Exploitation-Exploration Tensions and Organizational ... · Exploitation-Exploration Tensions and Organizational Ambidexterity: Managing Paradoxes of Innovation Constantine Andriopoulos

OrganizationScienceVol. 20, No. 4, July–August 2009, pp. 696–717issn 1047-7039 !eissn 1526-5455 !09 !2004 !0696

informs ®

doi 10.1287/orsc.1080.0406©2009 INFORMS

Exploitation-Exploration Tensions and OrganizationalAmbidexterity: Managing Paradoxes of Innovation

Constantine AndriopoulosBrunel Business School, Brunel University, Uxbridge, Middlesex UB8 3PH, United Kingdom,

[email protected]

Marianne W. LewisCollege of Business, University of Cincinnati, Cincinnati, Ohio 45221,

[email protected]

Achieving exploitation and exploration enables success, even survival, but raises challenging tensions. Ambidextrousorganizations excel at exploiting existing products to enable incremental innovation and at exploring new opportunities

to foster more radical innovation, yet related research is limited. Largely conceptual, anecdotal, or single case studies offerarchitectural or contextual approaches. Architectural ambidexterity proposes dual structures and strategies to differentiateefforts, focusing actors on one or the other form of innovation. In contrast, contextual approaches use behavioral and socialmeans to integrate exploitation and exploration. To develop a more comprehensive model, we sought to learn from five,ambidextrous firms that lead the product design industry. Results offer an alternative framework for examining exploitation-exploration tensions and their management. More specifically, we present nested paradoxes of innovation: strategic intent(profit-breakthroughs), customer orientation (tight-loose coupling), and personal drivers (discipline-passion). Building frominnovation and paradox literature, we theorize how integration and differentiation tactics help manage these interwovenparadoxes and fuel virtuous cycles of ambidexterity. Further, managing paradoxes becomes a shared responsibility, not onlyof top management, but across organizational levels.

Key words : innovation; organizational ambidexterity; paradox; tensionsHistory : Published online in Articles in Advance December 19, 2008.

1. IntroductionInnovation marks a vital but challenging managerialresponsibility. Innovation denotes intricate knowledgemanagement processes of identifying and utilizing ideas,tools, and opportunities to create new or enhancedproducts or services (Subramaniam and Youndt 2005).To prosper, or even survive, firms must excel at bothexploitative and exploratory innovation (Tushman andO’Reilly 1996), yet tensions emanate from their differ-ent knowledge management processes (March 1991). AsAtuahene-Gima (2005) explain, exploitation hones andextends current knowledge, seeking greater efficiencyand improvements to enable incremental innovation.Exploration, on the other hand, entails the developmentof new knowledge, experimenting to foster the variationand novelty needed for more radical innovation.Organizational ambidexterity signifies a firm’s ability

to manage these tensions (Duncan 1976). Ambidextrousfirms are capable of simultaneous, yet contradictory,knowledge management processes, exploiting currentcompetencies and exploring new domains with equaldexterity (Lubatkin et al. 2006). Related researchsuggests multiple paths to ambidexterity. Theories of

architectural ambidexterity propose dual structures andstrategies, differentiating efforts to focus on eitherexploitative or exploratory innovation (Gupta et al.2006). Contextual ambidexterity, in contrast, emphasizesbehavioral and social means of integrating exploitationand exploration (Birkinshaw and Gibson 2004). Social-ization, human resource, and team-building practices, forinstance, foster shared values and aid coordination, help-ing actors think and act ambidextrously on a daily basis(Ghoshal and Bartlett 1997).Whereas literature is replete with warnings about

the difficulties of managing exploitation-explorationtensions, related studies inhibit comprehensive under-standings, presenting either architectural or contextualapproaches and employing conceptual, anecdotal, orsingle-case studies (Atuahene-Gima 2005, Birkinshawand Gibson 2004, Jansen et al. 2006). In response,we sought to enable more holistic, empirical research.To investigate exploitation-exploration tensions and theirmanagement in practice, we conducted a comparativecase study of five ambidextrous firms that lead the prod-uct design industry. We chose the firms based on proxies

696

Andriopoulos and Lewis: Exploitation-Exploration Tensions and Organizational AmbidexterityOrganization Science 20(4), pp. 696–717, © 2009 INFORMS 697

of ambidexterity, because each had demonstrated con-sistent profitability and repeat clients (indicators of suc-cessful exploitative innovation) while being ranked inBusiness Week’s top 10 for cutting-edge design andreceiving prominent design awards (signs of successfulexploratory innovation).Our results offer an alternative framework for exam-

ining how ambidextrous organizations manage ex-ploitation-exploration tensions. More specifically, threetensions appeared highly robust, posed as paradoxesbecause informants described their opposing poles ascomplementary. These nested paradoxes swirl aroundstrategic intent (profit-breakthroughs), customer ori-entation (tight-loose coupling), and personal drivers(discipline-passion). Informants lauded integration anddifferentiation tactics as vital to their management. Inte-gration positions tensions as interwoven and synergistic,whereas differentiation focuses actions on each pole sep-arately. Building from innovation and paradox studies,we theorize how interplay among the paradoxes and theirmanagement fuels virtuous cycles of ambidexterity.

2. Innovation, Knowledge Management,and Organizational Ambidexterity

In his foundational work, March (1991) links innova-tion and knowledge management to explicate tensionssurrounding exploitation and exploration. Both forms ofinnovation entail combining knowledge—one employingexisting knowledge in well-understood ways, the otherleveraging varied and dispersed knowledge in new ways(Taylor and Greve 2006). As Wadhaw and Kotha (2006)explain, exploitation demands efficiency and convergentthinking to harness current capabilities and continuouslyimprove product offerings. Exploration, in contrast,entails search, variation, and experimentation efforts togenerate novel recombinations of knowledge. Studiesfind that excelling at both exploitation and exploration isvital to successful product development (e.g., Sheremata2000) and long-term performance (e.g., Tushman andO’Reilly 1996). Fundamentally, however, these knowl-edge processes conflict (Lubatkin et al. 2006).The complexity of the tensions intensifies manage-

ment challenges. Smith and Tushman (2005) positthat exploitation and exploration represent overarch-ing demands, spurring nested tensions throughout thefirm. Indeed, Dougherty (1996) depicts firms as grap-pling with multiple innovation tensions, such as conflictsbetween outside-inside, new-old, determined-emergent,and freedom-responsibility. Sheremata (2000) similarlydescribes innovation as a tug-of-war between centripetaland centrifugal forces that fuel discovery and syn-thesis, respectively, whereas Taylor and Greve (2006)stress that knowledge breadth and depth enable productdevelopment.

Innovation tensions also may trigger traps, viciouscycles that stem from increasingly one-sided focus oneither exploitation or exploration. Firms tend towardhomogeneity, finding comfort as they develop mindsetsand routines supporting one form of innovation, escalat-ing their efforts in their preferred mode to the neglectof the other (Smith and Tushman 2005). The results arecounterproductive and eventually destructive. Exploita-tion drives out exploration or vice versa (March 1991,Tushman and O’Reilly 1996). An exploitation bent mayengender “competency traps” (Gupta et al. 2006). Lever-aging current capabilities may enable immediate profits,but foster eventual stagnation, leaving firms vulnerableto market and technological changes (Atuahene-Gima2005). Likewise, Gupta et al. (2006) describe “fail-ure traps” triggered by gravitating toward exploration.Firms narrowly seeking exploration take escalating risks,attempting to negate past innovation failures while ignor-ing core competencies. Future opportunities are soughtat the expense of today’s operations (Birkinshaw andGibson 2004).Organizational ambidexterity is prized as a means

of managing such innovation tensions. As Beckman(2006) reviews, studies encourage top management tocreate supportive structures, strategies, and contexts(e.g., Gibson and Birkinshaw 2004, Smith and Tushman2005, Tushman and O’Reilly 1996). The ideal out-come has been described as balance. Such balancedoes not denote a mediocre split or bland compro-mise, but truly excelling at both exploitation and explo-ration (Atuahene-Gima 2005). Prescribed approaches toambidexterity, however, differ in their emphases, typ-ically advocating either differentiation or integrationtactics.Proponents of architectural ambidexterity, for in-

stance, stress using structure and strategy to enable dif-ferentiation. Segregated efforts target either exploitationor exploration. Related tactics rely on spatial and/or tem-poral separation (Puranam et al. 2006). Spatial separa-tion parses work into distinct units, whereas temporalseparation utilizes the same unit but at different timesfor either exploitation or exploration. For instance, par-allel structures may partition the firm (McDonough andLeifer 1983), whereas dual strategies aim the differentunits at opposing goals (Tushman and O’Reilly 1996).Such differentiation tactics help manage bounded ratio-nality by ensuring focus, but may engender isolation,engrain a preferred innovation mode, and impede coor-dination between varied efforts (Gibson and Birkinshaw2004).Contextual ambidexterity utilizes more behavioral and

social means to integrate exploitation and exploration.This approach has been prescribed at the executive leveland as a more pervasive approach. Hambrick (1994)describes behavioral integration as a unity of effortthrough which the top management team synchronizes

Andriopoulos and Lewis: Exploitation-Exploration Tensions and Organizational Ambidexterity698 Organization Science 20(4), pp. 696–717, © 2009 INFORMS

their social and task processes. Lubatkin et al. (2006)theorize that greater behavioral integration helps execu-tives cope with the contradictory knowledge processesof exploitation and exploration and enable their jointpursuit. Birkinshaw and Gibson (2004) present contex-tual ambidexterity as a higher-order approach. Support-ive social processes (e.g., socialization and recognitionpractices), culture, and interpersonal relations help actorsthroughout the firm think and act ambidextrously.Ghoshal and Bartlett (1997) depict context as the largelyinvisible set of stimuli and pressures that can shapeindividual and collective behaviors toward ambidexter-ity. However, proponents also warn that such integra-tive efforts may strain actors, given the complexity andpotential confusion posed by the contradictory forms ofinnovation (e.g., Gibson and Birkinshaw 2004).Although researchers call for more comprehensive

models of managing exploitation-exploration tensions,such studies are rare, and methodologically limited.Proponents of both approaches note the equifinality ofambidexterity, stressing multiple, and possibly mutuallysupportive pathways. Birkinshaw and Gibson (2004),for example, propose that contextual ambidexterity maycomplement dual structures and strategies by enablingshared values that aid coordination. Likewise, Smithand Tushman (2005) call for top management teams tothink ambidextrously, serving as the integrative lynchpinbetween differentiated units. Yet most research examinesspecific tactics within one or the other approach. Fur-thermore, existing works offer conceptual, anecdotal, orsingle case studies as evidence, inhibiting more holistic,empirical insights (Atuahene-Gima 2005, Birkinshawand Gibson 2004, Jansen et al. 2006).Our research responds by providing a comparative

case study of five ambidextrous firms, examining theirexploitation-exploration tensions and correspondingmanagement approaches (differentiation and/or integra-tion). The research design follows the recommendationsof Denison et al. (1995) as we sought to identify spe-cific tensions and their management in practice. It alsoextends the model provided by O’Reilly and Tushman(2004), which demonstrated the value of learning fromambidextrous leaders. However, whereas their workoffered anecdotal evidence from two such firms, weapplied a more systematic, empirical approach to studypatterns across five exemplars of ambidexterity. We nowdetail our methods.

3. MethodsOur research entailed a comparative case study. Thisinductive design applies replication logic. The cases aretreated as multiple experiments, each case helping toconfirm or disconfirm the findings drawn from the oth-ers (Yin 1994). Results are more focused than overar-ching theories and more valid and generalizable than

those of single-case studies, because findings are deeplygrounded in varied, empirical evidence (Eisenhardt andGraebner 2007, Yin 1994). Applying prescribed meth-ods of comparative case study (e.g., Eisenhardt 1989,Yin 1994), we selected theoretically relevant cases, col-lected case data, and conducted iterative, inductive anal-yses. This method fostered a vital mix of depth andbreadth, allowing immersion within multiple, ambidex-trous organizations.

3.1. Research ContextWe carried out this research in leading new productdesign (NPD) consultancies for several reasons. First,the NPD context has proven well suited to study-ing innovation challenges and their management (e.g.,Alvesson 1995, Robertson and Swan 2003, Sutton andHargadon 1996). NPD consultancies are innovation-intensive settings, demanding that firms excel at bothexploitation and exploration (Hargadon and Bechky2006, Hargadon and Fanelli 2002). Product develop-ment projects, driven by clients or the consultanciesthemselves, range from incremental improvements toradical inventions. Further, within such widely differ-ing projects, the firms must identify opportunities toboth leverage their existing competencies as well asbuild new capabilities. Similarities between NPD andother industries may also enable generalizability to awider population (see Hargadon and Sutton 1997, 2000).This industry resembles others that rely on knowl-edge workers, such as cultural (e.g., art, movies), high-technology (e.g., research, biotech), and professional(e.g., medicine, law) industries.Second, we theoretically sampled firms to fit our

research focus (Eisenhardt 1989). The five case firmsare models of ambidexterity, renowned for their excel-lence in exploitative and exploratory innovation withinthe intensely competitive NPD industry. As Sheremata(2000) notes, ambidexterity is essential to successfuland ongoing product development. Likewise, accordingto Tushman and O’Reilly (1996), ambidexterity enablesboth profitability and innovation over the long term.Indeed, each case firm has remained highly and consis-tently profitable, while receiving numerous awards andtop rankings for cutting-edge innovation.Lastly, within this setting, we sought firms with sim-

ilarities that would aid comparisons and replication, yetwith sufficient heterogeneity to help assess potential gen-eralizability. Selecting firms that offer common services,including product design, engineering, and branding,helped control for certain contextual factors. In addi-tion, all firms are headquartered in the United States.To provide variety, however, we chose firms with dif-ferent industry specializations, thereby offering eclecticsettings for exploitative and exploratory innovation, andwith clients ranging from entrepreneurial firms to For-tune 500 companies. We also selected firms of differing


Table 1 Overview of Case Firms

AnnualYear Number of revenues Number of Number of Number of Number of

Firmsa Servicesb Specializationc founded employees (millions)d officese awardsf projects patents

Firm A B., Eng., Computer hardware 1984 45 $6!0 2 200 3"500 93Gr.D., I.D. (e.g. desktops, 26

monitors, printers,memory cards)

Firm B B., Eng., Consumer and 1983 125 $19!6 3 140 2"700 280+E.D., Gr.D., health products 2 29I.D., In.D., (e.g. padlocks, 1P.D., P., R. digital thermometers)

Firm C B., Eng., Consumer electronics 1969 250 $37!5 8 500 10"000 200+Gr.D., I.D., and services (e.g. 2 41In.D., P.D. mobile phones, retail

environments)Firm D B., Gr.D., P.D. Computer hardware 1994 16 $1!5 1 15 250 20

and sports equipment 6(e.g. video game 1consoles, golf)

Firm E Eng., Gr.D., Consumer products 1985 65 $8!5 2 125 800 100+I.D., In.D., (e.g. kitchen tools) 24P.D., P., R.

aPseudonyms are used to protect anonymity of case firms and their members.bB. (Branding), Eng. (Engineering), E.D. (Environmental Design), Gr.D. (Graphic Design), I.D. (Industrial Design), In.D. (Interaction

Design), P (Packaging), P.D. (Product Design), R (Research).cMost of the firm’s revenue is derived from consulting in these markets.dInformation supplied by principals of each case during fieldwork.eFirst number is total offices; second is the number of offices outside the United States.fFirst number is total awards won; second is number of Design Business Week Awards; third is Catalyst Award (prize for products that

have excelled in the marketplace).

sizes and ages. Table 1 provides an overview of the casefirms.

3.2. Data CollectionOur data collection was intensive, extending overmore than four years. We used several data sources:(1) semistructured interviews, (2) archival data, and (3)observation (Table 2 summarizes the data sources percase). Informant interviews were our primary source ofinductive data, although archival materials and observa-tions expanded our understandings of each case context(e.g., strategic, operational, and cultural features), offer-ing insights that might refute or reinforce our interviewfindings (Forster 1994).

3.2.1. Interviews. We conducted a total of 86 semi-structured interviews with individuals (68 men and 18women) directly involved in the innovation process(e.g., senior executives, directors, designers, engineers).We asked firm founder(s)/CEOs to nominate partici-pating employees across levels, disciplines, and tenureto enable representative sampling. To further ensurethat our sample included the most knowledgeable infor-mants, we used a “snowballing technique.” We asked

initial informants to recommend others within their firmwho could offer further insight. All interviews (lasting70 minutes on average) involved two researchers andwere tape-recorded and transcribed verbatim to ensurereliability (Eisenhardt and Bourgeois 1988).An interview protocol was designed with exploitation-

exploration tensions in mind, but did not include termslike “tension,” “contradiction,” or “dilemma.” Rather,following Spradley (1979), interviews began with ques-tions covering general topics: company history andstructure, current projects, relationships with team mem-bers and clients, competitors, and a typical workday.Given our inductive aims, we encouraged informants towander freely in their answers and probed whenever pos-sible. Our interview protocol evolved systematically. AsGlaser and Strauss (1967) recommend, the study beganwith general research aims. Then, as data collectionand analysis unfolded, our interviews became increas-ingly focused. Within each firm, we continued recruitinginformants until additional interviews failed to disputeexisting, or reveal new, categories or relationships—thatis, until we achieved theoretical saturation (Strauss andCorbin 1990).


Table 2 Data Sources

Firm Interviewsa Archival materialb Observation

Firm A (at two locations) CEO (1) Company handbook Office observation (2 weeks)Senior managers/directors (10) Appraisal formsProduct designer (1) Values surveysSenior industrial designer (2) Marketing materialIndustrial designers (7) ArticlesGraphic designer (1)Senior engineers (2)Engineers (3)CAD and animation (3)

Total: 30

Firm B CEO (1) Company handbook Office observation (1 week)Senior managers/directors (5) Marketing materialSenior Industrial designer (1) ArticlesIndustrial designers (2)Graphic designer (1)

Total: 10

Firm C (at two locations) CEO (1) Company handbook Office observation (2 weeks)Senior managers/directors (7) Appraisal formsSenior industrial designer (1) Marketing materialIndustrial designers (4) ArticlesDigital designer (1)CAD and animation (1)

Total: 15

Firm D CEO (1) Company handbook Office observation (1 week)Directors (2) Marketing materialIndustrial designers (3) Articles

Total: 6

Firm E (at two locations) CEO (1) Company handbook Office observation (2 weeks)Senior managers/directors (13) Marketing materialSenior industrial designers (3) ArticlesIndustrial designers (3)Senior graphic designer (1)Senior engineer (1)Senior design researchers (1)Design researcher (1)Senior model maker (1)

Total: 25

aThis column summarizes the number of informants interviewed and roles within their firm.bThis column summarizes the forms of archival or secondary material collected for each firm.

3.2.2. Archival Data. Industry reports and internaldocuments were examined. Before each visit, we gath-ered articles and Web material related to that firm. Dur-ing the visit, we also collected documents produced bythe firm, such as employee handbooks, marketing mate-rial, and press releases.

3.2.3. Observation. Informal, nonparticipant obser-vations were made during site visits. We were invitedto examine and take notes of the work environment inall five firms. Within the firms, we shadowed membersin their daily routines (e.g., designing on computers,

sketching, handling client calls) and in social interac-tions (e.g., team meetings, impromptu discussions).

3.3. Data AnalysisThrough analysis we moved from raw data toward iden-tification of specific tensions and their management. Ourfour-stage process followed Glaser and Strauss (1967)and Miles and Huberman (1994). Systematic, iterativecomparisons of data, emerging categories, and existingliterature aided development of cohesive constructs andan integrative, theoretical framework.Stage 1. Identifying Initial, Broad Categories Within

Each Case. From our interview data, we first compiled


separate case studies of each firm. Examining all inter-view transcripts, we identified patterns and variancein descriptions of innovation tensions using languageindicators such as: tension, friction, yet, but, on onehand ! ! !on the other hand, juggle, balance, it can swingboth ways, there is a fine line, ! ! !how can you ! ! ! andstill ! ! ! ! We also looked for contradictory statementswithin the same transcript. Across all cases, referencesto tensions were identified in 78 of the 86 interviews.To further categorize the raw data, we applied tech-niques advocated by Van Maanen (1979). Specifically,our conceptual coding entailed using in vivo codes(i.e., first-order concepts comprised of language usedby informants) or a simple, descriptive phrase whenan in vivo code was not available (Strauss and Corbin1990). These first-order concepts offered general insightsinto innovation tensions and corresponding managementefforts as described by informants.To assess the reliability of the generated codes, we

then involved a second coder with considerable qual-itative research experience. Using standardized codinginstructions, the second coder examined all interviewsfor comments indicating tensions and their categoriza-tions. We compared codings, resulting in an intercoderagreement of k = 0!82 (Cohen 1960). Disagreementswere resolved through discussion between the firstauthor and second coder.Stage 2. Linking Related Concepts Within Each Case.

In stage two, we searched for links between and amongthe first-order concepts, which facilitated grouping themtogether into second-order themes. A core aspect ofthe inductive process was that we allowed conceptsand relationships to emerge from the data, rather thanbeing guided by a priori hypotheses (Strauss and Corbin1990). Each completed case write-up was provided tothat firm’s founder/CEO. Executives shared their viewsof the categorizations in follow-up sessions.Stage 3. Conducting Cross-Case Comparisons. Using

standard cross-case analysis techniques (e.g., Eisenhardt1989, Miles and Huberman 1994), we looked for sim-ilar concepts and relationships across cases, comparingthe categories produced in stage two. Similar themeswere gathered into aggregate dimensions that served asthe basis of our emerging framework. We labeled thesedimensions (e.g., strategic intent, integration) either bycapturing the content at a higher level of abstraction orby referring to existing literature that described highlysimilar notions. This process entailed seeking evidenceacross all cases for tension descriptors, resulting in 341instances (66 focused on strategic intent, 136 on cus-tomer orientation and 139 on personal drivers). We alsoidentified 597 depictions of corresponding managementapproaches (177 addressed strategic intent, 187 customerorientation, and 233 personal drivers).We assessed the reliability of each dimension in two

ways. First, we involved the second coder, comparing

codings of the first author and second coder. Intercoderagreement was k = 0!90 (Cohen 1960). Disagreementswere resolved through discussion between the coders.Next, the second author provided an outsider viewpoint.Through extensive discussions, the two authors debatedinterpretations of the data and probed how data fit withinthe proposed dimensions. All discrepancies were dis-cussed to reach agreement.Stage 4. Building a Theoretical Framework. In the

final stage, we drew on existing studies of ambidexterityand paradox to refine our labels and understandings. Toconverge on a parsimonious set of constructs, we focusedonly on the most robust findings. Figures 1(a) and 1(b)depict data structures for identified paradoxes of innova-tion and their management, respectively.

4. Findings: Paradoxes of Innovationand Their Management

Investigating exploitation-exploration tensions across thecase firms, we identified two overarching patterns. First,informants were acutely aware of tensions surroundingopposing forms of innovation, apparent in rich descrip-tions of three, highly robust tensions. Even more inter-esting, however, was that actors depicted the tensions asparadoxes—not as either/or dilemmas or trade-offs, butas synergistic and interwoven polarities. Second, inte-gration and differentiation tactics were lauded as vital to

Figure 1(a) Data Structure: Paradoxes of Innovation

First-orderconcepts

• Ensure stable revenues viarepeat clients

• Careful resource allocationfosters efficiency

• Focus on reputation building• Risk-taking ensures long-term adaptability

• Clients’ requirements help projects fulfill market needs

• Personal expression, challenge, and pride motivate knowledge workers’ creativity

Passion

• Well-defined development processempowers contribution

Second-orderthemes

Aggregatedimensions

Strategicintent

Customerorientation

Tight coupling

Profit emphasis

Breakthroughsemphasis

Loose coupling

• Achieving project goals fosters client satisfaction and loyalty

• Probing new products/technologies surfaces future opportunities • Ongoing experimentation extends firm knowledge base

• Targets (deliverables, budgets, deadlines) encourage execution

• Explicit roles enable focus

Discipline

Personaldrivers


Figure 1(b) Data Structure: Corresponding ManagementApproaches

Aggregatedimensions

Correspondingparadox

First-orderconcepts

• Stress need for andsynergies betweenprofit-breakthroughs

Cultivate aparadoxical

vision

• Leverage synergiesbetween current projectconstraints and emergingpossibilities

Temporally andstructurally

separate workmodes

• Begin by deepeningunderstanding of clientexpectations and market landscape

• Gradually pull awayfrom initial constraints,then return to addressclient requirements

Socialize“practical artists”

• Nurture paradoxicalidentities (discipline andpassion as interwovenelements of professionalsuccess)

• Vary nature of work during different projects and different project phases• Segregate routine/

administration and non-routine/creationresponsibilities

Personaldrivers

• Select exploitativeprojects that leveragecurrent skills Differentiation

Integration

Differentiation

Integration

Differentiation

Integration

Second-orderthemes

Strategicintent

Customerorientation

• Seek and/or create exploratory opportunities

Diversify projectportfolio

Improvisepurposefully

Iterate betweenproject

constraintsand freedom

managing each paradox. Although informants repeatedlyand across all cases described these tensions and corre-sponding management practices, variations were appar-ent, particularly between the smaller and larger firms.Cross-case comparison data tables (see appendix) andillustrative quotes help present these patterns.To further interpret our findings, we leveraged para-

dox literature as well as past studies of innovationand ambidexterity. Like other researchers, we came toapply paradox as a valued lens, integral to our emerg-ing framework (e.g., Lewis 2000, Poole and Van deVen 1989). Similarly, Dougherty (1996) wrote of inno-vation as a series of complex tensions, whereas Smithand Tushman (2005) called for paradoxical thinking toenable ambidexterity. Indeed, Raisch and Birkinshaw(2008, p. 377) review how March’s (1991) influentialtreatise on exploitation-exploration tensions spurred agradual movement from a trade-off to a paradoxicalmindset. Our results make that mind-set explicit, accen-tuating paradoxes of innovation revolving around strate-gic intent (profit-breakthroughs), customer orientation

(tight-loose coupling), and personal drivers (discipline-passion). To unpack each paradox, we examined itsunderlying tension and corresponding management tac-tics. Underlying tensions, as Lewis (2000) explains, arethe sources of paradoxes and evident in interrelated, yetseemingly contradictory polarities. Like the Taoist sym-bol of yin yang, the more actors move toward one pole(e.g., profit), the more they feel pull toward the other(e.g., breakthroughs). The resulting sensation is a cogni-tive tug-of-war. Although the typical reaction is to pullin one direction or the other, without its opposite, eachside in isolation is incomplete (Slaatte 1968).Managing paradox does not imply resolution or elimi-

nating the paradox, but tapping into its energizing poten-tial. According to Eisenhardt (2000, p. 703), effectivemanagement leverages paradox “in a creative way thatcaptures both extremes.” Our cases highlight integra-tion and differentiation approaches to paradox manage-ment. Integration efforts stress interdependence betweenseeming opposites and enable coordination. As notedearlier, advocates of contextual ambidexterity depictsocial (e.g., Birkinshaw and Gibson 2004) and behav-ioral (e.g., Lubatkin et al. 2006) means that help actorsshare and connect divergent knowledge. Similarly, para-dox literature encourages actors to find means of link-ing contradictions, thereby leveraging their synergies(Lewis 2000). In contrast, differentiation focuses effortson either exploitative or exploratory qualities of theparadox. Studies of architectural ambidexterity proposestructural and strategic means of differentiation (e.g.,Puranam et al. 2006, Tushman and O’Reilly 1996). Para-dox literature also articulates this approach. As Pooleand Van de Ven (1989) explain, by splitting a paradox,actors focus on each pole to accentuate its distinct value.We now leverage this paradox framework to exam-ine three interwoven paradoxes of innovation: strategicintent, customer orientation, and personal drivers.

4.1. Paradox of Strategic Intent: Profit andBreakthroughs

We focus on it every day, in everything we do—balancethe business side with the creative side to ensure that weare making money, because we are a business, but we arealso doing the world’s best creative work.

(Senior Manager, Firm C)

This quote exemplifies the strategic intent paradox astensions emanated from needing to emphasize profitand breakthroughs (see Appendix 1(a)). Similarly, pre-vious research presents strategic intent as a firm’sreason for being, often encompassing contradictions.Cameron and Quinn (1988), for example, claim thathigh-performing organizations seek to maintain internalmorale while meeting external demands. In their study ofcultural industries, Lampel and colleagues (2000) depicta tug-of-war between commercial success and artistic


expression, proposing that firms must target both toensure long-run survival.

4.1.1. Underlying Tension. Our findings suggest thatcase informants viewed the profit-breakthroughs tensionas paradoxical. Firm E’s CEO explained, “You must beprofitable and creative ! ! ! ! You need everybody thinkinglike that, right down to the interns. People must fullyinternalize and understand both sides of that equation—they are both important.” As March (1991) explains,exploitation and exploration seek opposing goals—stablerevenues that enable higher mean performance or frame-breaking opportunities that foster greater performancevariation, respectively. Taylor and Greve (2006) elab-orate. They explain that product extensions fuel posi-tive and consistent returns. In contrast, radically newproducts are characterized by high risk and uncertainty,resulting in varied mixes of high-visibility successes anddismal failures.Across the cases, a profit emphasis seemed infused

with conservatism, as informants stressed the value ofrepeat clients and efficiency. Indeed, most firms placed ahigh priority on retaining key clients. A senior industrialdesigner in Firm B explained,

We have this one, long-time client that has brought mil-lions and millions of dollars into this company ! ! ! ! Inaddition to that steady income, you are also looking tothe other opportunities that the client might bring to thetable, especially if the company has several brands undertheir control.

Efficiency appeared to be a secondary, yet valued,tenet. Informants noted how managers carefully allocateresources and select projects to leverage and hone orga-nizational capabilities. This was most evident in Firm A.Numerous informants stressed how specialization in anindustry enables efficiency. As a product design directorin that firm explained, “So we try to put the right peo-ple on (a project), those who have the background to bemost efficient because of their experience.”In contrast to profit’s conservatism, a breakthrough

emphasis entails risk taking in search of opportunitiesthat may enhance the firm’s reputation and adaptabil-ity. Reputation building was a paramount goal for allfirms, but the nature of the desired reputation varied. Thesmallest firms (A and D) wished to be viewed as “vision-ary,” thereby making a name for themselves. The otherfirms, however, sought high-profile projects to attractrepeat clients and prospective employees. According toa senior engineering director at Firm C: “People like towork on high profile projects, like X and Y, becausethey are Fortune 500 companies, so you feel like creat-ing the future of the industry.” Often informants statedthat being on the cutting edge is why they joined the

profession in the first place. A secondary and relatedadvantage of breakthroughs was that such risk takingenhances long-run adaptability. Continuously exploringnew domains fosters the flexibility vital in this intense,dynamic field. As a designer noted, “I mean this com-pany always has to stay cutting edge, keep young, youare forced into this position” (Firm B).Interviews also highlighted the paradoxical link

between profit and breakthroughs emphases. Profit, forinstance, can provide the slack and opportunity totackle potentially breakthrough projects. Firm D’s CEOexplained how repeat clients seek numerous, incremen-tal iterations on their products, building trust to employthe firm for more radical innovation. “We have somevery long-term clients, but we’ve been fortunate becausethey see us as the group to go to for the next new tech-nology, the next new introduction.” Breakthroughs alsocomplement financial goals. One director summed theirinterdependence: “If you’re doing world-class work, thenyou’re going to be happy; you’re going to be paid well.You know that (success) will follow” (Firm E). Likewise,reputation-building breakthroughs can reinforce current,and build new, client relationships. In Firm A, the Direc-tor of Product Design noted, “We are up there (BusinessWeek Top 10) with other innovative firms since we arewell-respected for pushing the limits. At the same timewe have a reasonably stable type of products we do, soif someone says ‘I want to do a desktop’ and know thatthis company is good, they will come to you.”

4.1.2. Management Approaches. Seeking profit andbreakthroughs and achieving both, however, are differ-ent matters, the latter demanding deft ambidexterity.In these firms, managing the strategic intent paradoxinvolved varied mixes of integration and differentiation.Integration entailed cultivating a paradoxical vision thataccommodates the dual emphases. Existing studies positthat holding competing strategies need not mean driftingtoward some mediocre compromise. Collins and Por-ras (1994, p. 44), for instance, depict great, enduringfirms as espousing pragmatic idealism. Such a visioncalls for firms to be highly profitable and highly ide-alistic. Likewise, these leading design firms embraceda “both/and vision.” According to Firm C’s CEO, “Idon’t believe that you can truly drive profitability if youdon’t focus on creativity.” A senior director in that samefirm noted the flip side, remarking that “as long as yourgroup or projects make money, you can pretty much dowhatever you want ! ! !which should foster more creativ-ity.” Informants across cases described how distinctivedesigns attract clients, whereas incremental innovationsfund risky work. Such an explicit, paradoxical visionhelped actors at all levels value the paradox. A princi-pal of industrial design at Firm B explained the result as“the ability to be creative in a controlled way that dealswith project profitability.”


Supportive communications help avoid paradoxicalvisions being interpreted as oversimplified or unrealis-tic. Reiteration also may build trust and avoid mixedmessages, such as employees perceiving cost controlsas assuming priority over innovation (Dougherty 1996).Communication was most frequent in the two small-est firms. For example, whereas Firm A holds staffmeetings twice a month and Firm D communicates itsvision in every project review, Firms B, C, and E havefirmwide meetings once a month. Firm A’s presidentexplained how his managers seek to provide enoughbusiness information to make employees feel at easewith firm decisions, but not so much as to feel micro-managed for financial purposes. A director in Firm Aillustrated how their dual purposes are reinforced: “Inthe first staff meeting that I went to, the CEO was talk-ing to the company about the fine percentage, whichis how much time is about current clients versus howmuch time is spent on exploratory projects.” Similarly,a design director in Firm D explained the need to infusetheir paradoxical vision throughout the firm: “It’s a dailystruggle between making money and doing what’s trueto your heart. Because your ability to make money thenext day depends on how true you are to your heart theday before.”In contrast to such integration, differentiation prac-

tices target distinctive efforts at each strategic goal.Specifically, these firms diversify their portfolios withmore routine, profitable projects and high-risk, break-through projects. Wheelwright and Clark (1992) sim-ilarly prescribe aggregate project plans. In their view,portfolios ideally include incremental projects that paythe bills and radical projects that build new capabili-ties and morale. This tactic spatially separates the ten-sions, enabling focus within distinct projects, yet build-ing a diverse portfolio. Mauzy and Harriman (2003)further describe inventive firms as leveraging currentstrengths, while taking risks that could drive futurebusiness. Case informants applied these tenets. To aidexploitation, the firms sought projects to leverage andhone their existing specializations and knowledge. InFirm B, a principal explained, “we go after certain typesof projects based on current skills and interests.” Sim-ilarly, a director described Firm D’s careful selectionof incremental projects: “We are trying to have ! ! ! adiverse range of programs that really appeal to whatwe are really all about: action, sports, but we still doour computer projects because that is the money makingactivity.”Exploratory projects often require more proactive

searches. Informants described widely varying means,such as collaborating with clients and/or suppliers injoint ventures or working for potential equity with start-up ventures. Because breakthrough projects are rare,firms initiate their own as well. Again size played a

distinguishing role between cases. Informants acrossFirms A and D, the smallest firms, claimed to takethe greatest initiative in creating exploratory projects.Indeed, these firms formalized the process (e.g., eachhas a director of ventures) and seemed highly attunedto seizing entrepreneurial opportunities. Firm A’s CEOexplained: “By creating new companies we can makework for ourselves rather than having to go out andchase new clients. Instead of waiting for opportunities tocome, we create them ourselves.” Whereas such exam-ples depict tensions swirling around strategic intent,we now examine how tensions also appeared nestedwithin particular projects, posing a paradox of customerorientation.

4.2. Paradox of Customer Orientation: Tight andLoose Coupling

There is usually a tension between what we believeshould be done and what the client believes should bedone ! ! ! ! Although the person we deal directly with maybelieve that the X we recommended is best, other forcesmight make it second best, like a dominant marketingdepartment or their CEO has a strong opinion or a prereq-uisite in the market. That is always a tension to manage.

(Industrial Design Director, Firm B)

Informants depicted a paradox of customer orienta-tion surfacing during projects, grappling with the needto be tightly and loosely coupled to the client (seeAppendix 1(b)). Whereas clients stress competitive, mar-ket, and their own constraints, case firms seek to exploreemerging technologies and trends. Similarly, Dougherty(1996) describes how product development projectsrarely leverage both current needs and future possi-bilities. Developers of cultural products (e.g., movies,music) face a related struggle, striving to meet mass mar-ket demand, and construct new niches (Delmestri et al.2005, Lampel et al. 2000). Tight coupling to the cus-tomer becomes blessing and curse, aiding insight intothe current market, but potentially inhibiting develop-ment of new segments (Danneels 2003). A designerexplained that regardless of project type—exploitative orexploratory—this tension persists. “Everybody is reallydedicated to their project and (the client), but they alsowant to push it further” (Firm D).

4.2.1. Underlying Tension. These cases present atension underlying customer orientation: Tight couplingstresses needs and constraints, whereas loose couplingnotes possibilities and freedom. It is the combination,however, that can fuel innovation success. Accordingto one designer, “It’s a tug-of-war ! ! !between the dif-ferent priorities of (our) departments ! ! ! and the clients’and market needs. It’s the friction of these priorities thatgives rise to higher levels of creativity” (Firm C).


Informants demonstrated great client respect, valuingtheir insights into their market and competition, andseeking to ensure their satisfaction and loyalty. Client-driven boundaries link a project to explicit market needsand to the client’s capabilities. As the Vice President ofProgram Development at Firm B explained, “The clientis the expert at what they do, specifically to their cate-gory of goods or their industry.” Further, tight couplinghelps ensure that projects meet client expectations. Inthe words of one creative director, “If we don’t serve theclients first ! ! ! then we can’t develop those client rela-tionships” (Firm C).On the other hand, loose coupling to clients was

stressed, even in the most exploitative, incrementalprojects, because development teams yearn to probeemerging opportunities and experiment continuously.Project-driven probing was stressed in all firms, becauseinformants shared their desire to push client-set bound-aries. According to a senior vice president, “We do it(break the rules) in every project. Because I think that’snecessary ! ! ! there’s a high potential that in that processyou’ll develop some ideas that are very applicable evenwith all the parameters put back on” (Firm A). Probingin Firms A and D, however, was also internally driven.Informants noted how they felt encouraged to investigateareas unrelated to any current projects, hoping to iden-tify new possibilities in different industries. On a relatedtheme, informants in all five cases advocated ongoingexperimentation to extend their firm’s knowledge base.As explained by a senior designer in Firm B, “In thiskind of business you have to evolve and try to stay aheadof potential trends and have a good understanding ofwhat is going to be around forever and what is truly atemporary fashion trend and a style ! ! !you have to lookat the bigger picture all the time.”The paradoxical link between these orientations,

however, was clear in repeated claims that the inter-play of tight and loose coupling enables project suc-cess. Without efforts to push project constraints, clientsmay force their views to a point that inhibits innova-tion. A likely result was described as the regurgitationof mundane, overcommercialized products. Similarly,Danneels (2003) finds that by building close ties withclients, projects can meet market demands, but at greatexpense as myopic vision curbs longer-term potential.At the opposite extreme, projects driven predominantlyby future-oriented designers can lack clear targets andbecome wasteful. As problems arise, these knowledgeworkers may become less willing to involve the client,intensifying reliance on their own views. Explained byFirm A’s founder, “With designers, they can get so inter-ested in doing cool design, that sometimes they missthe big picture, ignoring the actual business, and evenmissing opportunities within a project to do something

fun.” Likewise, Miner et al. (2001) warn of excessiveopportunism. Continuously seeking blue-sky possibili-ties can create “opportunity traps” that ignore marketdemands in favor of designer whims. In addition, justas designers may stretch a client’s comfort zone, clientsmay inhibit case firm complacency. As one industrialdesigner boasted, “We don’t have a ‘house style.’ Ourstyle is very client-driven, very project driven” (Firm B).

4.2.2. Management Approaches. These firms soughta customer orientation that is both tightly and looselycoupled to their clients. Integrative approaches were evi-dent in descriptions of purposeful improvisation. Projectwork seemed to stress exploiting existing routines andknowledge, while exploring within and pushing projectboundaries. Such descriptions fit the Miner et al. (2001)definition of improvisation as recombining existing ele-ments in new ways, much as a musician reassemblespreviously performed bundles of notes into a uniquemelody. These efforts enable serendipity, helping linkthe right idea at the right time with the right need(Robinson and Stern 1998). As one director noted, “Wehave to be flexible and see things as opportunities, notas constraints ! ! ! see potential in things that initially looklimited” (Firm E). A Firm A designer reiterated, “Youhave to cater to what the client wants and be creativewithin the restrictions they set.” In many of the firms,knowledge brokering—seeking unconventional connec-tions between varied, existing ideas (Hargadon andSutton 1997)—fueled improvisation.Differentiation seemed to entail temporal separation,

splitting the tensions by iterating between project con-straints and freedom. Most often, informants describedstarting a project by listening intently to the client,seeking to “walk in their shoes” to fully grasp projectgoals. Teams then begin to pull away from initial con-straints, most often using brainstorming to explore newdomains. As projects progress, constraint-freedom iter-ations may increase in frequency. A design directorexplained, “When we first start our brainstorming ses-sion we try to let (client constraints) go ! ! !because usu-ally the really crazy ideas tend to trickle down ! ! !”(Firm D). Danneels (2003) depicts the effect of such iter-ations as tight and loose coupling between creator andcustomer, enabling creative expression and commercialsuccess. A designer described this ebb and flow as help-ing clients expand their thinking. “This is part of whatthey are paying us for—to push them. And you have tocome up with creative ways to push them” (Firm B).However, such tensions of customer orientation appearedfurther nested, as project participants expressed theirown tensions, proposed now as a paradox of personaldrivers.

4.3. Paradox of Personal Drivers

Why does a lack of constraints equal creativity in peo-ple’s minds? ! ! ! ! You can argue that mathematicians are


probably the most creative people on the face of theearth—how they come up with something new in such ahighly constrained science.

(Industrial Design Director, Firm B)

Informants across cases described discipline and passionas interdependent, fueling their own innovative efforts(see Appendix 1(c)). Such depictions suggested a dualand paradoxical inner drive, whereas discipline, controland structure power execution, passion, pride, and risktaking mobilize creative expression. Such images extendstudies noting similar tensions among knowledge work-ers (e.g., Brown and Duguid 2001, Leonard and Swap1999).

4.3.1. Underlying Tension. Informants articulateddiscipline and passion as interwoven. In their view,product development challenges—clients seeking excit-ing, new products in short time frames with lim-ited budgets—demand both sides of the coin. Success-ful projects require knowledge workers who “live andbreathe design,” but also take advantage of streamlinedprocesses to speed development. A Firm B vice presi-dent explicated: “A lot of times (clients) will come tous when they have a tricky problem and they need theproduct to get out a lot faster than they have ever donebefore. So it is important to have a certain degree ofdiscipline.”Discipline was described in terms of control, account-

ability, and structure. Informants stressed how well-defined processes empower individual contribution,targets help ensure project execution, and explicit rolesenable focus. Similarly, Benner and Tushman (2002)stress that repetition of activities embedded in stan-dardized best practices increase the speed and effi-ciency of innovation. A principal engineer explained,“We have people ! ! !quite creative but if they are lazy,they are not really useful. You can sit around andcome up with a great idea but if you don’t executeit ! ! ! ” (Firm B). In Firms A and D, senior manage-ment defined employees’ targets and roles, whereasproject leaders assumed this responsibility in the largerfirms. Firm D’s CEO described how his senior man-agers support designers and engineers: “basically (wetell them) ‘here is the project, here is what we haveto do, here are some deadlines, some deliverables tomake’—that gives them some structure.” Interestingly,roles seemed most dynamic in Firm C, the largest orga-nization. Although roles are clearly assigned by projectleaders, their informants noted how employees are likelyto move around and between projects as their skills areneeded.In these firms, passion seemed to signify a pow-

erful blending of personal expression, challenge, andpride. Informants described their zeal as emanatingfrom an intense desire for work that is engaging and

challenging. They often noted how design permeatestheir lives, stressing that internalizing their work canopen opportunities for unexpected inspiration and artis-tic expression. A director at Firm C described his hiresas follows: “extremely passionate ! ! ! emotional ! ! !we arein an industry that kind of speaks to emotions, peopletake very seriously what they are doing and it’s part ofwho they are.” Such depictions exemplify the central-ity of intrinsic motivation to creativity (Amabile et al.2005).Interestingly, rather than lament their contradictions,

informants described discipline and passion as synergis-tic. Firm D’s CEO stated the link as follows:

My personal philosophy is that designers are like freeradical electrons. You let them do their thing, and everyonce in a while you gather them up and focus them onsomething, and they become very powerful. Then as soonas that is done, and they produce what they needed toproduce, you let them go back and be free radicals again.

Indeed, this and other studies suggest the dangers ofdiscipline without passion, or vice versa. On the onehand, excessive standardization may engender alienationand rigidity (Brown and Duguid 2001). Such resultswere described as a worst-case scenario by informants—a workplace devoid of inspiration. A senior industrialdesigner warned about using structure to impose disci-pline: “When you do that, people can start only actingon the responsibility frame that you give them, and youkind of lose the passion ! ! !you become a little screw inthe machine” (Firm E). Likewise, passion without disci-pline can fuel chaos (Levitt 2002). As stressed by onedesigner, “In such a creative environment, unless youhave some structure, it gets out of control very fast”(Firm B). Results can be escalating obsession for indi-viduals and inefficiency for firms.

4.3.2. Management Approaches. To manage theparadox of personal drivers, integration fostered both/and thinking, while differentiation compartmentalizedopportunities for routine and nonroutine work. Inte-gration entailed cultivating paradoxical work identities.Designers, engineers, and managers stressed the impor-tance of socialization—from hiring to mentoring toongoing reviews—in helping firm employees identifythemselves as “practical artists.” This seeming oxymoronhighlights the simultaneity of discipline and passion. Avice president explained how designer socialization, forexample, starts with hiring and self-selection:

You need the designers to be here because they enjoyhaving the engineers around to help them not createa piece of sculpture, but create a product. People hereare not sculptors. They decided not to be artists. Theydecided to be designers, and we feed on each other thatway. (Firm A)


The Creative Director of Firm C similarly elaborated:

Well, there’s right and left brain—we constantly talkabout teaching people how to switch from one side to theother because it is realistic, and as a business, we cannotstay in business if they don’t do their timesheets, or ifthey don’t fill their reports out because this is not an artstudio where we can sit around and paint all day.

A hybrid of temporal and spatial differentiation ap-peared to help manage the paradox of personal drivers.Varying the nature of work at different times—duringdifferent projects and project phases—enabled knowl-edge workers to leverage their discipline and their pas-sion. For instance, as projects evolve, a team member’sfocus systematically shifts between exploitation andexploration. Likewise, individuals rotate across projects,from the incremental product extension to the poten-tially frame-braking invention. An industrial designerat Firm B elaborated how mixed projects are vital forknowledge workers as well as for the firm:

We have some exciting projects and some less excitingprojects, and I think that the firm consciously tries to givedesigners a mix because if you are doing hard time allthe time, it pressures your soul. We do projects that aremuch more blue sky ! ! ! ! It is this wonderful deep waterthat designers can swim in for a while.

In contrast, distinct roles enabled spatial separation asdifferent individuals or units focused on either the dis-cipline of execution, budgeting, and other administrativeelements, whereas others tapped into the passion fos-tered by experimentation and ideation. For instance, thelarger firms (B, C, and E) used more division of labor,whereas employees in the smaller firms wear multiplehats. That said, all firms relied on paradoxical identities.In larger firms, this might be vital to help prevent knowl-edge workers from feeling like administrators. There-fore, although directors, who have risen through the firm,might now focus on billing, client relations, etc., theystill see themselves as practical artists. Spatial separa-tion, however, frees other designers and engineers tocreate.

Designers think in pictures, certainly not numbers but itdoesn’t always work well on all projects and all clients.Some clients want clear minutes distributed the day afterthe meeting and they want to be able to talk to someoneabout the billing, and the expenses and this and that. Evenif designers are good at it, “do we really want them tospend 3% of their time going over bills as opposed todesigning?” So, that is why we have project coordinators.

(Director, Firm E)

We now step back from case details to discuss theparadox of personal drivers, as well as those of customer

Figure 2 Virtuous Cycles of Ambidexterity

Profitemphasis

Discipline

Breakthroughsemphasis

Passion

Integration

Strategic intent

Differentiation

Exploitation Exploration

IntegrationCustomer orientation

Differentiation

IntegrationPersonal drivers

Differentiation

Loosecoupling

Tightcoupling

orientation and strategic intent, and propose a more com-prehensive framework.

5. Discussion: Fueling VirtuousCycles of Ambidexterity

Through this comparative case study we sought tocontribute an alternative framework for examiningexploitation-exploration tensions and their management.Our findings highlight three nested tensions, presentedas paradoxes of innovation. Further, managing theseparadoxes involved a mix of integration and differ-entiation tactics. We now build from our case find-ings and existing literature to theorize how firms mightmanage innovation paradoxes and thereby fuel virtuouscycles of ambidexterity (see Figure 2 for illustration).More specifically, we propose that three factors inter-act to reinforce and sustain organizational ambidexterity:a multilevel approach, complementary tactics, and learn-ing synergies. We now leverage existing literature andcase study examples to explicate each factor.First, we propose that a multilevel approach is vital

to managing nested paradoxes of innovation. As illus-trated in Figure 2, the identified paradoxes are inter-woven across levels. Supporting previous claims (e.g.,March 1991, Smith and Tushman 2005), exploitation-exploration conflicts spawn a host of tensions through-out organizations. Indeed, the recent review by Raischand Birkinshaw (2008) called for ambidexterity researchthat examines multiple levels and reflects this complex-ity. These case studies enable a more holistic perspec-tive. For instance, the strategic intent paradox appears


predominant at the firm level, whereas customer ori-entation particularly affects efforts within projects, andpersonal drivers appear most impactful on knowledgeworkers themselves. By managing innovation acrosslevels, firms reduce the threat of mixed messages.As Dougherty (1996) warned, efforts to manage onlyone tension may be contradicted by efforts aimed atother, interconnected tensions. In these design firms, forinstance, claims of seeking a both/and strategy aimedat profit and breakthroughs could be easily discountedby one-sided efforts in projects or individual efforts(e.g., overemphasis on tightly coupling projects to theirclients and on disciplined work practices might signalthat exploitation reigns supreme).By managing innovation paradoxes across levels,

management also becomes the responsibility of actorsthroughout the firm. Executives set the context, pro-viding strategic leadership and allocating resources thatdetermine their firm’s portfolio of projects. Directors andproject leaders, in turn, guide specific projects, ensuringadherence to clear development processes and encour-aging improvisation as well as iterations between workmodes. Finally, knowledge workers themselves choosewhen and how to best apply their discipline and passionto enhance product development. Together, these effortsreinforce each other in a virtuous cycle as ambidexteritybecomes pervasive.Second, we posit that integration and differentia-

tion offer powerful, complementary tactics for foster-ing ambidexterity. This combination helps reduce theanxiety and defensiveness that tensions spark and thatcan spur vicious, rather than virtuous, cycles. As Lewis(2000) explained, actors’ typical reactions to tensions aredefensive, trying to resolve or eliminate their anxiety bystressing their preferred pole. The results are counterpro-ductive. “In attempting to reduce the frustrations and dis-comfort of tensions, actors’ defensive behaviors initiallyproduce positive effects but eventually foster opposite,unintended consequences that intensify the underlyingtensions” (2000, p. 763). Such is often the case in inno-vation, as studies note the potential traps of one-sided,escalating efforts toward either exploitation or explo-ration (e.g., Gupta et al. 2006, Tushman and O’Reilly1996).Virtuous cycles, in contrast, stem from embracing

tensions, valuing their synergies, and their distinctions(Lewis 2000). Blending integration and differentiationfosters this effect. Integration tactics accentuate theimportance of both poles of exploitation-exploration ten-sions. These social and cultural approaches (e.g., orga-nizational vision, project norms for improvisation, andsocialized identities as practical artists) enable a para-doxical mindset. However, differentiation is also vital.Clearly focusing actions (e.g., on projects targeted at dif-

ferent strategies, at constraints or freedom during projectphases, and with segregated work modes) helps maxi-mize the distinct benefits of opposing poles.The final factor that may sustain ambidexterity is

the learning synergies enabled by exploitation andexploration efforts. Although March (1991) depictedthese as disparate learning activities, he stressed thatsuccess stems from their simultaneity. Jansen et al.(2006) elaborated, explaining that the interplay betweenexploitation and exploration enables absorptive capacity.Conceptualized by Cohen and Levinthal (1990), absorp-tive capacity denotes a firm’s ability to recognize, assim-ilate, and apply new knowledge. Exploitative efforts helptransform knowledge into commercial ends, but withoutexploration a firm’s stock of knowledge will wane (e.g.,being utilized repeatedly until a firm is stuck in a specificproduct or industry niche). Likewise, exploratory effortshelp continuously renew and expand a firm’s knowl-edge base, but without exploitation that knowledge maynot be utilized fully (e.g., recombined in varying waysacross projects or product iterations). In essence, the twomodes of innovation are mutually reinforcing. Accordingto Smith et al. (2005), innovation is a dynamic capabil-ity, demanding attention to both existing knowledge andknowledge creation.Our five case studies exemplify such learning syner-

gies. Absorptive capacity is demonstrated in their abilityto thoroughly utilize their knowledge base, deepening,honing, and leveraging their industry and product spe-cializations, while broadening their competencies byseeking more radical innovation and even developingtheir own projects to investigate emerging technologyand market domains. The results are impressive. Eachfirm has remained at the top of the product designfield in terms of financial performance and cutting-edgeproducts.In conjunction, a multilevel approach, complementary

tactics, and synergistic learning fuel virtuous cycles ofambidexterity. For instance, a paradoxical vision of thefirm as seeking both profit and breakthroughs may fosterimprovisation as project teams value client-set bound-aries, while pushing those boundaries to probe new pos-sibilities unconsidered by the client. In turn, the needto exploit and the opportunity to explore may foster thediscipline and passion that energize individual knowl-edge workers. As the cycle continues, interactions rein-force ambidextrous practices, positioning the firm tomeet financial goals and leverage frame-breaking oppor-tunities. Resulting benefits validate paradoxical thinkingand actions, and thereby sustain ambidexterity. However,given the need to manage such complexity in consistentyet paradoxical ways, it is not surprising that ambidex-terity remains a much lauded, but rarely achieved, orga-nizational capability. For this reason, we hope that these


case firms may serve as exemplars and motivate fur-ther research. In this light, we now examine the impli-cations of our study and suggest directions for futureresearch.

6. Implications and Future DirectionsLeveraging comparative case studies and blending inno-vation and paradox literature enabled development of aframework that explicates exploitation-exploration ten-sions and their management. More specifically, our workidentifies nested tensions, posed as paradoxes of innova-tion, illustrates the value of blending differentiation andintegration approaches, and theorizes the potential forinterwoven paradoxes, and their management to fuel vir-tuous cycles of ambidexterity. Given our goal of devel-oping a model that might spur more comprehensiveempirical research, we now examine opportunities fortesting and extending our framework.Seeking to learn from ambidextrous exemplars, we

studied firms that excel at exploitative and exploratoryinnovation. The NPD industry is acclaimed for its inno-vative intensity, and these five firms are industry lead-ers in terms of profits and awards. However, whetherand how their lessons apply elsewhere raises importantquestions. We encourage future studies to extend ourwork to other services as well as more conventionalmanufacturing. Existing research notes similar tensionsin cultural industries (Lampel et al. 2000) and productdevelopment groups within more traditional industries(Taylor and Greve 2006). Whether linking integrationand differentiation tactics enables ambidexterity in suchcontexts, however, requires investigation. Likewise, therelationship between organizational size and this blendedmanagement approach needs assessment. The patternspresented here were robust across cases, but size dif-ferences were evident. For instance, the smallest firmsseemed more intent on reputation building and risk tak-ing than their larger counterparts. Size, however, is rel-ative. All five firms would be deemed small in mostindustries. However, researchers note the need for suchalternative models of ambidexterity because managinginnovation tensions is increasingly vital in large corpo-rations (e.g., Dougherty 1996, Sharma 1999). Lubatkinet al. (2006), for example, suggest that executives oflarger firms reconsider separating units to focus oneither exploitation or exploration, and instead createunits capable of pursuing both. Further, we are left ques-tioning how/if nested tensions and their managementwould vary in less ambidextrous or even poorly perform-ing organizations. Research indicates the potential forvicious, rather than virtuous, cycles in nonambidextrousfirms, but call for more targeted, empirical research (e.g.,Jansen et al. 2006).Another direction for future research lies in apply-

ing other research lenses. Our study stressed the role

of management. Tapping innovation and ambidexter-ity literature, we delved into how firms might copewith innovation tensions, proposing that management isitself paradoxical—requiring integration and differenti-ation efforts throughout the firm. This finding connectsstudies of architectural (e.g., Tushman and O’Reilly1996) and contextual (e.g., Gibson and Birkinshaw2004) ambidexterity. It also elaborates calls in practi-tioner literature for paradoxical, both/and approaches tomanagement (e.g., Collins and Porras 1994). However,our management emphasis begs future studies of moreunderlying processes such as sensemaking and cogni-tion. As Smith and Tushman (2005) note, effectivelymanaging exploitation and exploration may require aparadoxical mindset. Indeed, paradox literature stressesthat paradoxical thinking is vital to combating natural,often counterproductive tendencies to overrationalize oravoid tensions (Eisenhardt 2000, Poole and Van de Ven1989). Delving into underlying sensemaking and cog-nition may suggest powerful enablers and deterrents topervasive ambidexterity.

7. ConclusionIn today’s dynamic world, innovation may posethe ultimate advantage and challenge for organi-zations. These leading design firms embrace para-dox as exploitation-exploration tensions swirled aroundseemingly contradictory but interwoven forces. Infor-mants richly articulated the benefits of both poles ofexploitation-exploration tensions and their synergies.Furthermore, their use of both integration and differen-tiation approaches to managing paradoxes of innovationdemonstrate managerial creativity. The cases presentedand resulting theory offer insights into the intricacies ofambidexterity and may spark future attempts to refute orelaborate our findings and the potentially valuable roleof a paradox perspective.

A paradox is an idea involving two opposing thoughtsor propositions which, however contradictory, are equallynecessary to convey a more imposing, illuminating, life-related or provocative insight into truth than either factcan muster in its own right ! ! ! ! What the mind seeminglycannot think, it must think. (Slaatte 1968, p. 4)

AcknowledgmentsThe Carnegie Trust for the Universities of Scotland supportedthis research with a grant to the first author. The authorsalso thank Gilbert Probst and their anonymous reviewers forexceptionally constructive feedback, as well as Shannon Funk,Charles Galunic, Manto Gotsi, Amy Ingram, Sarah Jack,Elizabeth Long Lingo, Ajay Mehra, Ruth Simpson, Vange-lis Souitaris, and Steve Wheelwright for their input on earlierdrafts of the paper, and seminar participants at Brunel Univer-sity and ALBA Graduate Business School for their insights onthis project.


Appendix 1(a). Data Table for Cross-Case Comparisons: Strategic Intent

Exploitation—exploration tension Management approaches

Integration: Cultivate a Differentiation:Profit emphasis Breakthroughs emphasis paradoxical vision Diversify project portfolio

1. Ensure stable revenues viarepeat clientsFirm A—Strong"

“We do a lot of work for Xand this has helped us lastyear to remain financiallystrong” (CEO).

Firm B—Strong“We have this one, long-time

client that has broughtmillions and millions ofdollars into this company”(Senior Industrial Designer).

Firm C—Strong“We have had a few big

projects for longer termclients, where we have donedozens of TVs, for example,or dozens of cell phones”(Director, MechanicalEngineering).

Firm D—Moderate“! ! ! and the other thing is, did

we help our client makemoney? If you get to satisfythem they’ll bring you morebusiness” (Design Director).

Firm E—Moderate“If you do work for this

company (computerperipheral), you might havesome breakthrough thinking,but then that breakthroughthinking gets applied tothree different products,which becomes more aboutexecution” (President).

2. Careful resource allocationfosters efficiencyFirm A—Strong #Weekly

resource schedulingmeetings by seniormanagement, stressindustry specialization$

“So we try to put the rightpeople on (a project) to havethe background to be themost efficient because oftheir experience, we need towork within parameters inorder to do this” (Director,Product Design).

1. Focus on reputation buildingFirm A—Strong #“Visionary”

through experimentalprojects, awards, press$

“I think the other part of ourreputation is really goodcreative work ! ! !we aretrying not to put too muchstock in awards, but they arereally nice. They are greatbecause I basically like thewhole idea of saying thingsto Business Week ! ! !because(clients) are going to callthese top firms and if wedon’t stay there we are introuble” (VP).

Firm B—Strong #“Highprofile” via breakthroughproducts, awards, press$

“You do great work, moreclients who want you to dogreat work come toyou ! ! ! company Y does allthese really high profileprojects that everybodyknows about ! ! ! and theprojects have to be sexy tobegin with” (SeniorIndustrial Designer).

Firm C—Strong #“Highprofile” via breakthroughproducts, awards, press$

“We were doing this shoeproject for the MOMA(Museum of ModernArt) ! ! ! that was one of themost satisfying I would say,to me and the company,because that won an IDaward, and went on to be inseveral magazines, it got alot of press” (IndustrialDesigner).

Firm D—Strong #“Visionary”through experimentalprojects, awards, press$

“Another project, it’s a veryhigh profile award for a TVshow so it’s pure kudos. It’sjust an opportunity tocreatively show off” (DesignDirector).

Stress need for andsynergies betweenprofit-breakthroughs

Firm A—Strong #Formalcommunication involvingall staff reinforcesdual-emphasis; two staffmeetings per month$

“In the first staff meeting thatI went to, the CEO wastalking to the companyabout the fine percentage,which is how much time isabout current clients versushow much time is spent onexploratory projects”(Director).

Firm B—Moderate #Formalcommunication involvingall staff reinforcesdual-emphasis; one staffmeeting per month$

“Every month we have a staffmeeting where we updatepeople on business that wehave done, new businesscoming in, awards orproducts featured in press,how each group doesfinancially” (Principal, ID).

Firm C—Moderate #Formalcommunication involvingall staff reinforcesdual-emphasis; one staffmeeting per month peroffice$

“We have a managementmeeting where we talk toeach other on a formal, butfrequent basis ! ! ! and thenwe have employees who areconfidently working onprojects that share stuff withother offices. So that tendsto create cross-communi-cation. Within each office,we meet the first Monday ofevery month, the wholeoffice, and have everyonedescribe what they aredoing” (VP Creative Media).

1. Select projects that leverageskillsFirm A—Strong“You may do computer

hardware for a year”(Director, Product Design).

Firm B—Strong“We go after certain types of

projects based on currentskills and interests”(Principal, ID).

Firm C—Strong“Maybe you take on a client

where the work is not socreatively joyful ! ! ! saysomething that we havedone 100 times ! ! !” (SeniorManager).

Firm D—Strong“We are trying to have ! ! ! a

diverse range of programsthat really appeal to whatwe are really all about:action, sports, but we stilldo our computer projectsbecause that is the moneymaking activity” (DesignDirector).

Firm E—Strong“I think that the focus here is

more on commercial,mass-produced products.That is what we’re reallygood at doing” (IndustrialDesigner).

2. Seek/create radicalopportunitiesFirm A—Strong #Initiating

own projects, receivingequity in high-risk clients,and developing jointventures—managed bythe Director of Ventures$

“By creating new companieswe can make work forourselves rather than havingto go out and chase newclients” (CEO).

Firm B—Moderate #Initiatingown projects andreceiving equity inhigh-risk clients$


Appendix 1(a). (cont’d.)


Integration: Cultivate a Differentiation:Profit emphasis Breakthroughs emphasis paradoxical vision Diversify project portfolio

Firm B—Moderate#Monthly resourcescheduling meetings byheads of groups, stressindustry and problemspecialization$

“I think ultimately the bestway to do it is by having areputation for doing reallyamazing things in difficultareas so people that havedifficult problems callyou” (CEO).

Firm C—Moderate#Monthly resourcescheduling meetings byheads of groups, stressindustry and problemspecialization$

“For efficiency, and theunderstanding that youdevelop—like if youworked on cell phones fora while, you know a lotabout them, and you aregoing to be very effectiveat doing the next one”(Director, MechanicalEngineering).

Firm D—Moderate#Resource schedulingmeetings by seniormanagement for newprojects, stress industryspecialization$

“When we get a new projectwe decide who’s the bestqualified to do this”(Design Director).

Firm E—Moderate#Monthly resourcescheduling meetings byheads of groups, stressindustry and problemspecialization$

“Everybody here is workingon like, 3 or 4 projects at atime ! ! ! so they might havea forté in certain areas ofproduct design” (Director,Industrial Design).

Firm E—Strong #“Highprofile” via breakthroughproducts, awards, press$

“Design is design, so we’repursuing bigger, moreFortune 500 companies”(Director, Industrial Design).

2. Risk-taking builds long-termadaptability

Firm A—Strong“The company does take risks:

employees are willing to dosomething differently interms of materials, trends,etc.” (Design Director).

Firm B—Strong“I mean this company always

has to stay cutting edge,keep young, you are forcedinto this position ! ! ! ”(Industrial Designer).

Firm C—Strong“People like to work on high

profile projects, like X andY, because they are Fortune500 companies, so you feellike creating the future ofthe industry ! ! ! a few workedon the X TV and it’s sellingreally well” (Senior Director,Engineering).

Firm D—Weak“We try to stay focused on the

top position of themarketplace and we getcertain types of projects thatfortunately are usually fun,fast, aggressive ones”(CEO).

Firm E—Strong“We are responsible for

shaping the future and tryingto make people’s lives easierand better” (Designer).

Firm D—Weak #Projectreviews reinforcedual-emphasis$

“We try to do a review afterevery project, just to look atwhat we accomplished whatwe got out of it” (DesignDirector).

Firm E—Moderate #Formalcommunication involvingall staff reinforcesdual-emphasis; one staffmeeting per month peroffice$

“! ! ! sometimes, they go downto talking figures, I thinkthat the more explicit theyare, the better. They also tellus if they are not doing thatwell, and what they’ll doabout it. There is also a partin which they discussmarketing—the projectscoming in and the proposalsthat we have sent out, andwhere we are. There is alsoone part which I think ispretty cool. It is called workin progress” (IndustrialDesigner).

“We may decide to take aproject in where it is a smallcompany that they have nomoney but we will get anequity interest in the thing”(CEO).

Firm C—Moderate #Initiatingown projects, receivingequity in high-risk clientsand developing jointventures$

“We developed a new kind ofproduct developmentpartnership where the designfirm, the retailer, themanufacturer, the licensebrand are all sitting at thetable” (Senior Manager).

Firm D—Strong #Initiatingown projects, receivingequity in high-risk clientsand developing jointventures—managed bythe Director of Ventures$

“We have hired a managingdirector just for that. He’sdeveloping this as a separatebusiness on how we getcompensated for the designwork that we do and theideas that we come up with.It’s royalty or equity inprojects” (CEO).

Firm E—Moderate #Initiatingown projects andreceiving equity inhigh-riskclients—managed by oneof the founders$

“We’ve done about 200 or 300X products since we sort ofco-founded the company”(Director, Industrial Design).

Notes. Strong: repeatedly indicated by a majority of the case informants; Moderate: indicated by some informants; Weak: indicated by afew informants.

"Strength of evidence.


Appendix 1(b). Data Table for Cross-Case Comparisons: Customer Orientation


Integration: Differentiation: Iterate betweenTight coupling Loose coupling Improvise purposefully project constraints and freedom

1. Achieving project goalsfosters client satisfactionand loyaltyFirm A—Strong"

“While listening to the clientsand trying to give them whatthey need, we will try to findopportunities to innovate”(Senior Industrial Designer).

Firm B—Strong“We spend a lot of time with

the client to understand whatthey are really trying to doand what we can do forthem” (CEO).

Firm C—Strong“That is where this firm’s

rationale comes in ! ! ! if wedon’t serve the clientsfirst ! ! !we can’t developthose client relationships”(Creative Director, IndustrialDesign).

Firm D—Strong“ ! ! ! and if you’re successful

the client keeps giving youmore and more” (CEO).

Firm E—Strong“So we kept raising the bar and

they kept wanting to go withus because we were willingto raise the bar with them”(Director, Engineering).

2. Clients’ requirements helpprojects fulfill market needsFirm A—Strong“We are in the business and

work about creating userexperiences that carry ourclients ! ! !we take clients’technology and their brandand create user experiencesout of that” (President).

Firm B—Strong“The client is the expert at

what they do, specifically totheir category of goods ortheir industry” (VP ProgramDevelopment).

1. Probing new products andtechnologies surfaces futureopportunitiesFirm A—Moderate #Project

and internally driven)“We are trying to give

employees the opportunityto be innovative and creativeand think about things in thefuture and aboutpossibilities” (GraphicDesigner).

Firm B—Moderate#Project-driven$

“In this kind of business youhave to evolve and try tostay ahead of potentialtrends and have a goodunderstanding of what isgoing to be around forever”(Senior Designer).

Firm C—Moderate#Project-driven$

“We pull a lot from ourhistory, but we are alsoconstantly looking forward”(Senior Manager).

Firm D—Moderate #Project-and internally driven$

“We are trying to helpestablish strategies formarkets to move forward”(Design Director).

Firm E—Strong#Project-driven$

“A product for a completelynew target market that youhave never really consideredis challenging. Bringingsomething new in terms ofmaterials” (IndustrialDesigner).

2. Ongoing experimentationextends firm knowledgebaseFirm A—Strong“They allow a lot of freedom

to experiment” (SVP,Design).

Leverage synergiesbetween current projectconstraints and emergingpossibilities

Firm A—Moderate“There are many pressures:ecological, businesses,opportunities tocollaborate in newdifferent ways, thecapabilities of theinternet to connect thingsand enable new things.

You’ve got designers whocan synthesizetechnology, reality(associated with ourphysical way with thingsthat would bemeaningful to society)and then they’ve got totake a point of view andput it together and usetheir own insight to dosomething with this. Ifyou try to throw one out,it is easy to misssomething” (CEO).

Firm B—Moderate“I think it’s very importantto the company that wedo work that has notonly immediate butlong-term impact on theworld and people ! ! ! thatwe do very thoughtfuldesign ! ! ! ‘how they aregoing to react to it?’‘how are they going tobe affected by it?’ ! ! ! Ithink we are aboutrespect ! ! ! respecting theuser, the client, eachother” (Director).

Firm C—Moderate“Another thing thatimproves creativity, Ithink, is creativeabrasion, where it’s justthe friction of thedevelopment process ! ! ! ”(Industrial Designer).

1. Begin by deepening understandingof client expectations and marketlandscapeFirm A—Strong“For example, last Friday we went

on the X store to see theproducts and projects that X isworking on” (SeniorMechanical Engineer).

Firm B—Strong“We take all these into

consideration and what we tryto do is always relate ourrecommendations back to theconsumer or the user or thefeedback of the marketplace”(Director, Industrial Design).

Firm C—Strong“We look at other products”

(Director, MechanicalEngineering).

Firm D—Strong“We try to really collect as much

information and immerseourselves in that as possible,anything from magazines andmovies to the internet to buyingall the competitors’ productsand tearing them apart” (CEO).

Firm E—Strong“The company’s philosophy ! ! !or

let’s say the positioning whichis somewhat based on thephilosophy is that we are reallytrying to improve products forpeople ! ! ! and we do thatthrough really observing andwe really try to understand howpeople live, how people behave,how people interact with theproducts and through thatprocess we try to identifyopportunities” (Senior VicePresident).

2. Gradually pull away from initialconstraints, then return to addressclient requirementsFirm A—Strong“I am often motivated by working

on the first phase, the


Appendix 1(b). (cont’d.)


Integration: Differentiation: Iterate betweenTight coupling Loose coupling Improvise purposefully project constraints and freedom

Firm C—Strong“I have the people involved in

the project right in with theclient meeting. I want themto hear first hand what theclient needs ! ! !because theyneed to understand theclient’s goal” (SeniorDirector, Engineering).

Firm D—Strong“We are very open and

collaborative with ourclients” (CEO).

Firm E—Strong“So usually all the client

criteria and marketingcriteria and all these thingsdon’t hamper the creativitytoo much, because youexpect these criteria”(Director, Industrial Design).

Firm B—Strong“There is something about

carving stuff with your ownhands that allows you tounderstand exactly what isgoing on with that form thatyou cannot necessarily getwith the mathematical modelof the computer” (Director,Industrial Design).

Firm C—Strong“In that early phase of concept

development, they might bebetter just building a quickmodel, or justcommunicating the ideadifferently” (Director,Mechanical Engineering).

Firm D—Strong“We get mock-ups done”

(CEO).

Firm E—Moderate“ ! ! ! and also materials and

technology—bringingsomething new in terms ofmaterials is challenging”(Industrial Designer).

Firm D—Moderate“Everybody is reallydedicated to their projectand (the client), but theyalso want to push itfurther” (Designer).

Firm E—Moderate“There was a little bit ofdirection, like anunderlying core ofdesign language comingfrom the clients’ side,that was developed there,and was part of theirproduct lineage history,but they are looking forthe next thing ! ! ! this is abig trend ‘plannedobsolescence’; it’s greatfor designers in a way,because companies needto update their products,so they are reallylooking for fresh ideas”(Director, IndustrialDesign).

conceptual phase ! ! !we giveourselves more freedom, like Itold you before presenting awide range of concepts and thatis pretty fun versus therefinement phase made for aclient, which does not involvetaking risks, and it is going tobe a matter doing a lot of CADwork ! ! ! it is not going to be asmuch exciting” (IndustrialDesigner).

Firm B—Moderate“Creativity actually is ! ! ! to

envision a possibility that doesnot exist. Designers removeconstraints ! ! ! to envision analternate product, an alternatetechnology, an alternate life,something very different ! ! ! !”(Principal Engineer).

Firm C—Strong“People are more open, like the

clients realize, like, ‘yeah, let’sgo a little bit wider first, andthen narrow it later’ ” (SeniorIndustrial Designer).

Firm D—Moderate“When we first start our

brainstorming session we try tolet (client constraints) go ! ! !because usually the really crazyideas tend to trickle down andmaybe a part of that can still beapplied to what the end resultwill be” (Design Director).

Firm E—Strong“Everybody contributes with their

expertise in all stages. But Ithink that in an initial phase,everyone is still involved togenerate ideas, we usually havebrainstorming sessions, but thenyou definitely need the teamleader to hold the vision of theproduct” (Senior IndustrialDesigner).




Appendix 1(c). Data Table for Cross-Case Comparisons: Personal Drivers


Integration: Socialize Differentiation: Temporally andDiscipline Passion “practical artists” spatially separate work modes

1. Well-defined development processempowers contributionFirm A—Strong"

“The way that we do it (minimize failure)is by having processes in place ! ! ! ifyou have a process, you buildprototypes and you do things to testthe design before you release” (COO).

Firm B—Strong“I see the structure of the process, I

understand the process, and how I cancontribute to the process ! ! ! ! We don’twant to say, ‘this is our process, justmemorise it’ it’s like a living body ! ! ! itkeeps changing” (Principal, ID).

Firm C—Strong“We change the definitions of these

things all the time, but right now, Iwould say that we have a 6-stageprocess that starts with zero, and goesto five” (Director, MechanicalEngineering).

Firm D—Moderate“Product Design is A to Z ! ! ! everybody

kind of touches one point or another ofthat” (Designer).

Firm E—Strong“ ! ! !we have team leaders and a team

structure to help make sure that theprojects are staffed for successfulresults. I think that our process hasbecome better refined over the years”(President).

2. Targets encourage executionFirm A—Strong #Defined by senior

management$“I am responsible for understanding what

is going on in the projects. I am alsoresponsible for scheduling,profitability, etc.” (Design Director).

Firm B—Strong #Defined by projectleaders$

“We have a lot more internal structureand discipline, true projectmanagement ! ! ! to make sure theproject is going according to plan fromboth sides. So it is important to have acertain degree of discipline” (VP,Program Development).

Personal expression,challenge andpride motivateknowledgeworkers’ creativity

Firm A—Strong“The creative side

for me is how Ido my work, whatnew tools I canuse, what I bringin, what kind ofexpertise I thinkwe’re going to beneeding in thefuture” (Director).

Firm B—Strong“Creativity occurs

within somestructure but theactual processitself is not overlystructured” (VP,ProgramDevelopment).

Firm C—Strong“I think that the

answer is todevelop the rightkind of flexibility,first of all, so thatyou can customizethe process toaddress theparticular clientproblem, and thensecondly, just tocommunicate wellenough within theteam to make thenew criteria andmilestones clear inevery case”(Director,MechanicalEngineering).

Nurture paradoxicalwork identities#discipline andpassion asinterwoven elementsof professionalsuccess$

Firm A—Strong“People haveflexibility toexercise theirpassion and thecompany hirespassionate peoplewho are alsobalanced anddiverse” (DesignDirector).

Firm B—Moderate“I think there areopportunities inphases for people tobe creative andopen-ended aboutwhat they’redoing ! ! ! and then,this is where thesenior people comein; my job isnavigation andediting” (Principal,ID).

Firm C—Strong“Well, there’s rightand left brain—weconstantly talkabout teachingpeople how toswitch from oneside to the otherbecause it isrealistic, and as abusiness, we cannotstay in business ifthey don’t do theirtimesheets, or ifthey don’t fill theirreports out becausethis is not an artstudio where wecan sit around andpaint all day”(Creative Director,Industrial Design).

1. Vary nature of work during differentprojects and project phasesFirm A—Strong“You can definitely use your creativity

for both things but it is a little bitdifferent in the blue-sky projects; wegive ourselves the subject and use ourimagination, versus some lessexploratory projects” (IndustrialDesigner).

Firm B—Moderate“There are times where we may explore

a bunch of different directions on whatthis project might work like, how itmight function, what the architecturallayout of the product might be, that isvery open. While in other times wehave to narrow the ideas down” (VP,Program Development).

Firm C—Strong“The main point of phase zero, I would

say, is to define the problem ! ! ! let’sgather all the information that we canabout what this product should be andthen distil that down to something thatis very clear. In Phase one, it isbasically concept exploration. And thegoal there ! ! ! is to come up with lots ofideas. And so it is sketching, it isbuilding little models ! ! ! it is trying tofigure out what the possibilitiesare ! ! ! ” (Director).

Firm D—Moderate“Making the product move along

efficiently and successfully based onthe fact that maybe one designer mightbe better at sketching, another mightbe better at modeling” (Designer).

Firm E—Strong“Well, we have you know five phases

that we all kind of follow, but withinthese phases there may be differentkind of things that we may do ! ! !butgenerally, you know that in thebeginning, you’re this wide and yournarrowing down, and narrowing down,that’s understood” (Senior DesignResearcher).


Appendix 1(c). (cont’d.)


Integration: Socialize Differentiation: Temporally andDiscipline Passion “practical artists” spatially separate work modes

Firm C—Strong #Defined by projectleaders$

“We set deliverables ! ! ! ! It is all aboutwhat is going to happen tomorrow, whatis going to happen by the end of theweek—that is like, as far out as welook sometimes” (Senior IndustrialDesigner).

Firm D—Strong #Defined by seniormanagement$

“Basically (we tell them) ‘here is theproject, here is what we have to do,here are some deadlines, somedeliverables to make’—that gives themsome structure” (CEO).

Firm E—Strong #Defined by projectleaders$

“Teams and clients understand what ismeant by certain deliverables” (VP,Engineering).

3. Explicit roles enable focusFirm A—Moderate #Static, assigned by

senior management$“There is one person who organizes

everything well, he seems to be able todelegate things well and get peopleinvolved” (Graphic Designer).

Firm B—Strong #Static, assigned byproject leaders$

“There is generally a fairly explicitstructure, we know who is running it.We know generally what our roles are”(Industrial Designer).

Firm C—Strong #Dynamic, assignedby project leaders$

“The process is mapped out from the start! ! ! as best as it can be, but it’s alwayson shifting sand” (Industrial Designer).

Firm D—Moderate #Static, assignedby senior management$

“They (employees) need to feel like partof a team but they have their own role”(CEO).

Firm E—Strong #Static, assigned byproject leaders$

“You would notice order because we’revery serious about these infrastructures,you know, meetings happeningregularly, team meetings and coaches,who are more experienced people,making sure that they’re keepingtabs ! ! ! ” (Director, Industrial Design).

Firm D—Strong“My personal

philosophy is thatdesigners are kindlike free radicalelectrons. You letthem do theirthing, and everyonce in a whileyou gather themup and focus themon something andthey become verypowerful. Then assoon as that isdone and theyproduce what theyneeded toproduce, you letthem go back andbe free radicalsagain” (CEO).

Firm E—Strong“Have a very

commonfoundation butwithin that weallow moreflexibility” (VP,Engineering).

Firm D—Moderate“There is only acertain type ofperson who canreally fitanyway ! ! ! a personwho could probablyadapt, but it isdefinitely the moreopen, moreself-motivatedindividual who canget stuff done andhas to do both”(Designer).

Firm E—Strong“Some things youmay pick upthrough some of theculture ! ! ! there arethings that peoplepick up throughosmosis” (Director,Industrial Design).

2. Segregate routine/administration andnonroutine/creation responsibilities

Firm A—Moderate #Unclear$“You’re either billing to clients or you’re

billing to nonbillable categories likemarketing, general administration,etc.” (COO).

Firm B—Strong #Clear$“I think that there’s a lot of effort that is

done with our business developmentgroup, our marketing group and withother key people in the company whoare looking to bring in new business”(Director).

Firm C—Strong #Clear$“We have a group of business

development people, we have a groupof strategists, a group of generalmanagers ! ! ! and then we haveproducers also ! ! ! ! These peoples’ jobis to think about the client, and thebusiness objectives. So, the designersare somewhat insulated ! ! ! ” (Director).

Firm D—Moderate #Unclear$“Everyone here wears a lot of hats ! ! ! it’s

a pretty flat organization ! ! !we have afew titles” (Design Director).

Firm E—Strong #Clear$“Designers think in pictures, certainly not

numbers but it doesn’t always workwell on all projects and all clients.Some clients want clear minutesdistributed the day after the meetingand they want to be able to talk tosomeone about the billing, and theexpenses and this and that. Even ifdesigners are good at it, ‘do we reallywant them to spend 3% of their timegoing over bills as opposed todesigning?’ So, that is why we haveproject coordinators” (Director).




ReferencesAlvesson, M. 1995. Management of Knowledge-Intensive Companies.

De Gruyter, Berlin and New York.Amabile, T. M., S. G. Barsade, J. S. Mueller, B. M. Staw. 2005. Affect

and creativity at work. Admin. Sci. Quart. 50 367–403.Atuahene-Gima, K. 2005. Resolving the capability-rigidity paradox

in new product innovation. J. Marketing 69 61–83.Beckman, C. M. 2006. The influence of founding team com-

pany affiliations on firm behavior. Acad. Management J. 49741–758.

Benner, M. J., M. Tushman. 2002. Process management and techno-logical innovation: A longitudinal study of the photography andpaint industries. Admin. Sci. Quart. 47 676–706.

Birkinshaw, J., C. Gibson. 2004. Building ambidexterity into an orga-nization. MIT Sloan Management Rev. 45 47–55.

Brown, J. S., P. Duguid. 2001. Creativity versus structure: A usefultension. MIT Sloan Management Rev. 42 93–94.

Cameron, K. S., R. E. Quinn. 1988. Organizational paradox andtransformation. R. E. Quinn, K. S. Cameron, eds. Paradox andTransformation: Toward a Theory of Change in Organizationand Management. Ballinger, Cambridge, MA, 12–18.

Cohen, J. 1960. A coefficient of agreement for nominal scales. Ed.Psych. Measurement 20 37–46.

Cohen, W. M., D. A. Levinthal. 1990. Absorptive capacity: A newperspective on learning and innovation. Admin. Sci. Quart. 35128–153.

Collins, J. C., J. I. Porras. 1994. Built to Last: Successful Habits ofVisionary Companies. HarperCollins, New York.

Danneels, E. 2003. Tight-loose coupling with customers: The enact-ment of customer orientation. Strategic Management J. 24559–576.

Delmestri, G., F. Montanari, A. Usai. 2005. Reputation and strengthof ties in predicting commercial success and artistic merit ofindependents in the Italian feature film industry. J. ManagementStud. 42 975–1002.

Denison, D. R., R. Hooijberg, R. E. Quinn. 1995. Paradox and perfor-mance: Toward a theory of behavioral complexity in managerialleadership. Organ. Sci. 6 524–540.

Dougherty, D. 1996. Organizing for innovation. S. R. Clegg, C. Hardy,W. R. Nord, eds. Handbook of Organization Studies. Sage,Thousand Oaks, CA.

Duncan, R. 1976. The ambidextrous organization: Designing dualstructures for innovation. R. Kilman, L. Pondy, eds. The Man-agement of Organizational Design. North Holland, New York,167–188.

Eisenhardt, K. M. 1989. Building theories from case study research.Acad. Management Rev. 14 532–550.

Eisenhardt, K. M. 2000. Paradox, spirals, ambivalence: The new lan-guage of change and pluralism. Acad. Management Rev. 25703–705.

Eisenhardt, K. M., L. J. Bourgeois. 1988. Politics of strategic deci-sion making in high-velocity environments: Toward a midrangetheory. Acad. Management J. 31 737–770.

Eisenhardt, K. M., M. E. Graebner. 2007. Theory building fromcases: Opportunities and challenges. Acad. Management J. 5025–32.

Forster, N. 1994. The analysis of company documentation. C. Cassell,G. Symon, eds. Qualitative Methods in OrganizationalResearch. Sage, Thousand Oaks, CA.

Ghoshal, S., C. A. Bartlett. 1997. The Individualized Corporation:A Fundamentally New Approach to Management. Harper Busi-ness, New York.

Gibson, C. B., J. Birkinshaw. 2004. The antecedents, consequencesand mediating role of organizational ambidexterity. Acad. Man-agement J. 47 209–226.

Glaser, B., A. Strauss. 1967. The Discovery of Grounded Theory:Strategies for Qualitative Research. Aldine, New York.

Gupta, A. K., K. G. Smith, C. E. Shalley. 2006. The interplaybetween exploration and exploitation. Acad. Management J. 49693–706.

Hambrick, D. C. 1994. Top management groups: A concep-tual integration and reconsideration of the “team” label.L. L. Cummings, B. M. Staw, eds. Research in OrganizationalBehavior, Vol. 16. JAI Press, Greenwich, CT, 171–213.

Hargadon, A., B. Bechky. 2006. When collections of creatives becomecreative collectives: A field study of problem solving at work.Organ. Sci. 17 484–500.

Hargadon, A., A. Fanelli. 2002. Action and possibility: Reconcilingdual perspectives of knowledge in organizations. Organ. Sci. 13290–302.

Hargadon, A., R. I. Sutton. 1997. Technology brokering and inno-vation in a product development firm. Admin. Sci. Quart. 42716–749.

Hargadon, A., R. I. Sutton. 2000. Building an innovation factory. Har-vard Bus. Rev. 78(3) 157–166.

Jansen, J. J. P., F. A. J. van den Bosch, H. W. Volberda. 2006.Exploratory innovation, exploitative innovation, and perfor-mance: Effects of organizational antecedents and environmentalmoderators. Management Sci. 52(11) 1661–1674.

Lampel, J., T. Lant, J. Shamsie. 2000. Balancing act: Learningfrom organizing practices in cultural industries. Organ. Sci. 11263–269.

Leonard, D., W. Swap. 1999. When Sparks Fly: Igniting Creativity inGroups. Harvard Business School Press, Cambridge, MA.

Levitt, T. 2002. Creativity is not enough. Harvard Bus. Rev. 80137–145.

Lewis, M. W. 2000. Exploring paradox: Toward a more comprehen-sive guide. Acad. Management Rev. 25 760–776.

Lubatkin, M. H., Z. Simsek, Y. Ling, J. F. Veiga. 2006. Ambidexterityand performance in small-to-medium-sized firms: The pivotalrole of top management team behavioral integration. J. Man-agement 32 646–672.

March, J. G. 1991. Exploration and exploitation in organizationallearning. Organ. Sci. 2 71–87.

Mauzy, J., R. A. Harriman. 2003. Three climates for creativity. Res.Tech. Management 46 27–30.

McDonough, E. F., R. Leifer. 1983. Using simultaneous struc-tures to cope with uncertainty. Acad. Management J. 26727–735.

Miles, M. B., A. M. Huberman. 1994. Qualitative Data Analysis.Sage, Thousand Oaks, CA.

Miner, A. S., P. Bassoff, C. Moorman. 2001. Organizational impro-visation and learning: A field study. Admin. Sci. Quart. 46304–337.

O’Reilly, C. A., M. L. Tushman. 2004. The ambidextrous organiza-tion. Harvard Bus. Rev. 82(4) 74–81.

Poole, M. S., A. H. Van de Ven. 1989. Using a paradox to buildmanagement and organization theories. Acad. Management R.14 562–578.

Puranam, P., H. Singh, M. Zollo. 2006. Organizing for innovation:Managing the coordination-autonomy dilemma in technologyacquisitions. Acad. Management J. 49 263–280.


Raisch, S., J. Birkinshaw. 2008. Organizational ambidexterity:Antecedents, outcomes, and moderators. J. Management 34375–409.

Robertson, M., J. Swan. 2003. “Control—What control?” Culture andambiguity within a knowledge intensive firm. J. ManagementStud. 40 831–858.

Robinson, A. G., S. Stern. 1998. Corporate Creativity. Berret-KoehlerPublishers, San Francisco.

Sharma, A. 1999. Central dilemmas of managing innovation in largefirms. California Management Rev. 41 149–165.

Sheremata, W. A. 2000. Centrifugal and centripetal forces in radicalnew product development under time pressure. Acad. Manage-ment Rev. 25 389–408.

Slaatte, H. A. 1968. The Pertinence of Paradox. Humanities Press,New York.

Smith, K. G., C. J. Collins, K. D. Clark. 2005. Existing knowledge,knowledge creation capability, and the rate of new productintroduction in high-technology firms. Acad. Management J. 48346–357.

Smith, W. K., M. L. Tushman. 2005. Managing strategic contra-dictions: A top management model for managing innovationstreams. Organ. Sci. 16(5) 522–536.

Spradley, J. P. 1979. The Ethnographic Interview. Holt, Rinehart andWinston, New York.

Strauss, A., J. Corbin. 1990. Basics of Qualitative Research, Sage,Newbury Park, CA.

Subramaniam, M., M. A. Youndt. 2005. The influence of intellectualcapital on the types of innovative capabilities. Acad. Manage-ment J. 48 450–463.

Sutton, R. I., A. Hargadon. 1996. Brainstorming groups in context:Effectiveness in a product design firm. Admin. Sci. Quart. 41685–718.

Taylor, A., H. R. Greve. 2006. Superman or the Fantastic Four? Knowl-edge combination and experience in innovative teams. Acad.Management J. 49 723–740.

Tushman, M. L., C. A. O’Reilly. 1996. Ambidextrous organizations:Managing evolutionary and revolutionary change. CaliforniaManagement Rev. 38(4) 8–30.

Van Maanen, J. 1979. The fact of fiction in organization ethnography.Admin. Sci. Quart. 24 539–550.

Wadhwa, A., S. Kotha. 2006. Knowledge creation through exter-nal venturing: Evidence from the telecommunications equip-ment manufacturing industry. Acad. Management J. 49819–835.

Wheelwright, S. C., K. B. Clark. 1992. Revolutionizing ProductDevelopment. Free Press, New York.

Yin, R. 1994. Case Study Research: Design and Methods. Sage,Beverly Hills, CA.

Documents

Exploitation-Exploration Tensions and Organizational ... · Exploitation-Exploration Tensions and Organizational Ambidexterity: Managing Paradoxes of Innovation Constantine Andriopoulos