S

    -I(pl

    annin

    g),S-

    II(op

    eratio

    ns).

    Vert

    ical

    coord

    inat

    ion

    betw

    een

    HQ

    and

    BU

    prod

    uct

    deve

    lopm

    enta

    ctiv

    ities

    .Hor

    izon

    talc

    oord

    inat

    ion

    betw

    een

    tech

    nolo

    gyde

    velo

    pmen

    tand

    com

    mer

    cial

    izat

    ion.

    MS3

    Stra

    tegi

    cpl

    anni

    ngVe

    rtic

    alco

    ord

    inat

    ion

    betw

    een

    HQ

    and

    BU

    inte

    rnat

    iona

    lbus

    ines

    sde

    velo

    pmen

    tact

    iviti

    es.

    MS4

    Tigh

    tcoupl

    ing

    Vert

    ical

    coord

    inat

    ion

    betw

    een

    HQ

    and

    BU

    prod

    uct

    deve

    lopm

    enta

    ctiv

    ities

    .Hor

    izon

    talc

    oord

    inat

    ion

    betw

    een

    tech

    nolo

    gyde

    velo

    pmen

    tand

    com

    mer

    cial

    izat

    ion.

    MS5

    Ope

    ratin

    gsy

    stem

    Ove

    rall

    arch

    itect

    ure

    tight

    lyco

    upl

    edth

    roug

    hto

    pdo

    wn

    stra

    tegi

    cin

    itiat

    ives

    and

    perio

    dic

    revie

    ws

    for

    botto

    mup

    issue

    s.Ch

    anne

    lssh

    ared

    issue

    sye

    thad

    spec

    ializ

    edfo

    cus

    ofa

    ttent

    ion:

    Boc

    a(pl

    annin

    g),Se

    ssio

    nC

    (HR

    ),S-

    I(pla

    nning

    ),S-

    II (op

    eratio

    ns).

    Vert

    ical

    coord

    inat

    ion

    betw

    een

    HQ

    and

    BU

    prod

    uct

    deve

    lopm

    enta

    ctiv

    ities

    ,and

    horiz

    onta

    lco

    ord

    inat

    ion

    acro

    ssal

    lBU

    san

    dH

    R(al

    lin

    supp

    ort

    ofi

    nitia

    tive).

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • Architecture, Attention and Adaptation 649

    and technological opportunities in a number ofways. First, it linked the Computer initiative tothe issue of cash flow, rather than to GEs abil-ity to provide a technical contribution or capitalizeon its engineering skills. GE was the only com-pany other than IBM with the cash flow necessaryto compete in the leasing business. Second, forCordiner, the interaction framed RCA (the No.2 player)not IBMas the primary competitor,but also aligned with Cordiners interest in beingan industry leader. The returns from the initial BoAcontract were substantial and the promise of a lineof compatible computers seemed like a way toachieve leadership in niche markets (such as pro-cess control) and launch into the general purposecomputer market. Finally, projected payback andprofitability for the new product family was highlyfavorable. Despite Cordiners earlier reservations,he approved the plan and budget with favorablesupport from GEs chairman and treasurer, whowere both present (Oldfield, 1996).

    A similar alignment of perspectives was madepossible though Sessions I and II, both cross-levelchannels. For example, Medical Systems decidednot to pursue the emerging CT imaging technol-ogy in 1971 because they viewed it as a nicheproduct for a small market. However, the corpo-rate planning staff (CES) was well aware of thetechnology and viewed it as a technological rev-olution (Session I, 1974). At 1973s Session II,corporate managers expressed concern over thebusiness units failure to move on CT technologyand drew attention to the need for more CorporateResearch and Development (CRD) research in thisarea. As the head of corporate planning recalls, wewere fiddling while Rome burned (Marone, 1993:30); he pushed the group executive (Jack Welch)and CRD to allocate resources for CT research.In early 1974, orders for new X-ray equipmentbegan to fall as hospitals diverted funds to buy thenew technology. At 1974s Session I, the negativeimpact of CT scanners on X-ray market share wasbrought to the attention of corporate headquarters,which imparted a sense of urgency at the corporatelevel and, at the same time, led Medical Systemsto conclude we have got to get into this busi-ness (Marone, 1993: 30). Consequently, duringthe Session II budgeting and operational reviewof the same year, CT technology was identified asa priority. Accordingly, funds were officially allo-cated to a crash development program starting inJanuary, 1995 (MS2).

    HQ and staff channelsIn contrast, interactions within headquarters orstaff channels failed to fully align corporate andbusiness unit perspectives on threats and opportu-nities. For example, in response to stagnant growthand profitability (C3 and MS1), the Growth Coun-cil was developed and charged with identifyinghigh-growth markets that GE had the capabil-ity to develop and support (Rothschild, 2007).The Growth Council targeted nine growth areasincluding Computers and Medical Services. Yetthe Growth Council was limited by the fact thatit was a staff channel. With little operational inputand no participation from senior management, itsplan failed to account for IBMs entrenchment ininformation systems departments (including thoseat GE), the high switching costs of computers,and corporate managements diminishing tolerancefor financing the venture (C5). A similar problemarose with the Growth Councils planning effortfor the Medical Systems business. Although bothcorporate and business unit executives shared acommon belief that growth would be achieved inthe X-ray department through investment in newventures, each focused on different opportunities.The Growth Council focused on medical infor-mation services, which were more profitable, lesscyclical, and less vulnerable to foreign competition(Rothschild, 2007). Medical Systems conducted itsown planning effort focused on opportunities suchas product line extensions and new technology inartificial organs and nuclear medicine (Janssen andMedford, 2009). The CXO had a different view ofthe business entirely and was considering divestingthe unit altogether because of GEs product safetyliability as a result of deaths from the use of GEpacemakers.

    Thus, the Growth Council, a staff channel,stands in contrast to the Annual Business Review(under Cordiner) and Sessions I and II, whichare cross-level channels. Annual Business Reviewsprovided a mechanism for business unit man-agers to directly influence attention at corporateheadquarters and a channel for implementing thecorporate agenda at the level of operating units.Session I reflected a similar pattern. Accordingly,in the cross-level channels, corporate and busi-ness unit managers could collectively reach somelevel of alignment on the opportunities to pur-sue. The staff and headquarters channels, absentcorporate and business unit interactions, allowed

    Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 633660 (2012)DOI: 10.1002/smj

  • 650 J. Joseph and W. Ocasio

    divergent perceptions and plans to persist. Underthese conditions, business unit managers remainedin their own worlds and conceived of opportu-nities more in line with existing product markets.Consequently, we suggest:

    Proposition 1: By creating interactions betweencorporate and business unit managers, cross-level channels align corporate and divisionalattention to threats and opportunities.

    SPECIALIZATION, INTEGRATION,AND CHANNEL COUPLING

    As Tables 3, 4, and 5 indicate, certain structuralcontingencies of cross-level channels shape issueidentification and responsiveness. Corporate andbusiness unit decision-maker interactions are notsufficient condition for channels to ensure greaterattention to strategy formulation and coordinatedresponses to opportunities and threats. Our analy-sis showed that adaptation requires channels thatare: (1) specialized to allow for focused attentionto strategic (competitive and long term) rather thanjust financial or operational issues and initiatives(Table 3); (2) cross-functional to ensure linkagesbetween planning and other functional activitiesin response to opportunities and threats (Table 4);and (3) tightly coupled in a coherent organizationalarchitecture to ensure vertical and horizontal coor-dination of initiatives (Table 5).

    Specialized channels and adaptive movesTable 3 provides a comparison of channel types,the degree to which they were specialized (i.e., hada focused agenda), and the effects on the types ofadaptive moves made. Of the 10 events studied,two resulted in no corporate-driven change (C4and MS1), one resulted in management changeswithout significant strategic adaptation (C2), andseven resulted in strategic adaptation (C1, C3, C5,MS2, MS3, MS4, and MS5). Five of the adap-tive moves involved cross-level and specializedchannels. For two of the seven adaptive moves,staff (C3) or headquarters (C5) channelsratherthan cross-level channelswere involved. Thesetwo involved major acquisitions or business unitdivestitures, respectively, suggesting that busi-ness unit adaptation through mergers, acquisi-tions, and divestitures may occur independently of

    cross-level channels. The acquisition of the Euro-pean Computer businesses (C3) was not success-ful, nor, in the long term, was the expansion ofthe Computer business (C1). The divestiture ofthe Computer business (C5) was a financial suc-cess, but corporate-business unit coordination wasnot consequential for the sale of the Computerbusiness.

    Our analysis suggests that while cross-levelchannels are critical to successfully aligning some-times divergent corporate and business unit percep-tions of the environment, they are not sufficient toensure that attention to long-range issues will notgive way to more immediate concerns. In com-plex organizations, routinized attention to internalfinancial and operational issues often drives outlong-term strategic planning (March and Simon,1958). The broader the agenda in the channel, thesmaller the likelihood that any single issue willreceive sustained attention (often, there is simplynot enough time to cover everything). This maybe particularly true if the strategic issues are notconsidered central to the agenda or share attentionequally with other issues. However, we find thatcross-level channels that primarily focus attentionon planning more readily foster adaptive activi-ties that reflect strategic issues. Focused cross-levelchannels ensure the engagement of decision mak-ers attention (Ocasio, 2011) and the intentionaland sustained allocation of cognitive resources, inthe case of GE, on early identification of newtechnologies and changes in the competitive envi-ronment (e.g., MS2-MS5). Such attentional engage-ment focuses time, energy, and effort on a selectedset of environmental stimuli (e.g., competitiveactivity), corresponding response initiatives, andthe relationships between them (Ocasio, 1997).

    For example, in 1962, the broad Annual Busi-ness Review agenda included financial perfor-mance, problem areas, key markets