The Importance of Proximity for the Start-Ups' Knowledge Acquisition and Exploitation

The Importance of Proximity for the Start-Ups’Knowledge Acquisition and Exploitationjsbm_331 361..389

by Manuela Presutti, Cristina Boari, and Antonio Majocchi

This paper intends to verify the impact of geographical proximity on the processesof knowledge acquisition and exploitation by high-tech start-ups considering at thesame time the role of both the social and cognitive dimensions of proximity. Our basicassumption is that proximity means a lot more than just geography. The findings fromthis research broaden our understanding of how start-ups located inside an indus-trial cluster acquire knowledge from their customers and exploit it in an innovativeway, underscoring the need to reconsider assumptions regarding the importance ofgeographical proximity between business partners during knowledge management.

IntroductionThe analysis of the potential forces

acting on knowledge acquisition andexploitation by start-ups has receivedmuch attention by strategic managementresearchers (Ahuja 2000; Yli-Renko,Autio, and Sapienza 2001). Much hasbeen written (Baptista 2000; Hagedoornand Duysters 2002; West and Noel 2009)on the impact of external social relation-ships on start-up knowledge acquisitionand exploitation processes (Amesse andCohendet 2001; BarNir and Smith 2002;

Lundvall 1988; Neck et al. 2004). Accord-ing to the idea that knowledge is themain source of their competitive advan-tage (Kogut and Zander 1992; Spender1996), start-ups have been thought to bethe most affected by external networksduring their knowledge processes(Eisenhardt and Schoonhoven 1996;Smith, Matthews, and Schenkel 2009).Geographical proximity between astart-up and its business partners is gen-erally assumed to reinforce these twoprocesses since knowledge is partiallytacit and localized (Cooke and Willis

Manuela Presutti is associate professor of management at the Department of Managementof the University of Bologna.

Cristina Boari is full professor of business strategy at the University of Bologna.Antonio Majocchi is associate professor of international business and management at the

Faculty of Economics, University of Pavia.Address correspondence to: Manuela Presutti, The University of Bologna, Department of

Management, Via Capo di Lucca, Bologna 34 40126, Bologna 40126, Italy. E-mail:[email protected].

Journal of Small Business Management 2011 49(3), pp. 361–389

PRESUTTI, BOARI, AND MAJOCCHI 361

1999; McEvily and Marcus 2005). In par-ticular, since Marshall’s (1920) seminalwork on agglomeration economies, bothstrategic studies and economic geogra-phers have considered geographicalproximity the key parameter that start-ups can use to increase their exposure topotential knowledge spillovers (Alcacerand Chung 2007; Audretsch andLehmann 2006; Freel 2003; Macphersonand Holt 2007).

However, these studies have paidlittle or no attention to other dimensionsof proximity (Beal and Gimeno 2001;Boschma and Frenken 2006; Boschmaand Lambooy 1999). Recent evidencefrom economic geographers suggeststhat although geographical proximitymay facilitate interactive learning, it isneither a prerequisite nor a sufficientcondition for reinforcing the processes ofknowledge acquisition and exploitationby co-localized start-ups (Antonelli 2000;Boschma 2005; Rallet and Torre 2000).These researchers point out that otherdimensions of proximity besides geo-graphical proximity are key in under-standing the acquisition and exploitationof knowledge by local start-ups(Boschma and Lambooy 1999; Torre andGilly 2000).

In this paper, we intend to verifyempirically the impact of geographicalproximity on the processes of knowledgeacquisition and exploitation by co-localized start-ups while also consideringthe role of both social and cognitivedimensions of proximity (Boschma 2005;Kaiser 2002; Liao and Welsch 2005;Powell et al. 2002). We agree with theidea of using the concept of proximity ina multidimensional way, analyzing othernontangible dimensions of proximity assuggested by Boschma (2005). Our finalaim is to integrate the framework ofstrategic management researchers—concerning the importance of geographi-cal proximity between a start-up and itsmain customers for the start-up’s pro-cesses of knowledge acquisition and

exploitation—with an in-depth analysisof both social and cognitive dimensionsof proximity according to the economicgeographic literature (Audretsch andLehmann 2006; Boschma 2005; Ralletand Torre 2000).

The findings from this researchbroaden our understanding of how start-ups located inside an industrial clusteracquire knowledge from their customersand exploit it for their innovation activ-ity. In particular, we recognize the needto reconsider assumptions regarding theimportance of geographical proximitybetween a start-up and its customersduring start-up processes of knowledgeacquisition and exploitation (Alcacer andChung 2007; Keeble et al. 1999; Soren-son, Rivkin, and Fleming 2006). Thepaper is structured as follows: thesecond section reviews the related litera-ture, whereas the third section developsthe set of hypotheses. The fourth sectionpresents the sample, the data, and theapplied methods. The fifth section pre-sents the results of the empirical analy-sis. Finally, the paper concludes with adiscussion and suggestions for furtherresearch.

Theoretical BackgroundAs maintained by the knowledge-

based view of the firm (Kogut andZander 1992; Spender 1996), the compe-tition among new ventures has shiftedfrom natural resources to knowledgeassets since knowledge is the mainsource of sustainable competitive entre-preneurial advantage (Nonaka, Toyama,and Nagata 2000; West and Noel 2009).The acquisition of knowledge opens newbusiness opportunities and reinforcesstart-up ability to exploit these opportu-nities because of the ambiguity and thedifficulty to imitate the knowledgeresource (Grant 1996). Based on the evi-dence that the accumulation of knowl-edge constitutes a driving force in thedevelopment of young firms (Autio, Sapi-enza, and Almeida 2000), in recent years,

JOURNAL OF SMALL BUSINESS MANAGEMENT362

the analysis of the potential forces actingon their processes of knowledge acqui-sition and exploitation has receivedmuch attention by strategic managementresearchers (Ahuja 2000; BarNir andSmith 2002; Smith, Matthews, and Schen-kel 2009; Yli-Renko, Autio, and Sapienza2001). These studies tend to concentrateon the role of interfirm networks in facili-tating both knowledge acquisition andknowledge exploitation by start-ups fortheir innovation activity (Baptista 2000;Dyer and Singh 1998; Liao and Welsch2005). They claim that the start-up’sgrowth is strongly dependent on a prof-itable combination between its internalspecific knowledge and that of externalbusiness partners because new firms areresource-constrained (Eisenhardt andSchoonhoven 1996; McDougall, Shane,and Oviatt 1994). Thus, since resourcelimitations cause the traditional prob-lems of liabilities of newness and adoles-cence (Stinchombe 1965; Yli-Renko,Autio, and Sapienza 2001), the ability ofstart-ups to leverage external networksfor knowledge acquisition and exploita-tion may assure their survival andgrowth (West and Noel 2009). In thisperspective, knowledge acquisition andexploitation are not seen as discrete pro-cesses resulting from knowledge devel-oped by isolated actors (Baptista andSwann 1998; Hagedoorn and Duysters2002), but they are conceptualized as aresult of interaction between businesspartners (Amesse and Cohendet 2001;Lundvall 1988). Indeed, since the knowl-edge is partially tacit and localized, theprocesses of knowledge acquisition andexploitation by a start-up require fre-quent interaction with external businesspartners, which is facilitated by geo-graphical proximity (Cooke and Willis1999; Hendry and Brown 2006; Mc-Kelvey, Alm, and Riccaboni 2003).

An apparent paradox is revealed inthese studies on the location of new ven-tures: despite the intangible nature ofnew ideas and their potential to diffuse

widely, small firms, especially in high-tech industries, are often geographicallyclustered (Bathelt, Malmberg, andMaskell 2004; Gittelman 2001; Gordonand McCann 2005). Definitions of indus-trial clusters range from studies that refermainly to geographic collectivity—as ageographic concentration of intercon-nected companies, specialized suppliers,related firms, and institutions in particu-lar fields that compete but also cooperate(Porter 1990)—to those that emphasizethe knowledge-sharing aspects of suchgroupings (Fujita, Venables, andKrugman 1999; Majocchi and Presutti2009).

Much has been written on the impor-tance of co-localization between astart-up and its partners inside anindustrial cluster for the start-up’s pro-cesses of knowledge acquisition andexploitation according to the economicgeography literature. In particular, sinceMarshall’s (1920) seminal work onagglomeration economies, disciplinessuch as economics, planning, sociology,strategic management, and businesshistory have emphasized the consider-able advantages of being colocated forknowledge processes in both the hori-zontal and the vertical dimension oflocal relationships (Gordon and McCann2005; Maskell and Malmberg 1999;Porter 2000). Some recent studies findthat collaborations with similar firms—horizontal dimension—in the localizedcluster are not significant mechanismsto access knowledge, differently fromvertical relationships (customer–supplier) (Audretsch and Lehmann2006; Malmberg and Power 2005).

In the horizontal dimension, the inter-action with local competitors with similarcapabilities and activities stimulates newexperiments and combination of knowl-edge. In the vertical dimension, learningbetween suppliers and customers isstimulated because low coordinationcosts in clusters encourage increasingspecialization, which differentiates the


internal knowledge bases of these localsuppliers (Porter 1990, 2000; Sorensonand Audia 2000). This is explained by theevidence that spatially concentratedstart-ups benefit from knowledge spill-overs as short distances literally bringbusiness partners together, favor infor-mation contacts, and facilitate theexchange of tacit knowledge (Gordonand McCann 2005; Howells 1999;Maskell 2001). Consequently, the largerthe distance between a start-up and itsbusiness partners actors, the less theintensity of these positive externalitiesand the more difficult it becomes for astart-up to acquire and exploit externalknowledge, especially in the case of tacitknowledge (Amin and Wilkinson 1999;Baptista and Swann 1998; Dahl and Ped-ersen 2005; Jaffe, Trajtenberg, and Hend-erson 1993). As a matter of fact, thisapproach claims that this may be trueeven for the use and spread of codifiedknowledge because its interpretationand assimilation may still require tacitknowledge and thus spatial closeness(Audretsch and Lehmann 2006; Freel2003; Howells 1999; Stuart and Sorenson2003).

In this respect, since start-up knowl-edge processes have been thought to bemost affected by external relationships, alot of recent studies on industrial clusters(Canina, Enz, and Harrison 2005; Folta,Cooper, and Baik 2006; Gilbert, McDou-gall, and Audretsch 2008) argue thatbecause of their liability of newness, lackof history and established proceduresand routines, and typically smaller sizes,new firms may be more likely to benefitfrom a cluster location than establishedfirms. For new ventures with limitedlegitimacy and whose resources are com-paratively limited, local knowledge spill-overs, guaranteed by the closeness tobusiness partners, can help reduce thetraditional uncertainty associated withknowledge acquisition and exploitation.Knowledge spillovers should help start-ups become knowledgeable about inno-

vation possibilities that have beensuccessfully employed by other colo-cated business actors in order to presentan offer aligned with market expecta-tions (Porter 2000). Moreover, local start-ups benefit because they are able to takeadvantage from the proximity with cus-tomers who may be attracted to coloca-tion by the reputation of the larger firmswithin the region (Antonelli 2006;Gilbert, McDougall, and Audretsch2008). It allows start-ups to more easilybring to the final market products thatare strongly in line with the needs andspecific interests of customers because ofthe significant knowledge transferassured by local knowledge spillovers(Howells 1999). In summary, the geo-graphical proximity with external busi-ness partners, reinforced by high levelsof trust and cognitive identification,increases the start-up knowledge acqui-sition from these external relationships,which is instrumental in fostering astream of innovation (knowledge exploi-tation) within the new ventures situatedinside a cluster (Freel 2003; Neck et al.2004). These studies also claim that start-ups located within industrial clusters arecharacterized by higher innovative per-formance, rates of growth, and survivalthan start-ups not located within geo-graphic clusters. New ventures assimilat-ing local knowledge spillovers shouldhave a better understanding of the indus-try and technological directions, higherlevels of performance than new venturesof the same industry located outside acluster (Canina, Enz, and Harrison 2005).

Although it is traditionally argued thatknowledge spillovers are geographicallyconstrained, recent studies in the geo-graphic field of research have pointedout that other dimensions of proximitybesides geographical proximity are keyin understanding start-up knowledgeacquisition and exploitation processes(Boschma 2005; Davenport 2005; Gittel-man 2001; Rosenkopf and Almeida2001). Although relatively few studies


have provided convincing empirical evi-dence of the superiority of local overnonlocal interactions in improving localstart-up knowledge acquisition andexploitation (Antonelli 2006; Oinas andMalecki 2002), some recent studies havehighlighted that the importance of geo-graphical proximity could stronglydepend on the nature of the exchangedknowledge (Sorenson, Rivkin, andFleming 2006). At the same time, a lot ofempirical studies have shown that distantrelationships between actors in globalproduction and distribution systems,rather than local social linkages, areindispensable for local start-up knowl-edge acquisition and exploitation pro-cesses (Carlsson 2003; Romanelli andKhessina 2005). In this respect, recentsignificant findings from innovation eco-nomics have shown that being sitedwithin an industrial cluster may constrainrather than enable the learning of start-ups because of the problem of lock-in(Beal and Gimeno 2001; Davenport2005; Nooteboom 2000). Too strong arelationship between colocated businesspartners may suffer from “overembed-dedness” (Uzzi 1997), both insulatinglocal start-ups from other external prof-itable sources of knowledge and infor-mation and inhibiting exchange andcombination of knowledge among localpartners (Burt 1992, 2000; Hansen 1999).

In short, this evidence suggests thatalthough geographical proximity mayfacilitate interactive learning, it is neithera prerequisite nor a sufficient conditionfor reinforcing knowledge acquisitionand its exploitation by start-ups(Antonelli 2000; Breschi and Lissoni2001, 2002) since other nontangibledimensions of proximity may act as sub-stitutes for geographical proximity(Boschma 2005; Boschma and Frenken2006; Rallet and Torre 2000). In particu-lar, these frameworks argue that theneed for geographical proximitybetween partners is weakened wherefour conditions are respected: the rela-

tions between actors are shared in anorganizational arrangement (organiza-tional proximity); the partners share thesame cognitive experience (cognitiveproximity); the relationships betweenactors are socially embedded, that is,they are characterized by a high level oftrust (social proximity); the actions ofactors are influenced by both formalinstitutions (i.e., rules and laws) andinformal institutions (cultural norms andhabits). However, in these frameworks,the cognitive proximity is separated fromthe organizational one and the institu-tional dimension is separated from socialproximity purely for analytical purposes(Boschma 2005). In fact, on one hand,the function of the institutions acts as asort of “glue” for collective actionbecause it reduces uncertainty andlowers transaction costs, reinforcing thedegree of trust (social proximity)between partners inside an intercon-nected system of relationships. On theother hand, organizational proximityincludes similarity: the actors are con-nected by sharing the same referencespace and knowledge so that they per-ceive, interpret, and evaluate the worldin a similar way (cognitive proximity). Insummary, these dimensions are so inter-related that it becomes very difficult toseparately consider their effects onknowledge acquisition and exploitationprocesses.

These frameworks point out that theinclusion of social proximity and cogni-tive dimensions of proximity reduces theneed for geographical proximity espe-cially in the case of tacit knowledge. As amatter of fact, also, these studies stressthat the exchange of codified knowledgestill requires physical copresence, which,however, could be organized by bringingbusiness actors together “through travelnow and then” (Boschma 2005). In otherwords, since these other intangibledimensions of proximity stress theimportance of networks as tools ofknowledge diffusion, social and cogni-


tive ties between actors and not geo-graphical proximity played a significantrole in knowledge spillovers becausethere is nothing inherently spatial aboutnetworks (Breschi and Lissoni 2001).Moreover, these studies empirically veri-fied that since networks are social andcognitive constructs, they can excludeoutsiders, whether or not they are localplayers (Breschi and Lissoni 2001;Hudson 1999; Morrison 2004), both inthe case of complex tacit knowledge butalso of simple codified knowledge(Pirolo and Presutti 2010; Sorenson,Rivkin, and Fleming 2006).

HypothesesThese different positions on the

importance of the geographical dimen-sion of proximity between business part-ners for their processes of knowledgeacquisition and exploitation suggest thatthere is a strong need for empiricalworks on these issues (Smith, Matthews,and Schenkel 2009; West and Noel 2009).We share the idea of using the concept ofproximity in a multidimensional way,including in the empirical analysis othernontangible dimensions of proximity assuggested by recent economic geogra-phers (Boschma 2005; Gittelman 2001;Presutti and Boari 2007). In particular,we intend to integrate the strategic man-agement framework on the importanceof geographical proximity for knowledgeacquisition and exploitation processeswith recent findings of economic geog-raphers that confirm also the significantinfluence of social and cognitive prox-imities on these two processes (Torreand Gilly 2000).

We discuss the research problemfocusing on the vertical business rela-tionships between a start-up locatedinside a cluster and its main customeraccording to several previous studies thathave empirically confirmed the impor-tance of a customer to reinforce bothnew product creation and the technologi-cal distinctiveness of its supplier (Yli-

Renko, Autio, and Sapienza 2001).Moreover, as already explained in thetheory section, the importance of a ver-tical dimension of local relationships hasbeen confirmed in several previousstudies on industrial clusters (BarNir andSmith 2002). The acquisition and exploi-tation of knowledge are treated as twoseparate steps of a start-up’s learningprocess (Lane and Lubatkin 1998). Westudy the process of knowledge acquisi-tion by a local start-up from its custom-ers, analyzing the importance of thecustomer’s providing strategic informa-tion about process, products, andresearch and development (R&D) activity(Antonelli 2000; Glaeser 2000; McEvilyand Zaheer 1999) useful to the localstart-up’s growth. We concentrate on thiskind of external technological knowl-edge (Yli-Renko, Autio, and Sapienza2001) because of the greater necessityfor a start-up to leverage interorganiza-tional customer relationships to broadenthe stock of knowledge concerning bothtechnological aspects and R&D activity(Pirolo and Presutti 2010). It is interest-ing to remark that a lot of previousstudies have considered this conceptual-ization of knowledge (technologicalknowledge) as a way to measure thetransfer of tacit knowledge by leveragingexternal networks (Huber 1991; Noot-eboom 2000; Von Hippel 1988).

The process of knowledge exploita-tion is studied through a networkapproach, assuming that in a start-up,the external knowledge acquired fromcustomers is exploited for increasinginnovation activity in terms of newproduct development (Antonelli 2000,2006). The importance of customers forthe launch of new products by a start-upis confirmed in a lot of traditional andmore recent studies (Zahra and George2002). A relationship with a customermay increase the start-up’s new productdevelopment in three ways: by enhanc-ing the breadth and depth of relation-specific knowledge; by reinforcing the


speed of product development throughreduced development cycles; and byincreasing the potential for new innova-tive combinations (Kaiser 2002; Yli-Renko, Autio, and Sapienza 2001).

We develop three research hypoth-eses according to three different selecteddimensions of proximity—geographical,social, and cognitive. In the analysis, weconcentrate only on these three dimen-sions of proximity for the followingreasons. First of all, as aforementioned inBoschma’s (2005) framework, the orga-nizational and institutional proximitiesare separated, respectively, from cogni-tive and social ones only for analyticalpurposes. It means that our measures ofcognitive and social proximity are able toindirectly capture both the organiza-tional and institutional dimensions ofproximity. Second, our interest in inter-organizational networks as profitabletools for a start-up to acquire and exploitexternal customer knowledge allows usto consider the networks as organiza-tional arrangements able not only tocoordinate transactions but also toenable the transfer and exchange ofinformation and knowledge (Cooke andMorgan 1998). In other words, sinceorganizational proximity is representedby the set of interdependence betweenactors “connected by a relation of eco-nomic or financial interdependence”(Kirat and Lung 1999), our paper consid-ers networks between a start-up and itsbusiness partners as a measure of orga-nizational proximity. Finally, as main-tained by a lot of studies on industrialclusters, the concept of geographicalproximity between colocated partnersalso corresponds to their organizationaland institutional proximity (Boschma2005; Porter 2000). It means that colo-cated actors are automatically subject tothe same institutional and organizationalarrangements. Inside an industrialcluster, the notion of institutional prox-imity includes both the idea of colocatedbusiness actors sharing the same institu-

tional laws and rules (formal institutions)and a set of cultural habits, norms, andvalues (Maskell and Malmberg 1999)(informal institutions). At the same time,inside an industrial cluster, the densenetwork of interfirm relationships ischaracterized by a certain level of orga-nizational proximity and is able toreduce uncertainty and opportunismbetween business partners, reinforcingtheir learning and innovation. In thisregard, the recent evolution of someindustrial clusters has resulted in thehierarchization of their network struc-ture, supporting the idea that colocatedactors are tied by a medium level oforganizational proximity since theirdense networks involve both cooperativeand competitive relationships (Albino,Carbonara, and Giannocaro 2005; Freel2003).

In our set of hypotheses, we first testthe impact of geographical proximitybetween a local start-up and its custom-ers on the start-up’s knowledge acquisi-tion and exploitation for its innovationactivity. In this respect, we dissociatefrom traditional assumptions of strategicand geographic researchers that maintainthe importance of geographical proxim-ity between partners for their knowledgeacquisition and exploitation processessince knowledge spillovers are geo-graphically concentrated (Almeida andKogut 1999; Audretsch and Feldman2003; Jaffe, Trajtenberg, and Henderson1993; Saxenian 1994). We agree with theidea that though a short geographicaldistance between business partnersbrings start-up and customers togetherand favors face-to-face interactions, it isneither a prerequisite nor a sufficientcondition for reinforcing the start-up’sknowledge acquisition and exploitation(Presutti and Boari 2007). Thus, a greatergeographical distance between businesspartners does not automatically reducethe intensity of the knowledge-basedexternalities (Liao and Welsch 2005). Atthe same time, the acquisition and


exploitation of knowledge betweendistant business partners is not more dif-ficult than between near business part-ners since actors do not always need tobe located nearby in order to take part ina process of knowledge diffusion(Hendry and Brown 2006; Presutti,Boari, and Fratocchi 2007). On the basisof these suggestions, we formulate thefirst research hypothesis, divided intotwo sub-hypotheses regarding, res-pectively, knowledge acquisition andexploitation:

H1a: A greater geographical proximitybetween a local start-up and its cus-tomers does not increase the probabil-ity that the local start-up acquiresknowledge from its customer.

H1b: A greater geographical proximitybetween a local start-up and its cus-tomers does not increase the probabil-ity that the local start-up exploitsexternal customer knowledge to rein-force its innovation activity.

In addition to the geographical prox-imity of partners, we discuss the impor-tance of both the social and cognitivedimensions of proximity between a localstart-up and its customers for the start-up’s knowledge acquisition and exploi-tation processes. Our interest in twodifferent dimensions of proximity is jus-tified by the recent evidence from geo-graphic studies showing how knowledgeacquisition and exploitation can beachieved between geographically dis-persed partners if they share the samecognitive experience (cognitive proxim-ity) or if their relationships are sociallyembedded, that is, involve high levels oftrust (social proximity) (Boschma andLambooy 1999; Boschma, Lambooy, andSchutjens 2002). However, we partiallydissociate from these assumptionsaccording to several researchers(Boschma 2005; Nooteboom 2000; Uzzi1997). In fact, we argue that though

social and cognitive proximities posi-tively impact on knowledge acquisitionof a local start-up from its customers,they influence in a negative way theknowledge exploitation by start-up forinnovation activity. We discuss sepa-rately these two dimensions of proximityand their impact on the knowledgeacquisition and exploitation processes bya local start-up.

First, we analyze the impact of socialproximity (Boschma 2005) between alocal start-up and its customers on thestart-up’s knowledge acquisition fromcustomers. The concept of social proxim-ity originates from the literature onembeddedness (Granovetter 1985),which suggests that the more sociallyembedded are the external business rela-tionships of a firm, the more interactivelearning it realizes with business part-ners reinforcing innovative performance.According to the embeddedness litera-ture, business relationships betweenactors are socially embedded when theyinvolve high levels of trust (Liao andWelsch 2005). The presence of trustinside business relationships makesreciprocal knowledge acquisition moreefficient, becoming a substitute for theformal contracts, incentives, and moni-toring mechanism associated with a busi-ness transaction (Zaheer, McEvely, andPerrone 1998). Indeed, interorganiza-tional trust reduces the risk of reciprocalopportunism, making relationships moreinformal and thus more effective foracquiring external knowledge (Lesser2000; Nahapiet and Ghoshal 1998)because of the effects of reputation andcooperation developed within trust-based networks (Burt 2000; Reagans andZuckerman 2001). These effects can bereinforced inside an industrial clusterwhere the way business partners coordi-nate their actions is strongly affected byboth formal and informal institutions. Insummary, looking at the issue of knowl-edge acquisition through the sociologicallens, social proximity between partners


reduces the two main critical obstacles toknowledge acquisition: the lack of cred-ibility and motivation of involved part-ners (Szulanski 1996) and the risk ofreciprocal opportunism (Uzzi 1997; Wil-liamson 1975).

Based on these suggestions, we for-mulate the second research hypothesis:

H2a: The greater the social proximitybetween a start-up and its customer,the higher the probability that thelocal start-up acquires knowledgefrom its customers.

The effects of social proximity onknowledge exploitation are quitecomplex. Although traditional studieshave considered interorganizational trusta necessary condition for innovation(Dyer and Singh 1998; Lundvall 1988;Nahapiet and Ghoshal 1998), a lot ofempirical studies confirm that high inter-organizational social proximity is detri-mental to learning and innovationprocesses implemented by involved part-ners (Glaeser 2000; Uzzi 1997). Weaccord to the idea that high social prox-imity has adverse impact on innovationaccording to the risk of over-embeddedness (Boschma, Lambooy, andSchutjens 2002; Uzzi 1997). In fact,embedded relationships in which muchsocial proximity is involved may lead toan underestimation of opportunism,weakening the capacity of a start-up toexploit external knowledge (Boschma2005; Gordon and McCann 2005). More-over, business relationships betweenstart-ups and customers with high levelsof social proximity are vulnerable toexogenous shocks, insulating suchstart-up from new and valuable informa-tion that exists beyond its actualnetwork. There is the danger of beingblind to new products development or ofbeing “locked-in” since high levels ofinterorganizational trust provide redun-dant information and knowledge, whichare not useful to the start-up’s innovative

performance (Gargiulo and Benassi1999; Hite and Hesterly 2001). In thisvein, strong social proximity becomes asignificant social liability for the start-up’s innovation growth. Based on theseconsiderations, we formulate the nextresearch hypothesis:

H2b: The greater the level of social prox-imity between a local start-up and itscustomer, the lower the probabilitythat the local start-up exploits cus-tomer knowledge to reinforce its inno-vation activity.

Then, we discuss the impact of cogni-tive proximity (Boschma 2005) betweenpartners on their processes of knowl-edge acquisition and exploitation(Kaplan and Norton 2004; Nooteboom2000). In literature, cognitive proximityis defined as the similarities in the wayactors perceive, interpret, and evaluatethe world (Nooteboom 2000). Therefore,a high cognitive proximity inside a rela-tionship can reduce the knowledge dis-tance among business partners,reinforcing the development of acommon knowledge base and expertise(Boschma 2005). We recognize thatshared cognitive proximity between alocal start-up and its customers is a nec-essary condition for the start-up’s knowl-edge acquisition from customers(Antonelli 2006; Lee, Lee, and Lee 2003)since the knowledge distance or “gap”between these two partners cannot begreat for improving the start-up’s acqui-sition of external knowledge. Sinceknowledge acquisition requires certainshared “interpretation systems” (Weick1979) or a “system of shared meanings”(Smircich 1983), a strong alignment interms of two partners’ knowledge is nec-essary for their knowledge acquisition(Cohen and Levinthal 1990; Szulanski1996). In essence, cognitive proximityallows a start-up to better understand itscustomers’ needs and to increaserelation-specific common knowledge


(Presutti, Boari, and Fratocchi 2007).Common knowledge, in turn, increasesrelation-specific absorptive capacitybecause the ability of a firm to acquire—that is, absorptive capacity—new exter-nal knowledge from a partner dependson the existing stock of related knowl-edge between these partners: similaritiesin current knowledge stocks seem toenhance the transfer of knowledge,whereas differences in knowledge stockstend to delay or prevent the absorptionof new knowledge from a partner(Cantwell and Santangelo 2002; Dahl andPedersen 2005; Lane and Lubatkin 1998;Zahra, Ireland, and Hitt 2000).

In summary, we argue that sharedexpectations and goals promote the cre-ation of compatible systems and culturesbetween a start-up and its customer,reinforcing the relative absorptivecapacities of the partners involved in aknowledge acquisition process (Maurerand Ebers 2006). Moreover, a significantlevel of cognitive proximity mitigates theinformation asymmetries concerning theinterfirm exchange by allowing a moreopen and honest acquisition of externalknowledge (Nooteboom 2000; Sako1998). These cognitive aspects are rein-forced inside an industrial cluster by thepresence of a deep and shared system ofrelationships that coordinate the localtransactions (organizational proximity),enabling the transfer and exchange ofinformation and knowledge betweenbusiness partners. According to theseconsiderations, we formulate the thirdresearch hypothesis:

H3a: The greater the level of cognitiveproximity between a local start-upand its customer, the higher the prob-ability that the local start-up acquiresknowledge from its customer.

Although cognitive proximity rein-forces the acquisition of external knowl-edge by a local start-up, we agree withseveral previous studies that suggest that

high levels of cognitive proximity may bedetrimental to the exploitation of exter-nal knowledge for innovation activity(Boschma, Lambooy, and Schutjens2002; Nooteboom 2000). First, cognitiveproximity tends to decrease the potentialprocess of learning by interacting, limit-ing the absorptive capacities of the part-ners useful to their innovation activity.Moreover, cognitive proximity may easilylead to cognitive lock-in in the sense thatfor a start-up, a routinized system ofbusiness relationships obscures the viewon new technologies and new marketpossibilities since a start-up is likely tostay with its existing familiar and closenetwork customers. Along this direction,a lot of studies explain the negativeimpact of cognitive proximity on innova-tion according to the network inertiaproblem (Boschma 2005; Burt 2000). Infact, when a start-up is developing a newproduct that deviates from existingknow-how shared with old customers,the presence of high interorganizationalcognitive proximity limits the ability of astart-up to exploit knowledge from dif-ferent external sources of innovation(Cantwell and Santangelo 2002;Lambooy and Boschma 2001; Uzzi 1997).Along this vein, the benefits of searchingfor and accessing heterogeneousresources have been stressed consider-ably in literature on innovation perfor-mance. Based on these suggestions, weformulate the last research hypothesis:

H3b: The greater the level of cognitiveproximity between a local start-upand its customer, the lower the prob-ability that the local start-up exploitscustomer knowledge to reinforce itsinnovation activity.

MethodsSample and Data

The field setting of this research con-sists of a geographical cluster of microand small high-tech firms in the electron-ics sector located about 14 km from the


center of Rome, in the so-called TiburtinaValley. This area, similar to Silicon Valleyin the United States, represents the mosttypical example of a high-tech metropoli-tan cluster in Italy. This cluster is char-acterized by a large concentration offirms with significant technological capa-bilities and a high-performing competi-tive presence both on national andforeign markets. The presence of a highnumber of specialized firms localized insuch a limited area led to the develop-ment of new businesses and the rapiddiffusion of knowledge among localactors. This confirms the idea of urbancluster as the result of a networkingprocess. As a result of these strong inter-connections, firm agglomerative tenden-cies can explain the process ofconcentration of economic activities inlocal clusters, especially in urban con-texts, where geographical proximity pro-motes social interaction. We tested thehypotheses using data from 54 electronicstart-ups located inside this cluster.

Before starting the empirical research,we carried out a preliminary detailedstudy through open interviews withselected actors within the cluster toreconstruct, at least from a qualitativeperspective, a map of the relationshipsamong the principal actors in the area.This qualitative step of the research hasbeen very useful to reinforce theresearch hypotheses, confirming theimportance of vertical relationships withcustomers for local start-up knowledgeacquisition and exploitation in compari-son with horizontal relationships charac-terized by a significant level ofcompetition. In particular, the start-upsof our sample belong to the “specializedsuppliers” category according to Pavittclassification since they produce andoffer technology and services mainly totheir customers. The sources of start-upknowledge acquisition and exploitationprocesses are strongly dependent on theinteractive learning with customers,which encourages local start-ups to

adopt a growing specialization causing astrong differentiation among theirknowledge internal bases. The launch ofnew products and the reinforcement ofR&D activity by these start-ups arestrongly conditioned by the knowledgeacquired from their customers since theoffer should be in line with the specificrequirements and needs of customers.For local start-ups, the lack of resourcesto patent could create a strong economicdependence on their customers (Presuttiand Boari 2007).

From data provided by the Chamberof Commerce of Rome, we observed that180 electronic firms were created in thisarea from 1985 to 2004, and of these,only 100 survived. Since we are inter-ested in analyzing start-ups, in our analy-sis, we only included firms less thaneight years old (Shane 2003; Shane andVenkatraman 2000). Moreover, to ensurethat sample start-ups were involved ininnovation activities, we checked theirbusiness descriptions in the source data-base, excluding start-ups with no R&Dinternal activity and able to offer non-technical services only. This provided uswith 80 start-ups.

The data to test our hypotheses comefrom a direct survey using a structuredquestionnaire. Our key informant wasthe entrepreneur, considered to be rep-resentative of the whole start-up. This isaccording to entrepreneurship literaturethat presumes that during the first stepsin the firm’s life cycle, the personal andsocial networks of the entrepreneur nor-mally coincide with the networks of thestart-up. We pretested the questionnaireas a result of a discussion with four start-ups located in the cluster. We contactedstart-ups by telephone to obtain thenames of the entrepreneurs and to besure they would agree to complete ourstructured questionnaire through face-to-face interviews. Of the 80 firms con-tacted, 54 accepted the proposal of apersonal interview to complete the ques-tionnaire, yielding a response rate of 70


percent. The data collection, started atthe end of 2004, lasted about 10 months.A comparison of differences in the meanvalues of the responding and nonre-sponding start-ups based on their three-year average sales revenues and numberof employees did not reveal any signifi-cant no-response bias. Rather thanimposing a numerical limit, we left theentrepreneurs free to quote their mostimportant customers, imposing amaximum number of 10. Overall, theanalysis was performed on 210customers—the sum of the customerslisted by our entrepreneurs—with anaverage of seven customers for eachstart-up. For every customer, using asame-structure questionnaire, we askedthe entrepreneurs to express an opinion,using Likert scales, about all itemsselected to measure both knowledgeacquisition and exploitation, cognitiveand social proximities. Moreover, weasked entrepreneurs to indicate the loca-tion of every customer, the relationshiplength, the size of the customer, and theeconomic importance of the customer forthe start-up’s performance. Finally, wecollected data about size of start-ups,their location, and age. Table 1 presentsthe descriptive statistics of our sample.On average, the start-ups in the samplewere about five years old, with annualrevenues of about €242,000 and about 15percent of total revenue coming from

foreign markets. The firms had anaverage of five employees.

MeasuresDependent Variables: Knowledge Acqui-sition (KA) and Knowledge Exploitation(KE). According to Yli-Renko, Autio,and Sapienza’s (2001) work, we mea-sured knowledge acquisition with twoitems (Table 2), previously used effec-tively by other studies such as Huber(1991), Nooteboom (2000), and VonHippel (1988). In these studies, theselected items for knowledge acquisi-tion were used to measure the techno-logical knowledge that a new venturemay acquire from its main customers.Thus, this measure of knowledge acqui-sition (technological knowledge)—aimed to discover the importance of acustomer in providing start-ups withprofitable information about technologyof process or product and R&Dactivity—has often been considered away to study the acquisition of tacitknowledge, differently from marketknowledge, which is based on codifiedknowledge (Powell et al. 2002; Yli-Renko, Autio, and Sapienza 2001). Inour study, during our pretest, theseitems proved to be appropriate forassessing knowledge acquisition by astart-up from its customers. The respon-dents were asked to answer using aseven-point Likert scale.

Table 1Descriptive Statistics

Age(Years)

Sales(in Thousands

Euro)

EmployeesMeans)

Foreign Sales(Percent onTotal Sales)

Mean 5 242 5 0.15Standard Deviation 2.14 263.5 4 0.05Range 1–8 35–900 2–8 0.01–0.18


Table 2The Dependent and the Explanatory Variables

Variables Definition Expected Effect

Ka Ke

Dependent VariablesKnowledgeAcquisition (KA)

Dummy variable (0,1) based on aseven-point Likert scale on two itemsaimed to verify the importance ofcustomers in providing local start-upswith profitable information abouttechnology of process or product andresearch and development activity

KnowledgeExploitation (KE)

Dummy variable (0,1) of log of number ofnew product developed as a result of thekey customer relationship

Explanatory VariablesCogn_Prox Seven-point Likert scale on five items

aimed to discover: development ofcommon goals, norms, and reciprocalexpectation between business partners;and overlap between business and socialaffairs among partners involved in aneconomic relationship

+ –

Social_Prox Seven-point Likert scale on six items aimedto discover lack of opportunisticbehavior among business partners,creation of common investments (degreeof commitment), and presence ofinformal relationships (log value)

+ –

Geographical_Distance Natural logarithm of the distance betweenthe firm and the client (in kilometers)

+ +

Control VariablesSize of Customersand Start-Up

Log value of total sales both of customerand start-up

= =

D_telecom/Informatics

Dummy variable for industry sector(computer industry, electronics andtelecommunications)

= =

Np_Sales The level of sales achieved by a start-upwith a specific customer in relation tototal sales of the start-up (in percentage)

= =

R&D Sales Ratio of the research and developmentexpenditure of start-ups on total sales

= =

Age_Start-Up Log age of start-up = =


We measured the knowledge exploita-tion by local start-ups for their innovationactivity by asking the start-ups to evaluatehow many new products or services theyhad developed specifically as a result ofthe relationship with the customer,according to Yli-Renko, Autio, and Sapi-enza’s (2001) work. The responsesranged from 1 to 15. We used the naturallog of this measure in the analysis tocompensate for skewness. Then, in orderto apply discrete choice modeling, wedichotomized the two dependent vari-ables using cutoff values that split thesample in a homogeneous rate using themedian value. In the case of knowledgeacquisition, the cutoff rate was the valueof 5, whereas it corresponds to 2 in thecase of log of knowledge exploitation.Therefore, the dependent variables weretransformed in dichotomous variablesthat assume a value of 0 for low values(0–5 for knowledge acquisition and 0–2for knowledge exploitation) and of 1 forhigh values. The percentage of low valueanswers is 42 percent for knowledgeacquisition and 53 percent for knowledgeexploitation.

Independent Variable: Geographical Dis-tance between Actors (Geographical_Distance). Following several studies(Boschma 2005; Kirat and Lung 1999;Rallet and Torre 2000), we measured thegeographical proximity between a localstart-up and its customer using thenatural logarithm of physical distance.We used the logarithm of distance as weexpect the relative changes in distance tobe more significant than simple absolutechange. If a very distant customer moves100 km away, the effects on the firms areminimal, whereas the same move from avery close customer can be much moresignificant.

Social Proximity (Social-Prox). As pre-viously mentioned, we measured socialproximity by the level of trust betweena local start-up and its customers

according to embeddedness literature(Granovetter 1985; Uzzi 1997). This litera-ture defines social proximity in terms ofsocially embedded relations betweenagents, where relations are socially em-bedded if they involve trust based onfriendship, kinship, and experience. Theliterature on interorganizational relation-ships provides two main definitions oftrust: confidence or predictability in one’sexpectations about another’s behaviorand confidence in another’s goodwill.The social capital theory uses and extendsboth concepts, analyzing three differentdimensions of trust between start-ups andcustomers (Nahapiet and Ghoshal 1998;Tsai and Ghoshal 1998): lack of opportu-nistic behavior; creation of commoninvestments (degree of commitment); andpresence of informal relationships. Basedon Yli-Renko, Autio, and Sapienza’s(2001) work, we selected six items tomeasure these three different dimensionsof trust using seven-point Likert scales.These items have been used in a lot ofstudies on trust, networks, and socialcapital (Leana and Van Buren 1999;Moran and Galunic 1998; Nooteboomet al. 2007).

Cognitive Proximity (Cogn_Prox). Inliterature, cognitive proximity is definedas the similarities in the way actors per-ceive, interpret, and evaluate the world(Nooteboom 2000). Therefore, a highcognitive proximity (Boschma 2005) canreduce the knowledge distance betweenbusiness partners, broadening theircommon knowledge base and expertise.According to a lot of previous studies(Nahapiet and Ghoshal 1998; Yli-Renko,Autio, and Sapienza 2001), we measuredthe cognitive proximity between busi-ness partners with five items usingseven-point Likert scales. These itemscapture two different interconnectedaspects of relational cognitive proximity.On one hand, they reflect the extent towhich the business relationship betweenstart-up and customer is characterized by


the development of common goals,norms, and reciprocal expectations con-cerning the goodwill trustworthiness ofthe exchange partner (Tsai and Ghoshal1998; Yli-Renko, Autio, and Sapienza2001). On the other hand, the selecteditems intend to verify the overlapbetween personal and business interestsamong partners, that is, the extent towhich the business relationship is char-acterized by personal social ties.

Control Variables. According to the lit-erature, an important factor influencingknowledge acquisition and exploitationprocesses is the size of the partnersinvolved in a business relationship (Tsaiand Ghoshal 1998). The analysis on sizeis very interesting as many studiessuggest that both superior resources andeconomies of scale allow larger firms toacquire and exploit external knowledgesuccessfully (Kogut and Zander 1992).However, other studies confirm that afirm’s motivation to acquire and exploitexternal knowledge tends to decreasewith size (Levinthal and March 1993). Wecontrol the effects of size variable (size)by computing the logarithmic transfor-mation of total revenues of both start-upsand customers.

Based on similar lines of reasoning(Yli-Renko, Autio, and Sapienza 2001),we included the start-up’s age as acontrol variable in our model as olderfirms may have higher capacity toacquire new knowledge and exploit itthan young firms. We calculated thefirm’s age as the number of years sinceits founding. This variable ranges fromone to eight years, with a mean of 4.7and a standard deviation of 2.1.

Moreover, as suggested by severalresearch studies, the effectiveness of theprocess of knowledge acquisition andexploitation between partners is stronglyinfluenced by the length of the businessrelationship between the partnersinvolved (Burt 1992, 2000). With theincrease in the length of the relationship

among business partners, the intensityand the strength of the business tie mayincrease, generating a growth of socialand cognitive proximity between part-ners useful to reinforce the acquisitionand exploitation of reciprocal knowledge(Nahapiet and Ghoshal 1998; Tsai andGhoshal 1998). We measured the rela-tionship length between start-up and itscustomers by the number of years ofrelationship between these two partners.

Another factor that can positivelyinfluence knowledge acquisition andexploitation by start-ups in their relation-ships with customers is the economicimportance of a customer for a start-up’sperformance (Yli-Renko, Autio, and Sapi-enza 2001). Although customers thathave a greater impact on start-up perfor-mance could become the favored part-ners in developing new products(knowledge exploitation), start-ups thatare particularly dependent on a customerfor business might be coerced intoadopting practices and techniques usefulto reinforce the acquisition and exploita-tion of knowledge (McEvily and Marcus2005). We measured the economicimportance of customers (Np_sales) byconsidering their percentage of the start-up’s sales. Based on several studies(Cohen and Levinthal 1990; Hagedoorn2002), in the analysis, we use R&Dspending as a control variable for a start-up’s willingness to invest in absorptivecapacities useful to its knowledge acqui-sition and exploitation from externalpartners. Finally, we inserted industrysector dummies for the three differentsectors (computer industry, electronics,and telecommunications).

The independent, dependent, andcontrol variables are listed in Table 2.

Reliability, Validity, andData Analysis

In order to obtain reliable and validdata, we took several precautions. First,as illustrated earlier, we pretested thesurvey with three entrepreneurs of our


sample. Second, we used multi-item mea-sures for measuring knowledge acquisi-tion and exploitation, and social andcognitive proximity. Multi-item measuresprovide considerable advantages oversingle-item measures (Churchill 1979). Ashighlighted earlier, these items on thequestionnaire were measured usingseven-point Likert scales anchored by “donot agree” and “completely agree.” As afirst step of measure validation, in orderto assess the unidimensionality of theresearch constructs (Churchill 1979),the scales were factor-analyzed by theprincipal axis method, positing a singlefactor (exploratory factor analysis). Afterexploring the factor structure of the data,we then submitted data to confirmatoryfactor analysis (CFA) that was accom-plished with the LISREL software package(Scientific Software International, Inc.,Lincolnwood, IL, USA). Table 3 summa-rizes the results of CFA on the measure-ment model. From the table, we can seethat the measurement model performedvery well because the selected constructsdemonstrate both good internal consis-tency and reliability. The standardizedfactors are all above the recommendedminimum of 0.40, and the average vari-ances extracted are all above the recom-mended minimum 0.50, ranging from0.50 to 0.84. The composite reliabilitiesare all above the recommended minimumof 0.70.

Hypotheses Test andResults

The main descriptive statistics arereported in Table 4. Focusing on inde-pendent variables, the correlation matrixreveals the potential for a multicollinear-ity problem between geographical dis-tance variable and social proximity. Inorder to check for possible problems, werely on variance inflation analysis (VIF)regressing the variable distance on theother explanatory variables. The toler-ance rate for the variable distance is 0.31,corresponding to a VIF value of 3.22. The

tolerance of the variable is well abovethe usual cutoff rate of 0.1, suggestingcollinearity is not a harmful problem.

Before testing the hypotheses, we veri-fied the statistical independence betweenknowledge acquisition and knowledgeexploitation equations that present a cor-relation coefficient (rho) equals to 0.013with a p-value = .8602. Since we didnot have an extremely large number ofobservations, we tested our hypothesesthrough binomial logit analysis using theSTATA 9 software package (StataCorp LP,College Station, TX, USA). The regressionresults for the complete models of knowl-edge acquisition and exploitation areshown in Table 5, where both control andindependent variables are inserted. Fitanalysis shows that the explanatorypower of two complete models is goodwith significant results for all the mea-sured variables. In particular, the percent-age of correctly predicted observationscorresponds to 89.47 percent for theknowledge acquisition model (72.34percent in the reduced model with theonly control variables), whereas it is equalto 80.32 percent in the knowledge exploi-tation model (68.13 percent in thereduced model with the only control vari-ables). We assure that the inclusion ofthree independent variables in both tworegression analyses adds considerableexplanatory power to the models by alikelihood ratio test with c2 values verify-ing that social, cognitive, and geographi-cal distance are jointly significant at 1percent in both two cases.

From this table, we can see that H1aand H1b are supported. In fact, the prob-ability that a local start-up acquiresknowledge from its customer tends toreduce when there is close geographicalproximity between start-up and customer(H1a). Moreover, the probability that astart-up is able to exploit the knowledgeacquired from a customer to reinforce itsinnovative activity is greater when thecustomer is not located near the start-up(H1b) since the coefficients of regression


Tab

le3

Mea

sure

men

tM

odel

Fac

tor

Nam

eM

easu

rem

ent

Item

Sta

ndar

diz

edLoad

ing

Cro

nbac

h’s

Alp

ha

Soci

al_P

RO

XThe

star

t-up

gran

tsa

def

erm

ent

of

pay

men

tto

its

cust

om

erw

ithout

pro

ble

ms

0.64

**0.

80The

firm

pro

cess

esth

ecu

stom

er’s

ord

erw

ithout

rece

ivin

ga

form

alord

er0.

65**

The

tran

sact

ion

with

the

cust

om

eris

bas

edon

info

rmal

pro

cedure

s0.

75**

*The

firm

sends

the

pro

duct

sto

the

cust

om

erbef

ore

rece

ivin

gth

een

tire

pay

men

t0.

70**

*The

tran

sact

ion

with

the

cust

om

eris

bas

edon

usu

alpro

cedure

sw

ithout

form

alag

reem

ents

0.65

**

The

cust

om

eran

dth

est

art-up

are

stro

ngl

ytied

by

com

mon

inve

stm

ents

aim

edto

rein

forc

eth

eir

grow

th0.

55*

Cogn

itiv

ePro

xim

ity

The

cust

om

eris

are

lative

of

the

entr

epre

neu

r/of

som

eone

who

work

sin

the

firm

0.88

***

0.76

The

cust

om

eroften

com

esto

the

firm

’shea

doffi

cew

ithout

reas

ons

rela

ted

tow

ork

0.70

***

The

purc

has

eord

erw

asdefi

ned

ina

nonw

ork

ing

envi

ronm

ent

0.66

**The

entr

epre

neu

rm

etth

ecu

stom

erbef

ore

star

ting

the

busi

nes

s0.

65**

The

star

t-up

has

nev

erhad

pro

ble

ms

with

the

clie

nt

since

ithas

sim

ilar

inte

rest

s0.

72**

*The

clie

nt

shar

esth

esa

me

obje

ctiv

esan

din

tere

sts

of

the

star

t-up

0.88

***

Know

ledge

Acq

uis

itio

nW

ear

eab

leto

obta

inva

luab

lein

form

atio

nab

out

rese

arch

and

dev

elopm

ent

activi

tyfr

om

this

cust

om

eruse

ful

toth

est

art-up’s

grow

th0.

88**

*

This

cust

om

erpro

vides

us

with

alo

tof

pro

fita

ble

info

rmat

ion

conce

rnin

gte

chnolo

gyof

pro

cess

or

pro

duct

use

ful

toth

est

art-up’s

grow

th0.

81**

*

Know

ledge

Exp

loitat

ion

Num

ber

of

new

pro

duct

sdev

eloped

asa

resu

ltof

the

key

cust

om

erre

lationsh

ip(l

og)

1.00

*coef

fici

ent

issi

gnifi

cant

atth

e0.

1le

vel

**co

effici

ent

issi

gnifi

cant

atth

e.0

5le

vel

***c

oef

fici

ent

issi

gnifi

cant

atth

e.0

1le

vel


Tab

le4

Des

crip

tive

Sta

tist

ics

and

Corr

elat

ion

Tab

le

Mea

nSta

ndar

dD

evia

tion

Min

Max

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Cogn

_Pro

x3.

814

2.03

71

81

Soci

al_P

rox

1.10

390.

652

01.

946

0.12

11

Geo

grap

hic

al_

Dis

tance

6.45

41.

487

2.30

27.

937

-0.2

42-0

.693

**1

Np_S

ales

0.28

310.

1498

90.

010.

7-0

.108

0.11

20.

005

1Rel

-Len

gth

5.52

31.

335

28

-0.0

290.

579*

*0.

670*

*-0

.009

1Si

ze_C

lien

t40

,965

72,7

7010

045

0.0

-0.0

63-0

.197

*0.

325*

-0.0

860.

252*

*1

Size

-Sta

rt-U

p5.

578

1.02

73.

526

6.80

20.

239*

*0.

024

-0.0

30-0

.172

*0.

024

-0.0

181

ReD

_Sal

es05

2303

420.

001

0.12

0.94

9**

0.01

0-0

.028

-0.2

41*

0.09

50.

010

0.68

9**

1A

ge-S

tart

-Up

4.17

62.

141

18

0.19

2*-0

.082

-0.0

320.

006

0.08

9-0

.082

0.68

7**

0.47

0**

1

N=

210.

*Sig

nifi

cant

at0.

1le

vel.

**Si

gnifi

cant

at0.

01le

vel.


concerning geographical distance have apositive and very significant impact.

Results concerning the positiveimpact of social proximity on knowl-edge acquisition confirm H2a. By con-trast, the coefficient of regressionrelated to knowledge exploitation ispositive and significant, contradictingour H2b: a tie based on high reciprocaltrust between start-up and customersreinforces the process of knowledgeexploitation by a local start-up for itsinnovation activity.

Finally, the data show that a strongcognitive proximity between start-up andcustomer impacts in a positive and sig-nificant way both knowledge exploita-

tion and knowledge acquisition,confirming H3a and disconfirming H3b.

DiscussionMany researchers from different disci-

plines have considered interorganiza-tional geographical proximity the keyparameter that start-ups can use toincrease their exposure to potentialknowledge spillovers (Almeida andKogut 1999; Audretsch and Feldman2003; Dahl and Pedersen 2005; Fujita,Venables, and Krugman 1999; Gittelman2001; Iammarino and McCann 2006;Kirat and Lung 1999). Although mostresearch in this stream of literaturefocuses on established firms inside an

Table 5Results of the Logistic Regression Analysis

Variables KnowledgeAcquisition

KnowledgeExploitation

D_Telecom 0.6467 (0.87) 1.018* (1.70)D_Informatics 0.9502 (1.38) 1.401* (2.16)Size_Start-Up -0.636 (-1.25) 0.101 (0.23)R&D Sales 12.398 (1.26) -2.124 (-0.11)Age Start-Up 0.542 (0.87) 0.081 (0.02)Np_Sales 2.2632 (1.33) 1.846* (1.08)Size_Client (*1.000) 0.000 (0.03) 0.021 (0.43)Rel_Length 0.0188 (-0.15) 0.307 (2.01)Cogn_Prox 0.5204** (1.96) 0.245* (1.11)Social_Prox 1.013* (1.61) 0.225* (0.11)Geographical_Distance 3.325*** (4.25) 2.345** (3.13)Const -25.125*** (-4.43) -20.265** (-5.28)

Total = 210Chi2(11) = 163.4p = .0000PsR2 = 0.5744Corr predobs. = 89.47 percent

Total = 210Chi2(11) = 114.5p = .0000PsR2 = 0.4034Corr predobs. = 80.32 percent

z-Statistic from Wald test in brackets.*p < .1**p < .05***p < .01


industrial cluster (Gilbert, McDougall,and Audretsch 2008), in this paper, weanalyzed start-ups and their interorgani-zational relationships with customers.Since start-ups are resource-constrainedand they depend upon innovatively com-bining their internal knowledge with thatof external partners (Autio, Sapienza,and Almeida 2000), we share the theo-retical idea that new ventures may bemore likely to benefit from traditionaladvantages of geographical proximitywith customers than old firms (Porter2000). However, our empirical researchseems to disconfirm these theoreticalassumptions, showing that a lot ofapproaches to industrial clusters postu-late more than they demonstrate inconsidering the importance of interorga-nizational geographical proximity for thestart-ups’ processes of knowledge acqui-sition and exploitation (Audretsch andLehmann 2006; Boschma and Frenken2006).

In fact, we verify that geographicalproximity is not a positive condition tobenefit from knowledge spillovers(Antonelli 2000, 2006; Breschi andLissoni 2001, 2002; Howells 1999; Jaffe,Trajtenberg, and Henderson 1993; Porter2000; Sorenson, Rivkin, and Fleming2006). It is an interesting result since themajority of studies have argued that start-ups localized inside a cluster, which arestrongly affected by external relation-ships, are more likely to benefit fromgeographical proximity with customersfor their knowledge acquisition andexploitation processes than establishedfirms (Gordon and McCann 2005; Hendryand Brown 2006; Lee, Lee, and Lee2003). Consequently, we empiricallyverify the theoretical assumptions ofseveral recent studies (Boschma 2005;Torre and Gilly 2000) that argued howgeographical proximity between busi-ness partners could limit their knowl-edge acquisition as actors do not alwaysneed to be located nearby to take part ina process of knowledge diffusion (Rallet

and Torre 2000). Thus, we recognize theneed to reconsider assumptions regard-ing the importance that a start-up islocated inside an industrial cluster whereit is maintained that both organizationaland institutional proximities influenceand improve the quality of a local busi-ness relationship (Porter 2000). Ifgeographical proximity facilitates inter-action, it does not constitute a sufficientpositive condition for a local start-up toreinforce its processes of knowledgeacquisition and exploitation; geographi-cal proximity does not correspond tocognitive and social proximity (Boschma2005; Breschi and Lissoni 2002; Glaeser2000).

This result is even more unexpected ifwe agree with previous studies in con-sidering our measure of knowledgeacquisition a way to study the tacitknowledge acquisition by a start-up fromits customer (Nooteboom 2000) since theadvantages of geographical proximityseem to be strongly evident in the case oftacit, complex, and noncodified forms ofknowledge (Ahuja 2000). At a generallevel, it means that local start-ups con-sider their close business partnersunlikely to possess the most profitableand valuable information concerningR&D activity and recent technologicaltrends (Ahuja 2000), differently from cus-tomers located very far from the cluster(Maskell 2001). This may be also due tothe fact that start-ups suppose that theirclose business partners have the sametechnological information as themselves,whereas exposure to many differentexternal sources of knowledge is essen-tial to innovate in new competitive inter-national environments (Alcacer andChung 2007; Neck et al. 2004; Zahra andGeorge 2002).

In contrast to geographical proximity,our empirical research shows that localstart-ups should be able to develop asignificant level of both cognitive andsocial proximity with their customers(Arita and McCann 2000; Boschma 2005)


to acquire external knowledge and tobenefit from knowledge externalitiesduring innovation activity. Although weempirically confirm the idea that themore a start-up has global-dispersed cus-tomers, the more it is able to absorbknowledge spillovers, and to reinforceits innovation activity (Bathelt, Malm-berg, and Maskell 2004; Cohen andLevinthal 1990), we also provide evi-dence of the importance of cognitive andsocial proximity between partners fortheir processes of knowledge acquisitionand exploitation. It is interesting to notethat Cohen and Levinthal (1990) do nottake account of the issue of localizationof agents in their analysis of the condi-tions of the absorption of spillovers(Cantwell and Santangelo 2002; Iam-marino and McCann 2006).

This positive influence of social andcognitive proximity is consistent withassumptions that both the acquisition andexploitation of external knowledge by alocal start-up are improved by the devel-opment of strong interactions with localand distant business partners (Hendryand Brown 2006; Tsai and Ghoshal 1998).Trust allows business partners to act in anonopportunistic way, reinforcing theircredibility and the motivation to acquireand exploit external knowledge. At thesame time, a high cognitive proximitybetween partners promotes the creationof compatible systems and cultures,reducing uncertainty and complexity,which are the main obstacles to bothknowledge acquisition and exploitationprocesses (Bathelt, Malmberg, andMaskell 2004; Fujita, Venables, andKrugman 1999; Glaeser 2000; Gordonand McCann 2005). An increase in cogni-tive and social proximity creates anopportunity for exploiting externalknowledge and not a problem. Insummary, the challenge then is to findpartners that are geographically far butcognitively and socially near as it stronglyreinforces the start-up’s knowledge acqui-sition and exploitation processes (Hage-

doorn and Duysters 2002; Nooteboomet al. 2007; Rowley, Behrens, and Krack-hardt 2000). In other words, differentlyfrom our hypotheses of research, we donot verify the traditional problems asso-ciated with high cognitive and socialproximity between business partners interms of lock-in and information redun-dancy (Burt 2000). In fact, though wesupposed in the set of hypotheses thatstrong ties—in terms of high cognitiveand social identification—are detrimentalto learning and innovation, by contrast,our results confirm the positive role of tieswith very high levels of social and cogni-tive proximity in triggering new ideas andcreativity (Uzzi 1997).

Main Implications andLimitations

The findings from this research under-score the need to reconsider traditionalassumptions regarding the importance ofgeographical proximity between busi-ness partners during knowledge manage-ment (Alcacer and Chung 2007;Audretsch and Lehmann 2006; Keebleet al. 1999). Moreover, though mostextant works in business strategy havefocused only on the importance of inter-organizational geographical proximityfor knowledge acquisition and exploita-tion processes (Almeida and Kogut 1999;Boschma 2005; Porter 2000), we extenda recent approach of economic geogra-phers that considers a mix of dimensionsof proximity influencing the ability of astart-up to acquire and exploit knowl-edge from its customers (Antonelli 2000,2006; Boschma and Frenken 2006; Dahland Pedersen 2005). By including in theanalysis the social and cognitive dimen-sions of proximity, we stress the impor-tance of studying the content ofnetworks between partners as vehicles ofknowledge acquisition and exploitation(Cooke and Morgan 1998; Gnyawali andPark 2009; Gordon and McCann 2005;McKelvey, Alm, and Riccaboni 2003;Reagans and Zuckerman 2001).


From a practical point of view, ourresults provide evidence that new entre-preneurs should be able to manage a lotof distant relationships with global cus-tomers to stimulate knowledge acquisi-tion and to reinforce knowledgeexploitation for their start-ups’ innova-tion activity. Thus, a well-developedsystem of “pipelines” connecting thestart-ups to customers all over the worldis beneficial for the start-up’s knowledgeacquisition and exploitation (Arita andMcCann 2000; Rowley, Behrens, andKrackhardt 2000). However, knowledgeflow through pipelines is not automatic,and participation is not free; therefore,the process behind the creation andmaintenance of global pipelines must bepredesigned and planned in advance,requiring specific investments to assurelong and stable relationships betweenbusiness partners (Sako 1998; Tsai andGhoshal 1998). In fact, a very highcontent of social and cognitive proximityis necessary to make effective these pro-cesses of knowledge acquisition andexploitation from global customers. Thisinvolves a complex and costly processthat requires the selection by local start-ups of distant customers according to theability to develop very high levels ofreciprocal social and cognitive proximity(Owen-Smith and Powell 2004; Tolstoy2009). It is coherent with various bodiesof innovation literature where cognitiveand social distance between businesspartners is presented as only a probleminstead of an opportunity. In terms ofpolicy implications, these results under-line that though the majority of initiativeshave been aimed at providing potentialfor interactive learning and knowledgediffusion across near actors within a spa-tially defined cluster (Macpherson andHolt 2007; Maskell and Malmberg 1999),it should be more important to supportthe development of global relationshipswith very far business partners to rein-force the knowledge diffusion andexploitation processes (Audretsch and

Feldman 2003; Sorenson, Rivkin, andFleming 2006). However, a lot of mecha-nisms able to assure high levels of socialand cognitive proximity with global cus-tomers should be assured in order tomake effective the process of knowledgeacquisition and exploitation (Tolstoy2009).

Given these strengths, however, weknow that the present study suffers alsofrom some weaknesses. First, the sampleis small and stems from a single high-tech industry, located in a geographicalarea developed over time along anendogenous path of development.Therefore, our findings may not be trans-ferable to other industries or clusters,and their applicability needs to be testedin other contexts. Moreover, since only35 of 210 analyzed customers are locatedvery near to start-ups of sample, the busi-ness model of the start-ups of the sampleseems to be most likely oriented towardcustomers geographically distant to thestart-ups. This could create some inter-pretative problems with our results.Indeed, as it is usual in survey-basedwork, in this research, we have nocontrol for survivorship bias. Moreover,though our measure of knowledge acqui-sition is based on items previously usedto measure external technological knowl-edge and then indirectly tacit knowl-edge, we did not consider the differentways that are useful in fostering differentkinds of knowledge acquisition—tacitversus codified—and exploitation,whereas the topic of dynamic learningprocesses in interorganizational relation-ships deserves closer examination.

Finally, we investigated only verticalrelationships rather than horizontal rela-tionships between similar firms (Baptista2000; Yli-Renko, Autio, and Sapienza2001). Future research could broadenthese results with an in-depth longitudi-nal analysis on the repeat of dyadic rela-tionships between start-ups and theirdifferent business partners—such as cus-tomers, suppliers, similar firms, and so


on—in order to confirm or disprove ourresults on the basis of a static approach.Finally, although the importance of infor-mal institutions has been indirectly con-sidered in our research (cognitiveproximity), future works could test ourresults inside an industrial cluster wherethe role of formal institutions has beensignificant and efficient in structuringlocal relationships and accelerating localgrowth. It could thus be verified that theimportance of geographical proximity forstart-up knowledge acquisition andexploitation is confirmed inside indus-trial clusters that are strongly supportedby local institutions.

ReferencesAhuja, G. (2000). “Collaboration Net-

works, Structural Holes, and Innova-tion: A Longitudinal Study,”Administrative Science Quarterly 45,425–455.

Albino, V., N. Carbonara, and I. Gianno-caro (2005). “Industrial Districts asComplex Adaptive Systems: Agent-Based Models of Emergent Phenom-ena,” in Industrial Clusters and Inter-Firm Networks. Eds. C. Karlsson, B.Johansson, and R. Stough. CheltenhamGlos, UK: Edward Elgar Publishing,78–82.

Alcacer, J., and W. Chung (2007). “Loca-tion Strategies and Knowledge Spill-overs,” Management Science 53(5),760–776.

Almeida, P., and B. Kogut (1999). “Local-ization of Knowledge and the Mobilityof Engineers in Regional Networks,”Management Science 45(7), 905–917.

Amesse, F., and P. Cohendet (2001).“Technology Transfer Revisited fromthe Perspective of the Knowledge-Based Economy,” Research Policy 5,100–115.

Amin, A., and F. Wilkinson (1999).“Learning, Proximity and IndustrialPerformance: An Introduction,” Cam-bridge Journal of Economics 23(2),121–125.

Antonelli, C. (2000). “Collective Knowl-edge Communication and Innovation:The Evidence of Technological Dis-tricts,” Regional Studies 34, 535–547.

——— (2006). “The Governance of Local-ized Knowledge. An Information Eco-nomics Approach to the Economics ofKnowledge,” Industry and Innova-tion 13(3), 227–261.

Arita, T., and P. McCann (2000). “Indus-trial Alliances and Firm LocationBehaviour: Some Evidence from theUS Semiconductor Industry,” AppliedEconomics 32, 1391–1403.

Audretsch, D., and M. P. Feldman (2003).“Knowledge Spillovers and the Geog-raphy of Innovation,” in Handbook ofUrban and Regional Economics. Eds.H. Vernon and J. Thisse. Amsterdam:North Holland Publishing, 80–110.

Audretsch, D., and E. Lehmann (2006).“Entrepreneurial Access and Absorp-tion of Knowledge Spillovers:Strategic Board and Managerial Com-position,” Journal of Small BusinessManagement 44, 155–166.

Autio, E., H. J. Sapienza, and J. Almeida(2000). “Effects of Age at Entry,Knowledge Intensity and Imitabilityon International Growth,” Academy ofManagement Journal 43, 909–924.

Baptista, R. (2000). “Do InnovationsDiffuse Faster within GeographicalClusters?,” International Journal ofIndustrial Organization 18(3), 515–535.

Baptista, R., and P. Swann (1998). “DoFirms in Clusters Innovate More?,”Research Policy 27(5), 525–540.

BarNir, A., and K. A. Smith (2002). “Inter-firm Alliances in the Small Business:The Role of Social Networks,” Journalof Small Business Management 40,219–232.

Bathelt, H., A. Malmberg, and P. Maskell(2004). “Clusters and Knowledge:Local Buzz, Global Pipelines and theProcess of Knowledge Creation,”Progress in Human Geography 28(1),31–56.


Beal, B., and J. Gimeno (2001). “Geo-graphical Agglomeration, KnowledgeSpillovers, and Competitive Evolu-tion,” Best Paper Proceedings of theAcademy of Management Conference,Washington, DC.

Boschma, R. A. (2005). “Proximity andInnovation: A Critical Assessment,”Regional Studies 39(1), 61–74.

Boschma, R. A., and K. Frenken (2006).“Why Is Economic Geography not anEvolutionary Science? Towards anEvolutionary Economic Geography,”Journal of Economic Geography 6(3),273–302.

Boschma, R. A., and J. G. Lambooy(1999). “Evolutionary Economics andEconomic Geography,” Journal ofEvolutionary Economics 9(4), 411–429.

Boschma, R. A., J. G. Lambooy, and V.Schutjens (2002). “Embeddedness andInnovation,” in Embedded Enterpriseand Social Capital: International Per-spectives. Eds. M. Taylor and S.Leonard. Aldershot: Ashgate, 19–35.

Breschi, S., and F. Lissoni (2001). “Loca-lised Knowledge Spillovers vs. Inno-vative Milieux: Knowledge ‘Tacitness’Reconsidered,” Papers in RegionalScience 80, 255–273.

——— (2002). “Mobility and Social Net-works: Localised Knowledge Spill-overs Revisited,” paper presented atClusters in High-Technology: Aero-space, Biotechnology and SoftwareCompared, Montreal, Canada, Novem-ber 2002.

Burt, R. S. (1992). Structural Holes. Cam-bridge: Harvard University Press.

——— (2000). “The Network Structure ofSocial Capital,” in Research in Orga-nizational Behavior, 22nd ed. Eds. R.I. Sutton and B. W. Staw. Greenwich:JAI Press, 345–423.

Canina, L., C. A. Enz, and J. S. Harrison(2005). “Agglomeration Effects andStrategic Orientations: Evidence fromthe US Lodging Industry,” Academy ofManagement Journal 48, 565–581.

Cantwell, J. A., and G. D. Santangelo(2002). “The New Geography of Cor-porate Research in Information andCommunications Technology (ICT),”Journal of Evolutionary Economics12, 163–197.

Carlsson, B. (2003). “Innovation Systems:A Survey of the Literature from aSchumpeterian Perspective,” in TheCompanion to Neo-SchumpeterianEconomics. Ed. A. Pyka. Cheltenham:Edward Elgar, 160–183.

Churchill, G. A. Jr. (1979). “A Paradigmfor Developing Better Measures ofMarketing Constructs,” Journal ofMarketing Research 16(1), 64–73.

Cohen, W. M., and D. A. Levinthal(1990). “Absorptive Capacity: A NewPerspective on Learning and Innova-tion,” Administrative Science Quar-terly 35, 128–152.

Cooke, P., and K. Morgan (1998). TheAssociational Economy: Firms,Regions and Innovation. Oxford:Oxford University Press.

Cooke, P., and D. Willis (1999). “SmallFirms, Social Capital and the Enhance-ment of Business Performancethrough Innovation Programmes,”Small Business Economics 13, 219–234.

Dahl, M. S., and C. R. Pedersen (2005).“Social Networks in the R&D Process:The Case of the Wireless Communica-tion Industry around Aalborg,Denmark,” Journal of Engineeringand Technology Management 22,75–92.

Davenport, S. (2005). “Exploring theRole of Proximity in SME KnowledgeAcquisition,” Research Policy 34, 683–701.

Dyer, J. H., and H. Singh (1998). “TheRelational View: Cooperative Strategyand Sources of InterorganizationalCompetitive Advantage,” Academy ofManagement Review 23, 660–679.

Eisenhardt, K. M., and C. Schoonhoven(1996). “Resource-Based View of Stra-tegic Alliance Formation: Strategic and


Social Effects in EntrepreneurialFirms,” Organization Science 7, 136–150.

Folta, T. B., A. C. Cooper, and Y. Baik(2006). “Geographic Cluster Size andFirm Performance,” Journal of Busi-ness Venturing 21, 101–123.

Freel, M. S. (2003). “Sectoral Patterns ofSmall Firm Innovation, Networkingand Proximity,” Research Policy 32,751–770.

Fujita, M., A. Venables, and P. Krugman(1999). The Spatial Economy: Cities,Regions and International Trade.Cambridge: MIT Press.

Gargiulo, M., and M. Benassi (1999).“The Dark Side of Social Capital,” inCorporate Social Capital and Liability.Eds. R. Leenders and R. Gabbay.Boston, MA: Kluwer Academic Pub-lisher, 100–120.

Gilbert, B. A., P. P. McDougall, and D. B.Audretsch (2008). “Clusters, Knowl-edge Spillovers and New Venture Per-formance: An Empirical Examination,”Journal of Business Venturing 23,405–422.

Gittelman, M. (2001). Mapping NationalKnowledge Networks: Scientists, Firmsand Institutions in Biotechnology inthe United States and France. NewYork: NYU Stern School of Business.

Glaeser, E. L. (2000). “The New Econom-ics of Urban and Regional Growth,” inThe Oxford Handbook of EconomicGeography. Eds. G. L. Clark, M. P.Feldman, and M. S. Gertler. Oxford:Oxford University Press, 83–98.

Gnyawali, D. R., and B. J. R. Park (2009).“Co-Opetition and TechnologicalInnovation in Small and Medium-Sized Enterprises: A Multilevel Con-ceptual Model,” Journal of SmallBusiness Management 47, 308–330.

Gordon, I., and P. McCann (2005). “Inno-vation, Agglomeration and RegionalDevelopment,” Journal of EconomicGeography 5(5), 523–543.

Granovetter, M. S. (1985). “EconomicAction and Social Structure: The

Problem of Embeddedness,” Ameri-can Journal of Sociology 91, 81–150.

Grant, R. M. (1996). “Toward aKnowledge-Based Theory of theFirm,” Strategic Management Journal17, 109–122.

Hagedoorn, J. (2002). “Inter-Firm R&DPartnerships: An Overview of MajorTrends and Patterns since 1960,”Research Policy 31(4), 477–492.

Hagedoorn, J., and G. Duysters (2002).“Learning in Dynamic Inter-Firm Net-works: the Efficacy of Multiple Con-tacts,” Organization Studies 23(4),525–548.

Hansen, M. T. (1999). “The Search–Transfer Problem: The Role of WeakTies in Sharing Knowledge acrossOrganizations Subunits,” Administra-tive Science Quarterly 44, 82–111.

Hendry, C., and J. E. Brown (2006).“Organizational Networking in UKBiotechnology Clusters,” BritishJournal of Management 17(1), 55–73.

Hite, J. M., and W. S. Hesterly (2001).“The Evolution of Firm Networks:From Emergence to Early Growth ofthe Firm,” Strategic ManagementJournal 22, 275–286.

Howells, J. (1999). “Regional Systems ofInnovation?” in Innovation Policy in aGlobal Economy. Ed. D. Archibugi.Cambridge: Cambridge UniversityPress, 67–94.

Huber, G. P. (1991). “OrganizationalLearning: The Contributing Processesand the Literatures,” OrganizationScience 2(1), 88–115.

Hudson, R. (1999). “The LearningEconomy, the Learning Firm and theLearning Region. A Sympathetic Cri-tique of the Limits to Learning,” Euro-pean Urban and Regional Studies 6,59–72.

Iammarino, S., and P. McCann (2006).“The Structure and Evolution ofIndustrial Clusters: Transactions,Technology and Knowledge Spill-overs,” Research Policy 35(7), 1018–1036.


Jaffe, A., M. Trajtenberg, and R. Hender-son (1993). “Geographic Localizationof Knowledge Spillovers as Evidencedby Patent Citations,” QuarterlyJournal of Economics 108, 577–598.

Kaiser, U. (2002). “Measuring Knowl-edge Spillovers in Manufacturing andServices: An Empirical Assessment ofAlternative Approaches,” ResearchPolicy 31(1), 125–144.

Kaplan, R. S., and D. P. Norton (2004).“Measuring the Strategic Readiness ofIntangible Assets,” Harvard BusinessReview 82(2), 52–63.

Keeble, D., C. Lawson, B. Moore, and F.Wilkinson (1999). “Collective LearningProcesses, Networking and Institu-tional Thickness in the CambridgeRegion,” Regional Studies 33, 319–332.

Kirat, T., and Y. Lung (1999). “Innova-tion and Proximity. Territories as Lociof Collective Learning Processes,”European Urban and Regional Studies6, 27–38.

Kogut, B., and U. Zander (1992). “Knowl-edge of the Firm. Combinative Capa-bilities and the Replication ofTechnology,” Organization Science 3,383–397.

Lambooy, J. G., and R. A. Boschma(2001). “Evolutionary Economics andRegional Policy,” Annals of RegionalScience 35, 113–131.

Lane, P. J., and M. Lubatkin (1998).“Relative Absorptive Capacity andInterorganizational Learning,” Strate-gic Management Journal 19(5), 461–477.

Leana, C. R., and H. J. Van Buren (1999).“Organizational Social Capital andEmployment Practices,” Academy ofManagement Review 24, 538–555.

Lee, J., J. Lee, and H. Lee (2003). “Explo-ration and Exploitation in the Pres-ence of Network Externalities,”Management Science 49(4), 553–570.

Lesser, E. L. (2000). Knowledge andSocial Capital. Foundations andApplications. Boston, MA: Butter-worth Heinemann.

Levinthal, D. A., and J. C. March (1993).“The Myopia of Learning,” StrategicManagement Journal 14(Winter),95–112.

Liao, J., and H. Welsch (2005). “Roles ofSocial Capital in Venture Creation:Key Dimensions and Research Impli-cations,” Journal of Small BusinessManagement 43, 345–362.

Lundvall, B. (1988). “Innovation as anInteractive Process: From User–Producer Interaction to the NationalSystem of Innovation,” in TechnicalChange and Economic Theory. Eds. C.Dosi, C. Freeman, G. Silverberg, andL. Soete. London: Frances Pinter, 349–362.

Macpherson, A., and R. Holt (2007).“Knowledge, Learning and Small FirmGrowth: A Systematic Review of theEvidence,” Research Policy 36(2),172–192.

Majocchi, A., and M. Presutti (2009).“Industrial Clusters, EntrepreneurialCulture and the Social Environment:The Effects on FDI Distribution,”International Business Review 8(1),76–88.

Malmberg, A., and D. Power (2005).“(How) Do (Firms in) Clusters CreateKnowledge?″ Industry and Innova-tion 12, 409–431.

Marshall, A. (1920). Principles of Eco-nomics. London: Macmillan.

Maskell, P. (2001). “Towards aKnowledge-Based Theory of the Geo-graphical Cluster,” Industrial andCorporate Change 10, 921–943.

Maskell, P., and A. Malmberg (1999).“The Competitiveness of Firms andRegions and the Importance of Local-ized Learning,” European Urban andRegional Studies 6, 9–25.

Maurer, I., and M. Ebers (2006). “Dynam-ics of Social Capital and Their Perfor-mance Implications: Lessons fromBiotechnology Start-Ups,” Administra-tive Science Quarterly 51(2), 262–292.

McDougall, P., S. Shane, and B. M. Oviatt(1994). “Explaining the Formation of


International New Ventures: TheLimits of Theories from InternationalBusiness Research,” Journal of Busi-ness Venturing 9, 469–487.

McEvily, B., and A. Marcus (2005).“Embedded Ties and the Acquisitionof Competitive Capabilities,” StrategicManagement Journal 26, 1033–1055.

McEvily, B., and A. Zaheer (1999).“Bridging Ties: A Source of Firm Het-erogeneity in Competitive Capabili-ties,” Strategic Management Journal20, 1133–1156.

McKelvey, M., H. Alm, and A. Riccaboni(2003). “Does Colocation Matter forFormal Knowledge Collaboration inthe Swedish Biotechnology Pharma-ceutical Sector?,” Research Policy32(3), 483–501.

Moran, P., and C. Galunic (1998). “Har-nessing Social Capital for ProductiveResource Exchange,” Academy ofManagement Conference, San Diego,CA.

Morrison, G. (2004). “Gatekeepers ofKnowledge within Industrial Clusters:Who They Are, How They Interact,”poster session presented at Schum-peter Society Conference, Milan, Italy,June 2004,.

Nahapiet, J., and S. Ghoshal (1998).“Social Capital, Intellectual Capitaland the Organizational Advantage,”Academy of Management Review 23,242–266.

Neck, H. M., G. D. Meyer, B. Cohen, andA. Corett (2004). “An EntrepreneurialSystem View of New Ventures Cre-ation,” Journal of Small BusinessManagement 42, 190–208.

Nonaka, I., R. Toyama, and A. Nagata(2000). “A Firm as a Knowledge-Creating Entity: A New Perspectiveon the Theory of the Firm,” Indus-trial and Corporate Change 9, 1–20.

Nooteboom, B. (2000). “Learning byInteraction: Absorptive Capacity, Cog-nitive Distance and Governance,”

Journal of Management and Gover-nance 4(1–2), 69–92.

Nooteboom, B., W. Van Haverbeke, G.Duysters, V. Gilsing, and A. Vanden Oord (2007). “Optimal Cogni-tive Distance and AbsorptiveCapacity,” Research Policy 36, 1016–1034.

Oinas, P., and E. Malecki (2002). “TheEvolution of Technologies in Timeand Space: From National andRegional to Spatial InnovationSystems,” International RegionalScience Review 25(1), 102–131.

Owen-Smith, J., and W. W. Powell(2004). “Knowledge Networks asChannels and Conduits: The Effects ofSpillovers in the Boston Biotechnol-ogy Community,” OrganizationScience 15, 5–21.

Pirolo, L., and M. Presutti (2010). “TheImpact of Social Capital on the Start-up’s Performance Growth,” Journal ofSmall Business Management 48, 197–227.

Porter, M. E. (1990). The CompetitiveAdvantage of Nations. Worchester:Billing and Sons Ltd.

Porter, M. (2000). The ComparativeAdvantage of Nations. New York:New York Press.

Powell, W., K. Koput, L. Smith-Doerr,and J. Bowie (2002). “The SpatialClustering of Science and Capital,”Regional Studies 36, 291–305.

Presutti, M., and C. Boari (2007). “Space-Related Antecedents of Social Capital:Some Empirical Inquiries about theCreation of New Firms,” InternationalEntrepreneurship and ManagementJournal 4, 1–18.

Presutti, M., C. Boari, and L. Fratocchi(2007). “Knowledge Acquisition andthe Foreign Development of High-Tech Start-Ups: A Social CapitalApproach,” International BusinessReview 16(1), 23–46.

Rallet, A., and A. Torre (2000). “Is Geo-graphical Proximity Necessary in theInnovation Networks in the Era of


Global Economy?,” GeoJournal 49,373–380.

Reagans, R., and E. Zuckerman (2001).“Networks, Diversity, and Productiv-ity: The Social Capital of CorporateR&D Teams,” Organization Science12(4), 502–517.

Romanelli, E., and O. M. Khessina(2005). “Regional Industrial Identity:Cluster Configurations and EconomicDevelopment,” Organization Science16, 344–358.

Rosenkopf, L., and P. Almeida (2001).Overcoming Local Search throughAlliances and Mobility. University ofPennsylvania and Georgetown Uni-versity: Mimeo.

Rowley, T., D. Behrens, and D. Krack-hardt (2000). “Redundant GovernanceStructures: An Analysis of Structuraland Relational Embeddedness in theSteel and Semiconductor Industries,”Strategic Management Journal 21,369–386.

Sako, M. (1998). “Does Trust ImproveBusiness Performance?” in Trustwithin and between Organizations.Conceptual Issues and EmpiricalAnalysis. Eds. C. Lane and R. Bach-mann. Oxford: Oxford UniversityPress, 88–117.

Saxenian, A. (1994). Regional Advan-tage: Culture and Competition inSilicon Valley and Route 128. Cam-bridge, MA: Harvard University Press.

Shane, S. (2003). A General Theory ofEntrepreneurship: The Individual–Opportunity Nexus. Aldershot:Edward Elgar.

Shane, S., and S. Venkatraman (2000).“The Promise of Entrepreneurship asa Field of Research,” Academy ofManagement Review 25(1), 217–226.

Smircich, L. (1983). “Concepts of Cultureand Organizational Analysis,” Admin-istrative Science Quarterly 28(3), 339–358.

Smith, B. R., C. H. Matthews, and M. T.Schenkel (2009). “Differences inEntrepreneurial Opportunities: The

Role of Tacitness and Codification inOpportunity Identification,” Journalof Small Business Management 47,38–57.

Sorenson, O., and P. G. Audia (2000).“The Social Structure of Entrepreneur-ial Activity: Geographic Concentrationof Footwear Production in the USA,1940–1989,” American Journal ofSociology 106, 424–426.

Sorenson, O., J. W. Rivkin, and L.Fleming (2006). “Complexity, Net-works and Knowledge Flow,”Research Policy 35(7), 994–1017.

Spender, J. C. (1996). “Making Knowl-edge the Basis of a Dynamic Theory ofthe Firm,” Strategic ManagementJournal 17, 5–9.

Stinchombe, A. L. (1965). “Social Struc-ture in Organizations,” in Handbookof Organizations. Ed. J. G. March.Chicago, IL: Rand McNally, 142–193.

Stuart, T., and O. Sorenson (2003). “TheGeography of Opportunity: SpatialHeterogeneity in Founding Rates andthe Performance of BiotechnologyFirms,” Research Policy 32, 229–253.

Szulanski, G. (1996). “Exploring InternalStickiness: Impediments to the Trans-fer of Best Practice within the Firm,”Strategic Management Journal 17,27–43.

Tolstoy, D. (2009). “Knowledge Combi-nation and Knowledge Creation in aForeign-Market Network,” Journal ofSmall Business Management 47, 202–220.

Torre, A., and J. P. Gilly (2000). “Debatesand Surveys: On the AnalyticalDimension of Proximity Dynamics,”Regional Studies 34(2), 169–180.

Tsai, W., and S. Ghoshal (1998). “SocialCapital and Value Creation: The Roleof Intra-Firm Networks,” Academy ofManagement Journal 41, 464–476.

Uzzi, B. (1997). “Social Structure andCompetition in Interfirm Networks:The Paradox of Embeddedness,”Administrative Science Quarterly 42,35–67.


Von Hippel, E. (1988). The Sources ofInnovation. New York: Oxford Uni-versity Press.

Weick, K. E. (1979). The Social Psychol-ogy of Organizing. New York: Edi-tions McGraw-Hill, Inc.

West III, G. P., and T. W. Noel (2009).“The Impact of Knowledge Resourceson New Ventures Performance,”Journal of Small Business Manage-ment 47, 1–22.

Williamson, O. E. (1975). Markets andHierarchies. New York: Free Press.

Yli-Renko, H., E. Autio, and H. J. Sapienza(2001). “Social Capital, KnowledgeAcquisition, and Knowledge Exploita-tion in Young Technology-Based

Firms,” Strategic Management Journal22, 587–613.

Zaheer, S., B. McEvely, and V. Perrone(1998). “Does Trust Matter?,” Organi-zation Science 9, 141–159.

Zahra, S. A., and G. George (2002).“Absorptive Capacity: A Review,Reconceptualization and Extension,”Academy of Management Review27(2), 213–240.

Zahra, S. A., R. D. Ireland, and M. A. Hitt(2000). “International Expansion byNew Venture Firms: InternationalDiversity Mode of Market Entry, Tech-nological Learning and Performance,”Academy of Management Journal43(5), 925–950.


The Importance of Proximity for the Start-Ups' Knowledge Acquisition and Exploitation

Documents