The themes of technological innovation, entrepreneurship, and organizing

Firm-Specific Factors and the Degree of Innovation Openness

Valentina Lazzarotti

Carlo Cattaneo University, Italy

Raffaella Manzini

Carlo Cattaneo University, Italy

Luisa Pellegrini

University of Pisa, Italy

ABSTRACT

This chapter investigates the topic of how open innovation is actually implemented by companies, accord­ing to a conceptual approach in which open and closed models ofinnovation represent the two extremes of a continuum of different openness degrees; though, these are not the only two possible models. By means of a survey conducted among Italian manufacturing companies, this chapter sheds light on the many different ways in which companies open their innovation processes. Four main models emerge from the empirical study, which are investigated in depth in order to understand the relationship between a set of firm-specific factors (such as size, R&D intensity, sector of activity, company organization) and the specific open innovation model adopted by a company.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

INTRODUCTION

The concept of “Open Innovation” (OI) is often studied supposing an artificial dichotomy between closed and open approaches, whilst the idea of exploring different degrees and types of openness in a sort of continuum seems to provide a more interesting avenue (Chesbrough, 2003b). Prior research has highlighted that open innovation may

DOI: 10.4018/978-1-61350-165-8.ch010 be pursued in different ways, which are identifi­able in terms of organisational form of acquisition or commercialization, number and typologies of partners, phases of the innovation process that are actually open, the direction of opennes (inbound and/or outbound) and governance (hierarchical or flat).

Moreover, previous research has also attempted to study the relationships among different OI models and several contextual factors, driven by the idea that these factors could explain or,

at least, characterize the companies’ choices in terms of degree of openness. Lastly, different OI models, defined according to this concept (i. e. degree of openness and models within their specific context), have been analysed in some preliminary work in terms of their performance (Lichtenthaler, 2008; 2009).

The objective of this chapter is thus three­fold: first, to provide evidence in support of Chesbrough’s (2003b) theoretical proposition that businesses may be located along an Open Innovation Continuum, second, through the use of extensive study, to identify any potential inter­mediate states between the extreme points of the Continuum - Open and Closed Innovators - and, third, to identify the contextual factors that affect the choices firms make along the Open Innova­tion Continuum.

In particular, for the identification of the po­tential intermediate states in the OI Continuum, we focalized on two variables representing the openness degree, which are not still deeply in­vestigated: (1) the number and type of partners (partners variety) and (2) the number and type of phases of the innovation process open to external contributions in and/or out (innovation phase variety). It should be noted that we assume that the innovation process is composed of different phases: idea generation (identification of a tech­nology opportunity through scouting, monitoring, market analysis, trends analysis); experimentation (from the idea to the prototype); engineering (transforming the prototype into an industrial project); manufacturing (defining and organis­ing the “plant”); commercialisation (planning of commercialisation and promotional activities).

The choices in terms of OI will be investigated in terms of those contextual factors whose role is still controversial (Lichtenthaler, 2008), or oth­erwise it can be better understood in light of the concept of openness suggested here. Our inves­tigation was carried out in Italy, where empirical evidence about OI is still poor. However, there are many pressures, arising from institutions, too, towards the establishment of collaborative models (Global Business Summit, 2010). Thus, investi­gating if, how and with what results companies work together becomes a relevant issue for both Italian scholars and practitioners.

We would also like to emphasize that our en­deavour to identify any in-between states along the Open Innovation Continuum is the first at­tempt to research this topic and that the subject indeed requires further research in order to better characterise such intermediate states.

The following sections are divided into sub­topics: a description of the pertinent literature (so as to better understand the research questions we posed), a description of the empirical study we carried out and the methodology used, the main research results, a discussion of the results, con­clusions and future research.

THEORETICAL BACKGROUND The Theoretical Framework and the Research Questions

Traditionally, large firms relied on internal research and development (R&D) to innovate and, in many industries, large internal R&D labs were a strategic asset and firms could internally discover, develop and commercialize technologies. This approach has been labelled the “closed innovation model” (Chesbrough, 2003a). Although it worked well for quite some time, the current innovation landscape has changed. Due to labour mobility, increasing R&D costs, abundant venture capital and widely dispersed knowledge across multiple public and private organizations and the need for specialisa­tion in knowledge production, enterprises can no longer afford to innovate on their own, but rather need to engage in alternative innovation practices. In this regard, Open Innovation (OI) represents an important innovation practice that can help firms to innovate without having to rely only on their in­house strengths. Since Chesbrough published his book in 2003, the concept of “Open Innovation” has received a considerable amount of attention from practitioners and researchers. A large num­ber of studies are adopting this term to describe the phenomenon where firms rely increasingly on external sources of innovation, which means that ideas, resources and individuals flow in and out of organizations (Chesbrough, 2003a). While contributions are still growing (Gassmann, 2006; Enkel et al., 2009; Enkel et al., 2010), the debate on innovation management is enriched by stud­ies that critically examine the Open Innovation concept by exposing its weakness and limitations (Dahlander & Gann, 2007; Trott and Hartmann, 2009). In particular, the concept of Open Innova­tion is criticized because of the widespread view that the concept highlights an artificial dichotomy between closed and open approaches. On the other hand, the idea of exploring different degrees and types of openness in a sort of continuum (i. e. the openness degree concept) seems to provide a more interesting and richer avenue to investigate, (Chesbrough, 2003b; Dahlander & Gann, 2007). Indeed, this view allows for a deeper and more real investigation into company behaviour and into the particular nature and context ofinnovation sources (Chesbrough, 2003b; Gassmann, 2006; Dahlander & Gann, 2007).

In any case, the era of open innovation has begun and many firms are opening their innovation process to the outside world (Enkel et al., 2009). The way the innovation process can be opened has been studied in management literature from a variety of perspectives. Although the perspec­tive that has received most of the attention in the literature is undoubtedly the direction of openness, other approaches have also been investigated. More specifically, these look at the number and types of partners, the kind of governance in the innovation networks, and the organisational forms chosen to define the links among partners (high vs. low integration level).

As regards the perspective connected with the “direction of openness”, three models of open innovation can be observed: the inbound, exploration or outside-in process, the outbound, exploitation or inside-out process, and the coupled process (Keupp & Gassman, 2009; Lichtentaler, 2008; Enkel, et al., 2009). Thanks to the outside - in process, firms aim at enriching the company’s own knowledge base through the integration of external knowledge sourcing, and hence increase their innovativeness (Enkel et al., 2009). Through the inside-out process, firms aim at earning prof­its by bringing ideas to market, selling IP, and multiplying technology by transferring ideas to the outside environment, in order to bring ideas to market faster than they could through internal development (Enkel et al., 2009). The coupled process combines the two abovementioned pro­cesses to simultaneously gain external knowledge and bring ideas to market.

As regards the perspective connected with the “types of partners” (Enkel et al., 2009), literature has highlighted the interactive character of the innovation process, suggesting that innovators use ideas and knowledge of external actors in their innovation processes: firms rely heavily on their interaction with lead users, suppliers, and a range of institutions inside the innovation system (von Hippel, 1988; Lundvall, 1992; Brown & Eisenhardt, 1995; Szulanski, 1996). With each innovation source, an organization can achieve different intensity levels of collaboration (Laursen & Salter 2006; Keupp & Gassman 2009). Hence, it is possible to define different open innova­tion models depending on both breadth (i. e., the number of sources used for innovation activities) and depth (i. e., the intensity of collaboration with each source).

As regards the perspective connected with the “kind of governance” in the innovation net­works, there are two dimensions which need to be considered (Pisano & Verganti, 2008): open­ness, i. e. a large number of involved partners and hierarchy, i. e. the level of ‘democracy’ in decision making. On the basis of two such aspects, four open innovation models emerge: (1) the open/ hierarchical model, in which anyone can offer ideas but only one company defines the problem and chooses the solution; (2) the open/flat model, in which anyone can generate ideas, and no one has the authority to decide what is or is not a valid innovation; (3) the closed/hierarchical model, in which a company selects certain participants and decides which ideas are to be developed; (4) the closed/flat model, in which a selected group offers ideas, while making critical decisions together.

As regards the “organisational forms” chosen to define the links among partners (high vs. low integration level), there are four technology sourc­ing modes that firms can adopt: corporate Venture Capital investments, non-equity alliances, equity alliances and acquisitions. Each form carries with it different implications in terms of the investing company’s reversibility and commitment (Chiesa & Manzini, 1998; van de Vrande et al., 2006). More precisely, corporate Venture Capital invest­ments and non-equity alliances are reversible to some extent and involve a relatively low level of commitment from the investing company, while equity alliances and acquisitions require a high level of commitment and are hardly reversible.

In our opinion, all these contributions share two aspects: on one hand, they have a common interpretation of open innovation, while, on the other, they have a weakness.

Regarding common interpretation, all these contributions share the understanding that the open innovation models which the firms follow are not exclusively open or closed, but rather show varying degrees of openness: i. e. between the two pure models - open or closed, which represent the two extremes of a continuum - there are many shades of grey (Chesbrough, 2003b). Indeed, according to Dahlaner and Gann (2007), the dichotomy between open vs. closed is artificial and it is necessary to explore different degrees and types of openness: this can yield more insight in understanding openness.

With regards to weakness, the perspectives used in the previous contributions are not exhaus­tive in explaining the open innovation models followed by firms. In other words, the latest lit­erature still does not fully explain in what ways the degree of openness can happen. Indeed, to the best of our knowledge, the literature does not address the question whether some firms conduct open innovation in many phases oftheir innovation funnel and if others focus only on a very few of them; if this is the case, we must consider which phases ofthe innovation funnel are open or closed.

Hence, literature does not help companies to find the right balance between closed and open phases of the innovation funnel. Neither is it clear if the phases of the innovation funnel that are permeable are open to many or just a few partners. With few exceptions, it is not even clear if the involved partners are different in terms of typologies or not. For instance, De Backer et al.

(2008) analyzed such a problem and found that universities and government research institutes are generally considered to be an important source of knowledge transfer for the innovation activities of companies, especially regarding more upstream/ research activities.

On the basis ofthese premises, our objective is to contribute to the literature which sustains that business reality is not based on pure open innova­tion, but on companies that invest simultaneously in closed as well as in open innovation activities (Enkel et al., 2009) throughout the innovation funnel with different partners. Hence, we will introduce a new perspective that considers both the number/typology of partners and the number/ typology of phases, in order to understand if such a perspective can confirm the existence of different models of open innovation. Within this context we will try to answer the following specific research questions:

• Do different firm-specific factors charac­terize the models of open innovation?

• Do such different models show a different level of innovative performance?

Figure 1 depicts the constructs of our theo­retical framework. The operationalisation of each construct is reported in detail in the Appendix (all questions were measured on a four-point Likert scale to indicate the frequency of use, with 1 = disagreement and 4 = agreement).

Figure 1. Theoretical framework

As explained in the introduction, the main objective of the chapter is to provide empirical evidence to the notion of OI as a continuum, that is to say that different OI models may exist. Before characterizing them by means of our empirical analysis, in the following we will analyze what the literature says about the relationships between some contextual factors (i. e. R&D intensity, size, type of industry, approach to innovation, com­pany’s objectives for collaboration, managerial - organizational actions supporting open innova­tion) and OI models and their performance, by highlighting areas that are lacking which justify our subsequent empirical analysis. First, an analysis of the relationships between the firm - specific factors and the open innovation models will be made. Then, an analysis of the impact of the open innovation models on innovative per­formance will follow. Specifically, what is lacking in the literature has been highlighted for each of the relationships studied.

The Firm-Specific Factors and Open innovation

Relationship between R&D Intensity and Open Innovation Models

As regards the role played by R&D intensity, Lichtenthaler (2008), Lichtenthater and Ernst

(2009) , Calantone and Stanko (2007) and Sofka and Grimpe (2008) investigated this role from different viewpoints. Lichtenthaler (2008) and Lichtenthater and Ernst (2009) analysed the effect of R&D intensity and found that the greater the level of R&D intensity the greater the techno­logical exploration. This provides support for the assumption that firms pursue external technology acquisition as a complement to internal R&D and not as a substitute (Cohen & Levinthal, 1990, Zahra & George, 2002).

Calantone and Stanko (2007) underpin that firms’ exploration activities cannot occur frequent­ly: therefore, given the high costs for developing specialized structures, firms are more likely to resort to outside expertise. Moreover, they state that firms performing a great deal of in-house exploratory research are likely to be led by this exploration away from their competencies, and will therefore be more likely to seek out outside expertise.

Sofka and Grimpe (2008) studied the effect of internal R&D investments on “breadth” (i. e. extent of partner typology) and “depth” (i. e. intensity of collaboration for each partner) of research strategies. They hypothesize that internal R&D investments lead to deep research strategies rather than broad ones. With their survey, which involved firms from twelve European countries, they argue that firms building absorptive capacities through internal R&D have both broader and deeper re­search strategies. However, the effect on depth is stronger than the effect on breadth. In other words, committing internal resources to in-house labs and specialized scientists and engineers is therefore the primary path for innovation managers to achieve more depth in their search strategies.

Hence, on the basis of these contributions, the role played by R&D intensity is studied in rela­tion to two of the abovementioned perspectives, through which it is possible to investigate how the innovation process can be opened: the perspec­tive connected with the “direction of openness” and the perspective connected with the “types of partners”. Thus, literature lacks the investigation of the role that R&D intensity plays with regards to the perspective offered in this chapter, which will consider both the variety ofpartners involved and the variety of stages in which companies collaborate.

Relationship between Size and Open Innovation Models

Size is one of the most investigated of the contex­tual factors and it is still a controversial subject.

On the one hand, empirical literature suggests that open innovation is mainly driven by larger companies. Empirical investigations show that size impacts on two variables representing the openness degree: the extent of both technology exploitation and exploration (Lichtenthaler, 2008; Lichtenthaler & Ernst, 2009). Indeed, as regards technology exploitation, larger companies seem to have a bigger technology portfolio than smaller companies and hence have wider technological knowledge that is potentially suitable for com- mercialization. As regards technology exploration, larger firms do not seem to be able to completely rely on internal activities due to the diversity ofthe technological knowledge that they use. In general, the fact that larger firms seem to drive the open­ing of the innovation process can be justified by the more systematic approach they have in their innovation processes (Lichtentaler, 2008) and the larger resources they possess with respect to small and medium enterprises (Lichtentaler, 2008; De Backer et al., 2008; van de Vrande et al., 2008). In addition, according to Lichtentaler (2008), it should be noted that the effect of size seems to be stronger in the case oftechnology commercializa­tion than in technology exploration, in that com­mercialization is rather a newer phenomena than acquisition. As the external mode of technology exploitation has become a broader trend only in recent years, it is still driven by large pioneering firms, while the acquisition of external technology is distributed more evenly across firms of different sizes. Keupp and Gassman (2009) analyse the ef­fect of size on two different variables representing the openness degree: the number of knowledge sources used for OI activities (i. e. the breath of OI) and the intensity of collaboration with each source (i. e. the depth of OI) and show that there is a positive and significant effect of firm size on both the breadth and the depth of OI.

On the other hand, some literature emphasises that especially small companies, often lacking resources and competence to innovate by them­selves, would have great benefits from exploit­ing the OI model. In fact, SME are increasingly adopting OI practices (van deVrande et al., 2008).

Hence, on the basis of these contributions, it is possible to draw considerations similar to those regarding the role played by R&D intensity. Indeed, the role of size is studied in relation to the same two abovementioned perspectives: the perspective connected with the direction of open­ness and the perspective connected with the types of partners. In addition, its role is still controver­sial. As a consequence, it is possible to assert that literature lacks the investigation of the role that size plays with regard to the perspective offered in this chapter (number/typology of partners and number/typology of phases).

Relationship between Type of Industry and Open Innovation Models

Some authors investigate the impact that industry exerts on OI, interpreting industry as the typology of sector in which firms operate. Also in this case, literature is not unidirectional in that empirical findings show contrasting results. What’s more, the same Lichtenthaler in different publications (Lichtenthaler, 2008 and Lichtenthaler & Ernst,

2009) , while referring to the same sample, finds different results.

In fact, on the one hand, Lichtenthaler and Ernst

(2009) show that a firm belonging to a particular industry does not produce any impact either on the external technology acquisition or on external technology exploitation. Similarly, Lichtenthaler

(2008) states that his findings demonstrate the insignificant effect which industry differences have across the clusters. Thus, the openness of the innovation process does not seem to be determined principally by industry characteristics.

On the other hand, the studies by Gassman and Enkel (2004) and De Backer et al. (2008) demonstrate the opposite. More in particular, Gassman and Enkel (2004) state that the relative importance of internal and external sources var­ies across different industries. De Backer et al.

(2008) , although focusing on particular aspects, such as patent licensing, find important differences among industries, with chemical/drugs, electronic/ electrical/semiconductors and machinery/equip­ment/computers as the industries where licensing deals take place more frequently than in others.

Relationship between Approach to Innovation and OI Models

A relevant concept investigated in the literature is that of “technology aggressiveness” (measured by three items, among them “the emphasis on radical innovation rather than incremental innovation”)1. Lichtenthaler and Ernst (2009) find that technol­ogy aggressiveness is negatively related to the extent of external technology acquisition and is positively related to external technology exploita­tion, in that commercialization nurtures benefits in terms of setting industry standards, entering into new markets, and realizing learning effects.

In another publication, Lichtenthaler (2008) studies the implications connected with firms’ emphasis on radical innovation and finds that the degree of openness seems to rise with the degree of emphasis on radical innovation, especially concerning the degree of external technology commercialization. There are two reasons for this: first, the opportunity to commercialize knowledge which, when not applied in the organization, turns out to be residual; second, the possibility to facili­tate acceptance on the market and the creation of a standard. Lichtenthaler (2008) also finds that firms which emphasize radical innovation are obviously not able to develop all knowledge internally, but they have to strongly rely on complementary external sources and thus they use technology acquisition (Perrons et al., 2005).

Hence we can draw even more restrictive con­siderations than those regarding R&D intensity: technological aggressiveness is studied in the literature in relation only with the perspective connected with the direction of openness. If we add that this factor’s role is still controversial, it emerges that new empirical investigation is needed to analyze the impact exerted by technological aggressiveness on OI models.

Company Collaboration Objectives and OI Models

The main reasons that push firms towards choices of open innovation are, on the one hand, the need to reduce innovation costs and business risks, and on the other, the need to extend skills, competences and creativity (Huang et al., 2009).

As regards the objective of diminishing costs and risks, Calantone and Stanko (2007) analyze outsourcing as a tool for increasing staffing ef­ficiency measured in terms of employee sales efficiency. They infer that the decision to reduce the number of employees is related to the outsourc­ing of innovation in the short run but not over the long term. Gassmann and Enkel (2004) state that research-driven companies usually aim at reducing the R&D’s fixed costs and sharing risk. Chiaroni, Chiesa and Frattini (2009) state that the reason for accessing external sources is the willingness to minimize risk by investing in technologies that are already proven in other applications.

Another main reason for firms to undertake R&D outsourcing includes accessing specialized skill sets and creativity, which exposes the internal development staff to new knowledge, technol­ogy, and organizational development processes (Catalone & Stanko, 2007; Chesbrough and Teece, 1996; Linder, 2004; Lynch, 2004), even if this strategy has drawbacks in terms of opening the market to new entrants (Porter, 1980) and exposing core competencies to imitation and substitution (Piachaud, 2005).

In comparison with the other firm-specific variables, the objectives of collaboration are stud­ied in the literature even more restrictively; not only are they studied in relation to the perspective connected with the direction of openness, but also mainly in relation to one ofthe two directions, i. e. with the inbound process. Hence, for this firm - specific variable, too, there is a gap in the research literature which needs to be filled, that being an analysis of the impact exerted by collaboration objectives on the open innovation models.

Managerial-Organizational Actions Supporting Open Innovation and OI Models

Managerial-organisational actions allow open innovation to be pursued easily and more deeply, Some of these actions include the commitment of top management to promote the transition towards an open innovation approach (Vanhaver - beke, 2006; vandeMeer, 2007; Chiaroni et al., 2009, Pisano & Verganti, 2008); the need for a champion supporting the integration of external technology into an existing product develop­ment phase-gate process (Chesbrough, 2006; Chesbrough & Crowther, 2006); the exploitation of the personal relationship of the R&D manag­ers for starting technological collaborations; the formal evaluation of collaboration objectives and risks, as well as the analysis and selection of the potential partners with a formal and explicit process (Sakkab, 2002; Huston & Sakkab, 2006). Although the works cited have shed light on how organizational and managerial factors facilitate the implementation of open models, we believe that enriching this line of inquiry with new empirical evidence is in any case quite important.

The Impact on Performance

The debate is still open on whether and how openness degree and contextual factors impact on innovative and economic performance. The results are still quite limited and contradictory, although very recent contributions (Chiang & Hung, 2010; Sofka & Grimpe, 2010) shed more light on the topic.

A widely accepted assumption is that the rela­tions between openness degree and performance must be analyzed considering the moderator role of external environmental moderators (e. g. patent protection status: Lichtentaler, 2009; Slowinsky & Zerby, 2008; MacCormack & Iansiti, 2009). Indeed, regarding performance, it should be noted that the analysis ofthe company’s financial performance is a complex topic due to the fact it can be explained only by considering a wide set of factors that can have contrasting effects.

Probably, the concept of innovative perfor­mance (Chiang & Hung, 2010; Sofka & Grimpe,

2010) is more understandable. The impact of open innovation models on innovative performance has been analysed in terms ofa company’s competence base, development costs, time to market and the level of innovation of new products/processes. Literature is unidirectional in showing the impact of the outside-in process on the access and inte­gration of internal company capabilities with new and complementary knowledge of external firms (Gassmann and Enkel, 2004). Instead, literature re sults are not unidirectional as far as the reduction of development time is considered: for instance, on one hand, Gassmann and Enkel (2004) state that the benefits of co-operation are seen in an improvement in the competitive position and in risk minimisation, but not in a reduction of devel­opment time; on the other hand, according to Kolk and Puumann (2008) firms not concentrating on Open Innovation strategies fail, as rising develop­ment costs and shorter product life cycles make it increasingly difficult to justify investments in innovation. Other studies (e. g. Dahlaner & Gann,

2007) show that relationships with other actors help firms to increase the level of innovativeness.

In summary, as suggested by the literature listed above, certain relationships between the selected firm-specific variables and the openness degree are still controversial or lacking in depth. Below, we suggest improving the empirical evi­dence available by adopting a perspective based on number/typology of partners and the number/ typology of phases, with particular reference to Italy, where partnerships are desired by many subjects, including institutional ones, though the issue is still poorly studied.

RESEARCH DESIGN AND METHODOLOGY Survey Design

The empirical study has focused on companies located in Lombardy, a Northern Italian region; in 2008 the companies had applied for funding from the Chamber of Commerce to conduct in­novative activities within different manufacturing sectors, including the mechanical and machinery sectors, as well as in sectors dealing with auto­motive, metallurgy, textiles, food, electronics, chemicals, pharmaceuticals, plastic, rubber, paper and paperboard, publishing and printing, wood and wood products (NACE rev.2 codes). This engagement in innovation by such companies, combined with the fact that Lombardy is marked by a particular propensity for innovation (if mea­sured by the number of patents, Lombardy ranks first among the Italian regions according to the European Patent Office data for Italy elaborated by the Unioncamere Observatory of Patents and Brands, 2008) make them very interesting topics of innovation study. The data was collected by means of questionnaires distributed by email to participants. The advantages of such a method include low cost, completion at the respondent’s convenience, absence of time constraints, guar­antee of anonymity and reduction of interviewer bias (Forza, 2002). Its shortcomings, on the other hand, are represented by lower response rate as compared to other methods, longer comple­tion times and greater effects due to the lack of both interviewer involvement and open-ended questions. The survey tool was conducted as a questionnaire whose items regarded company characteristics (sector, size), innovation strategy; organization for innovation; collaborations and innovative performance, as will be clarified in more detail below.

Before sending the questionnaire to the com­panies, a pilot test was conducted to assess the quality of the measure items. The items were tested by a group of senior managers and academ­ics with working experience in innovation. They were asked to analyze the questionnaire in order to eliminate items not having strong content validity. After this stage, the resulting questionnaire was sent to the key informants of the companies that we identified as the R&D manager (if present in the company) or the company owner, if deeply involved in the definition of the company’s in­novation strategy (as is very common in Italian companies).

Statistical Analysis

Among the companies that have applied for fund­ing (about 500) 99 firms have responded during a four-months period in 2009 (i. e. with a response rate equal to 20%). A general premise should be made as concerns company size (in terms of number of employees and revenues): except for few big subjects, the size ofthe studied companies can be classified as middle/small2. This imbal­ance can hardly be avoided because it is due to the intrinsic major sectoral composition in Lom­bardia, where the small size plays an important role. If, from the one hand, it is also found in non-respondents and thus it protects against the potential non-response bias, on the other hand it may prevent capture size differences when we will analyze firms characterized by different openness degrees. We must therefore bear in mind that this cannot make next comparisons between companies significant because of the intrinsic nature of the sample. However, this is the typi­cal situation in Lombardia as well as in Italy. As clarified above in the theoretical background, we adopt the partner variety and the phase variety as relevant variables to represent and to investigate the degree of openness. Regarding their operation­alization, we used subjective measures based on four-point Likert-type scales (1=strongly disagree; 4 = strongly agree) as given in the Appendix. In order to better specify the partner variety and the phase variety we introduced the variables: intensity of collaboration with partner and intensity of col­laboration on phases that make up the innovation funnel (measures also based on the Likert-type scale). The study of correlations between these two variables allowed us to understand which are the most typical combination partner/phase and thus to characterize the innovation process in practice. To study firms’ approaches to open innovation, we firstly carried out a cluster analy­sis (“complete linkage method”, recommended when researcher wants to identify groups which are distinct from each other as much as possible; Barbaranelli, 2006) based on the partner variety and the phase variety. Secondly, concerning the firm-specific variables, with which we intended to describe the companies belonging to different clusters, we carried out the following procedure. Items of the questionnaire were defined on the basis of scales already used in previous works or coming from partial reworking of such scales (still Likert-type). Anyway, we applied to the gathered data an exploratory factor analysis (principal axis factoring as extraction method and promax rotation in the case of initially unclear solution) in order to delete weakly related items and to understand the factor structure and the measurement quality. An evaluation of the Eigenvalues and the Scree plot were used to identify the number of factors to retain. In addition, all factor loadings were above the acceptance level of 0.50 (Hair et al., 2006; Barbaranelli, 2006; Cheng and Shiu, 2008), thus indicating the unidimensionality of the various factors. These were saved as variables and em­ployed in the subsequent analysis. The factors/ firm-specific variables were the following (see the Appendix for detail):

Objectives of collaborations classified in two factors:

1. aims to extend skills, competences and cre­ativity (three items, inspired by the work of Huang et al., 2009; Cronbach’s a: 0,71);

2. aims to share risks and costs (two items, based on Calantone and Stanko, 2007; Cronbach’s a: 0,84).

Approach to innovation: technological aggres­siveness with emphasis on radical innovation (five items, inspired by Lichtenthaler and Ernst, 2009, that use suggestion by Brockoffand Pearson, 1992. We re-adapted the scale also on SURVEY TOOL 2.1 basis, a questionnaire sponsored by Industrial Research Institute, Cronbach’s a: 0,71)

Organizational and managerial actions for open innovation (five items, scale based on SURVEY TOOL 2.1, Cronbach’s a: 0,85)

Some other variables, not presented in Ap­pendix, were measured directly (and eventually transformed in logarithmic scale to improve nor­mality), such as:

• R&D intensity (i. e. percentage of R&D expenses/sales)

• Revenues (i. e. to operazionalize size)

• Number of employees (i. e. to operazional - ize size)

• Indicators of company’s performance (ROA - Return On Assets - and ROS - Return On Sales).

As concerns company’s results, also a factor representing innovative performance was defined (five items reported in the Appendix, our scale based on Calantone et al. (2002); Cronbach’s a: 0,82). After all, dummy variables were included for: the type of industry; the existence of orga­nizational unit specifically devoted to support collaboration; the type of organizational structure used by companies for innovation activities (input - oriented, output-oriented, matrix; Chiesa, 2001).

Finally, as concerns data analysis, we applied the one-way variance analysis (i. e. ANOVA), in order to appreciate differences among clusters in terms of scale variables, and Chi-square test to compare the frequency on nominal variables.

RESULTS

Figure 2 illustrates the results of cluster analysis based on the partner variety and the phase variety.

This has resulted in a solution with four groups of firms. The decision on the number of clusters has been determined by the criterion that suggests of stopping the aggregation process at the stage that precedes the one with the highest increase in the coefficient of agglomeration (Barbaranelli, 2006). In the four-cluster solution, the variance inside clusters is about 21% whereas the variance among clusters is about 72% (F-tests sig. <.001). Despite the quite high correlation between partner variety and phase variety variables, which sug­gests that most firms adopt open or closed innova­tion approach on both dimensions, small inter­mediate clusters - 2 and 3 - exist (i. e. they open their innovation process more strongly in one direction rather than in the other) and thus they are worth to be analyzed although they only pro­vide clues for further analysis.

Cluster 1 refers to the open innovators, com­panies that make up the largest group. From data about partner and phase variety variables (reported in Table 1), we found that these companies are really able to manage a wide set of technological relationships, that impact on the whole innova­tion funnel and involve a wide set of different partners. Although the open innovators strongly collaborate especially with the supplier in the engineering and experimentation phases3, many other types of partners (particularly, firms oper­ating in different sectors of activity, customers, universities, technical and scientific service com­panies, governmental institutions) are involved at different stages (especially in the idea generation,

experimentation and engineering). An example is a medium-sized company from the machinery industry that conceptualizes and produces boilers for industrial use. Cluster 4 refers to the closed innovators, companies that form the second largest group. From data in Table 1, we realized that these companies access to external sources of knowledge only for a specific, single phase of the innovation funnel and typically in dyadic collaborations. The prevalent partners are in fact suppliers and cus­tomers, especially on the idea generation phase. Particularly little-used in the closed companies are the collaborations with universities and firms operating in different sectors of activity. A good example here is a small-sized textile company that has declared in some follow-up interviews, carried out after data-analysis stage, to follow a traditional innovation approach (i. e. internal research and development procedures “jealously preserved”) by using the low-intensity contribu­tion of its customers and suppliers on the idea generation phase. The companies in the smallest cluster 2 can be named integrated collaborators. Already found in the evidence emerged by a mul­tiple case-study in a previous work (Lazzarotti & Manzini, 2009), these companies are the most similar to the closed ones: collaborations are with few types of partners (typically suppliers and customers), but instead of being tightly focused on one stage, they can be extended to the whole funnel. This means that the integrated collabora­tors “share” with their few trusted partners the whole process of innovation. An example is a small company in the electronics industry that produces and commercializes panels and electrical equipment: suppliers and customers, with whom the company has a longstanding relationship, support it through the whole process of innova­tion. Finally, companies in the other small cluster 4 can be classified as specialized collaborators. Already also emerged in our previous work, they form a group similar to open innovators regard­ing the variety of partners (suppliers, customers, universities, governmental institutions) but they concentrate their collaborations in a single/few points of the innovation funnel (typically the idea generation and the experimentation). From the follow-up interviews, these companies still seem a “bit behind the open”, but it is a matter of time: with increasing confidence in partners, the cooperation will also increase by covering all the innovation process. An example here is a medium­sized company in the electronics industry that open to many partners (universities and governmental institutions as well as the traditional customers and suppliers) the idea generation phase.

To analyze differences across clusters, in par­ticular the open and closed ones, one-way variance analysis (i. e. ANOVA) has been used for compar­ing means of scale variables (i. e. company’s size expressed by revenues and number of employees, R&D intensity, approach to innovation, types of objectives, organizational and managerial actions for open innovation, performance expressed by innovative performance) and Chi-square test has been employed to compare the frequency on nominal variables (i. e. type of industry, existence of organizational units supporting collaborations). It must be said that the scarcity of observations on clusters integrated and specialized makes not applicable Chi-square tests as well as the results regarding scale variables are often not significant. Anyway, although the following evidence is use­ful above all to compare open and closed compa­nies, some clues on other two clusters also emerge and thus, if interesting, they will be briefly pre­sented in order to deepen them with next research.

Regarding R&D intensity, we found significant difference among open and closed clusters (see Table 1): the open companies invest more on average than the closed companies. This finding provides support for the assumption that firms consider open innovation as complementary with internal R&D and not a substitute (Lichtenthaler, 2008). This is consistent with the theory ofabsorp - tive capacity (Cohen & Levinthal, 1990, Zahra & George, 2002) in the sense that to be able to absorb from the outside, a company must have the appropriate skills and competences. This does not mean that the closed invests little, just it seems they invest less than open. As clues relatively to integrated and specialized clusters, interestingly the integrated shows the lowest average of R&D intensity whereas the specialist is more similar to the open one. Perhaps the integrated innovator invests internally less because it relies on a few trusted collaborators along the whole innovation process. In addition, in similar vein with previous absorptive capacity interpretation, Chi-square tests on internal organizational structure for in­novation activities show that the input-oriented one (i. e. where people are organized according to their specific area of expertise, whose growth is continuously fed - Chiesa 2001) is typical for open innovators rather than for closed (for which an informal type of organization is prevalent). Indeed, the competence-building receives also a formally structured attention and thus it could suggest competence is considered a pre-requisite to openness.

As concerns size (see Table 1), we did not find significant differences among clusters. However, we reiterate that this may be due to the inherent imbalance of the investigated sample (i. e. high weight of medium-small sizes for companie s) that reflects the Lombardia’s and Italy’s condition. But what perhaps we can say is that Italian companies, despite being small, are nevertheless brought to open up to outside sources, in keeping the stream of literature that argues that small firms behave like this (van de Vrande et al., 2008).

Groups do not seem different even for type of industry (see Table 2). As suggested by the follow-up interviews, the degree of open innova­tion seems to be mainly determined by the indi­vidual strategic choice of a company rather than by industry characteristics (for similar evidence, see Lichtenthaler, 2008).

As concerns approach to innovation, with emphasis on radical innovation rather than incre­mental innovation, we found that the open cluster has a higher mean (in the factor score resulting by factor analysis) than the closed and this dif­ference is statistically significant. This is consis­tent with the literature and empirical evidence that suggest that companies, when focalized on radical innovations, must collaborate because they are not able to internally develop all relevant knowledge (Lichtenthaler, 2008; Perrons et al.,

2005) . “Relevant knowledge” that our data seem to suggest is coming from a higher degree of openness in term of wide partner variety and wide phase variety. Interestingly, the specialized, al­though very little, shows a mean higher than in­tegrated, which might suggest that perhaps the partner variety is more relevant.

Regarding the type of objectives pushing com­panies to collaborate, open cluster shows higher mean (and statistical different) in the first-type goal “aim to extend skills, competences, creativity” with respect to closed companies. Very similar to each other (and in intermediate position between open and closed), the integrated and the specialized cluster. This finding suggests that companies look for competences and creativity by opening up in some way: to a wide variety of partners (even if on few phases), to a wide variety of phases (even if with few partners) or, at the highest level, in

Variables

Sample

(n=99)

Cluster 1

(n=43)

Open

Cluster 2 (n=9) Integrated collaborator

Cluster 3 (n=11)

Specialized

collaborator

Cluster 4 (n=36) Closed

Significance (Anova test)

Partner and phase vari­ables

Partner variety

2.63

3.44

1.89

3.18

1.67

.000

Phase variety

2.61

3.49

3.11

1.82

1.67

.000

Intensity of collabora­tion with University and Research centres

1.38

1.59

1.16

1.36

1.18

.005

Intensity of collaboration with Technical and Scien­tific Service Companies

1.37

1.56

1.29

1.05

1.25

.02

Intensity of collabora­tion with Governmental institutions

1.11

1.20

1.00

1.18

1.01

.03

Intensity of collaboration with customers

1.70

1.77

1.56

1.80

1.61

.63

Intensity of collaboration with suppliers

1.93

2.12

1.80

1.87

1.74

.17

Intensity of collaboration with competitors

1.09

1.08

1.18

1.29

1.03

.01

Intensity of collaboration with firms operating in dif­ferent sectors of activity

1.40

1.61

1.29

1.33

1.19

.04

Intensity of collaboration on Idea generation

1.51

1.62

1.38

1.52

1.39

.11

Intensity of collaboration on Experimentation

1.56

1.72

1.35

1.69

1.36

.000

Intensity of collaboration on Engineering

1.44

1.61

1.40

.1.32

1.29

.002

Intensity of collaboration on Manufacturing set up

1.34

1.49

1.13

1.26

1.22

.002

Intensity of collaboration on Commercialization

1.28

1.35

1.35

1.25

1.18

.23

Firm-specific contextual variables/factors

Revenues (Log)

7.35

7.5

7.5

7.35

6.9

0.53

Employees (Log)

1.52

1.63

1.58

1.66

1.32

0.41

R&D intensity (Log)

0.59

0.86

0.20

0.59

0.35

.01

Innovation approach

-0.04

0.33

-0.07

0.30

-0.47

.000

Objective of extending skill and competence

-0.20

0.45

-0.13

-0.11

-0.47

.000

Objective of sharing risks and costs

0.02

0.28

-0.29

-0.13

-0.22

.05

Organizational and mana­gerial actions for OI

0.01

0.37

0.26

0.34

-0.61

.000

continued on following page

Table 1. Continued

Variables

Sample

(n=99)

Cluster 1 (n=43) Open

Cluster 2 (n=9) Integrated collaborator

Cluster 3

(n=11)

Specialized

collaborator

Cluster 4 (n=36) Closed

Significance (Anova test)

Performance

Innovative performance

-0.04

0.30

0.08

0.29

-0.48

.001

ROS (Log)

0.73

0.76

0.80

0.65

0.70

0.89

ROA (Log)

0.78

0.78

0.69

0.75

0.81

0.94

Table 2. Information on open and closed and main differences (dummy variables)

Variables

Sample

(n=99)

Cluster 1 (n=43) Open

Cluster 4 (n=36) Closed

Significance (Chi-square test)

Industry

Mechanic/machinery

41%

46%

34%

0.29

Metallurgy

14%

12%

18%

0.38

Textile

8%

7%

9%

0.7

Food

4%

5%

3%

0.7

Electronics

7%

7%

6%

0.9

Chemical/pharmaceuticals

10%

14%

6%

0.42

Organizational context

Organization input-oriented

44%

53%

33%

.05

Existence of organizational unit supporting collaboration

35%

47%

22%

.02

both directions. Similarly, the second-type goal of sharing risks and costs is related to the degree of openness in our conception. It is also interesting to note the prevalent objective in each cluster. Whereas in open cluster the main goal is the first, closed companies are pushed to open by the objective of sharing costs and risks.

As concerns the organizational and managerial actions, open cluster shows an average intensity on these tools statistically higher than closed (inte­grated and specialised still similar each other and in intermediate position between open and closed). As suggested by literature (Pisano & Verganti,

2008) these type of actions are necessary to ensure successful collaborations and this is confirmed by our evidence. Integrated and specialised have got a lower degree of openness (the integrated lower than the specialized) and probably a lower complexity in the collaborations. Thus, it is not necessary to introduce high-intensive managerial and organizational actions. Also a significant Chi - square test on the existence of an organizational unit supporting collaboration (see Table 2) gives evidence of the organizational and managerial differences between open and closed approaches.

Regarding performance (see Table 1) we obtained only some preliminary indications and mainly focus on innovative performance. Indeed, we believe that the analysis of overall company’s performance is a complex topic due to the fact it can be “explained” only by considering a wide set of factors that can have opposite impacts. With this premise, we studied the differences between clusters only with an explorative purpose to define
next steps of research and in terms of innovative performance (i. e. factor that is a combination of five items). We found that open cluster seems more performing than closed (and better than the sample average). Moreover, by studying correla­tions between innovative performance and partner variety, we found a high and significant relation. Particularly strong were the relations between the single item “The company’s competence base was enlarged” and partner variety and “the level of innovativeness of new products/processes was improved” and still the partner variety, suggest­ing that the open is more innovative and that the innovativeness seems to be linked to the partner variety. Another clue for this type of interpretation is given by the specialist innovative performance: higher than integrated just in these two items. Anyway, it is important to keep in mind that they are only clues, not confirmed by the analysis of company’s overall performance (measured by means of ROS and ROA), that is even greater in the closed than in the open cluster. Thus, further investigation on performance is surely required.

DISCUSSION

In this chapter, different models for open innova­tion are studied, by means of a survey conducted among 99 Italian manufacturing companies, with respect to two variables: the partner variety and the innovation phase variety. Although these two variables are highly correlated (.71; p<.001), inter­mediate cases (i. e. companies for which the two variables are different in their value, low or high) were found among companies. As a result, four different models for open innovation were found in the practice of companies: open innovators, specialised collaborators, integrated collaborators, closed innovators. The two extreme models - open and closed - are far more diffused (in coherence with the correlation found between the two variables), so the intermediate ones need a more dedicated analysis to confirm what is emerging here, that is difficult to generalise because of the limited number of companies included in these clusters. Open and closed innovators actually emerge from this analysis as two significantly different open innovation models, especially in terms of:

• Approach to innovation: open innovators are those who choose an aggressive tech­nology and innovation strategy, in which they work to be technological leaders, to come first to the market with new products, to lead the technology evolution with su­perior know-how, to pursue even radical innovations. In other words, it can be ar­gued that opening the innovation process to a wide variety of partners and all along the innovation funnel is conceived as part of an aggressive strategy;

• R&D intensity: open innovators invest more in R&D than closed, and this in some way confirms the difference in the strategy for the two clusters; aggressive innovators spend significantly more in R&D and, as part of their effort, they spend for opening up their innovation process. Another inter­esting explanation of this result refers to the need to invest in internal competences in order to be open, as the absorptive ca­pacity of the company is critical to iden­tify and exploit potential collaborations and exchanges with external partners. In perfect coherence with this result, there is evidence that an input-oriented organi­sational structure for the R&D activities, which maximises the absorptive capacity, is typical for open innovators rather than for closed;

• Type of objectives: in coherence with the two results above, open innovators, with respect to closed ones, mainly open their innovation process to achieve benefits re­lated to internal competences, i. e. to deep­en and integrate the knowledge base, to in­crease creativity and flexibility, to achieve excellence in knowledge production. On the other side, closed innovators are rather focused on reducing the costs and risks of innovation, by sharing them with external partners;

• Organisational and managerial actions implemented to support openness: open innovators have actually modified their organisational structure and management techniques, by introducing roles, routines and tools especially dedicated to the de­sign, development and implementation of collaborations with external partners;

• The results concerning the two “inter­mediate” models of open innovation, i. e. specialised collaborators and integrated collaborators, are less robust, because of the limited number of companies found for these two clusters. As a consequence, only some tentative interpretation of the achieved results can be put forward. However, it is relevant to reflect on such results since they can represent the start­ing point for a future research aimed at verifying whether these two open innova­tion models can actually represent a valid alternative to open and closed models and, if so, in which specific context condi­tions. By making a synthesis of all results achieved for specialised and integrated collaborators, it seems that there is a sort of “continuum” in the openness of compa­nies, in terms of the most relevant context conditions emerged in this study, as shown in Figure 3.

Figure 3 clearly shows that specialised and integrated collaborators can be really considered as “intermediate” models: the most significant variables that characterise the open and closed models, in fact, have values that are between the two extremes4. Integrated and specialised col­laborators are thus viable options for companies that don’t have a highly aggressive approach to innovation and that don’t want to invest too much for opening up the innovation process. As a consequence, these companies have limited expectations in terms of benefits deriving from open innovation, but do not want to completely abandon the opportunity to access to external sources of knowledge and competencies. As an example, let’s take a specialised collaborator, the electronic company cited above: the general manager wanted to spend a limited amount of resources on studying opportunities for opening the innovation process, but, at the same time, R&D managers clearly felt the need to integrate their knowledge with external contributions com­ing from other industries, universities, excellent research centres. As a consequence, they decided to open only the idea generation phase to a wide variety of partners.

Even in terms of performance the four models are different and a first tentative conclusion in this sense is that the degree of openness is posi­tively correlated to the innovative performance: from closed innovators to open ones the level of innovative performance increases (in terms of new products and services, time to market, level of novelty, learning, costs for new products). But this does not seem to have an effect on the com­pany’s economic performance in the short term, as already discussed above. This result is in con­trast with other studies (Lichtenthaler2009), which found a positive correlation between the degree of openness (measured in terms of intensity of outbound licensing) and the economic perfor­mance (measured through ROS and ROI). In our opinion, the relation between performance and open innovation is very complex to be studied: performance is intrinsically a multi-dimensional concept (think, just to quote a well known frame­work, to the balanced scorecard concept), as well as the degree of openness. This can probably lead to many different measures for evaluating the relationship between openness and performance that certainly requires further in depth studies.

1=0.20 C=0.35 S=0.59 O=0.S6

C=-0.47

R&D intensity I=-0.07

--------------------------- ►

S=0.30 0=0.33

Approach to innovation

C=-0.47

I=-0.13 S=-0.11

0=0.45

Objective of extending skills and competences

C--0.61

1=0.26

S=0.34 0=0.37

Organisational and managerial actions for OI

Figure 3. Specialised and integrated collaborators between the two extreme models (C=closed innova­tors; I=integrated collaborators; S= specialised collaborators; O=open innovators)

CONCLUSION AND FUTURE RESEARCH DIRECTIONS

Our area of research is related and contributes to innovation because collaborations and networks (in particular, technological collaboration), and the proper ways to manage them, are today largely recognized as means to improve, or at least to sup­port, firms’ innovation capabilities (Chesbrough,

2006) . The study is conducted by means of a survey involving 99 Italian companies operating in manufacturing industries. Different models for open innovation are found in the practice of companies: open innovators, specialized collabo­rators, integrated collaborators, closed innovators. The two extreme models - open and closed - are far more diffused and actually emerge as two significantly different open innovation models, in terms of approach to innovation, R&D intensity, type of objectives, organizational and manage­rial actions implemented to support openness. The two intermediate models - specialized and the integrated collaborators - although in need of further empirical investigation, provide evidence in support of Chesbrough’s (2003b) theoretical proposition that businesses may be located along an Open Innovation Continuum.

In conclusion, the chapter introduces a new perspective that integrates both the number/ty­pology of partners and the number/typology of phases, in order to understand if such perspec­tive can confirm the existence of different open innovation models. Moreover, it provides useful managerial implications because it suggests that OI is not an “on/off’ choice, but it can be interpreted and adopted with different degrees (Chesbrough, 2003b), consistently with the company’s specific context. Thus, intermediate open innovation mod­els (i. e. integrated and specialized collaborators) are viable options for companies that do not have a highly aggressive approach to innovation and that do not want to invest too much for opening up the innovation process. As a consequence, these companies have limited expectations in terms of benefits deriving form open innovation, but do not want to completely abandon the opportunity to access to external sources of knowledge and competencies. We suggest that intermediate mod­els for opening the innovation process can be a first relevant topic for future research; performance of open innovation can be a second one and, fi­nally, a third one may concern the study of open innovation models in a dynamic perspective, i. e. analyzing the path followed by companie s to open their innovation process (Chiaroni et al., 2009).

Adopting a dynamic perspective, the different models found in this study may be interpreted as different steps in a long term path towards open innovation: starting from a closed innovation process, companies may gradually open to a very limited set of well known partners (suppliers and customers) in a integrated collaboration model, or may decide to open only a single phase of the innovation process to a wide variety of partners with a specialized collaborator approach. Some of the cases studied in a previous work (Lazzarotti & Manzini, 2009) seem to confirm this hypothesis. A future research with longitudinal case studies can probably improve the understanding of this dynamic path to open innovation.

Finally, it should be noted that the number of respondents is still very limited. Moreover, it is studied only the relationship between some firm-specific factors and the degree of openness (defined specifically in terms of partner variety and phase variety): a wider investigation is recom - mendable to include more contextual factors, i. e. external/environmental ones, or more variables that can help to define the openness degree.

The themes of technological innovation, entrepreneurship, and organizing

About the Contributors

Farley S. Nobre (PhD, MSc, BSc) is Professor at the School of Management of Federal University of Parana, Brazil. His research interests include organizations, knowledge management systems, innova­tion and sustainability. …

The Roles of Cognitive Machines in Customer — Centric Organizations: Towards Innovations in Computational Organizational Management Networks

Farley Simon Nobre Federal University of Parana, Brazil ABSTRACT This chapter proposes innovative features of future industrial organizations in order to provide them with the capabilities to manage high levels …

Tools That Drive Innovation: The Role of Information Systems in Innovative Organizations

Jason G. Caudill Carson-Newman College, USA ABSTRACT The purpose of this chapter is to examine computer technology as a tool to support innovation and innovative processes. The primary problem that …

Как с нами связаться:

Украина:
г.Александрия
тел./факс +38 05235  77193 Бухгалтерия
+38 050 512 11 94 — гл. инженер-менеджер (продажи всего оборудования)

+38 050 457 13 30 — Рашид - продажи новинок
e-mail: msd@msd.com.ua
Схема проезда к производственному офису:
Схема проезда к МСД

Партнеры МСД

Контакты для заказов шлакоблочного оборудования:

+38 096 992 9559 Инна (вайбер, вацап, телеграм)
Эл. почта: inna@msd.com.ua