In many ways innovation is in tension with strategy. Innovation is serendipitous and uncertain and technological change is seldom, if at all, under the control of a single organisation. The flexibility to adapt to environmental contingencies is an indispensable competence, and strategies can slow down this adaptation, especially in rapidly changing environments. It is no wonder that Quinn (1985) described innovation management as ‘controlled chaos’ and Mintzberg observed that strategies impede the use of ‘peripheral vision’ (Mintzberg, 1987). In fact, it is precisely this peripheral vision that is indispensable for innovation and where a network approach is most valuable: in recognising opportunities, increasing variety, accessing resources and timeliness in the face of uncertainty and sponsoring the diffusion of innovations in dynamic environments.

Mintzberg and Waters’ (1985) incrementalist approach suggests a continuum of strategies ranging from the most deliberate to the most emergent. Deliberate strategies are those in which the distinction between formulation and implementation stages is most clear-cut: there is a detailed plan, followed by pre-programmed implementation. Deliberate strategies leave little space for coping with uncertainties and are better suited to stable environments. As strategies move along the spectrum from most deliberate to most emergent, the distinction between formulation and implementation phases becomes fuzzy. At the far end of the spectrum strategies are realised ex post, without prior plans.

‘Why do people put solar on their roofs? Because other people put solar on their roofs’, read the title of an article in the Washington Post on 25 October 2014. This logic of influence and imitation applies to a diverse range of behaviours, attitudes, beliefs, opinions and strategies – as there are numerous contexts in which actors are influenced by the behaviour of others they admire or perceive to be similar to themselves. The findings about solar panels were based on research confirming this influence, controlling for other factors, in Connecticut and California (Bollinger and Gillingham, 2012; Graziano and Gillingham, 2015).

How can the network perspective help us understand the process of diffusion and learn how to direct it? Valente (2012: 49) defines network interventions as ‘purposeful efforts to use social networks or social network data to generate social influence, accelerate behaviour or change, improve performance, and/or achieve desirable outcomes among individuals, communities, organisations, or populations’. The network perspective can help us detect potentially influential actors to promote diffusion, understand barriers to diffusion and discover which networks are better for diffusion in a given context.

Diffusion Studies: From Past to Present

Diffusion studies have a long history in the social sciences. Three major studies have been influential in the growth of this field of research. The first (Ryan and Gross, 1943) points to the time lag in innovation diffusion.

SNA refers to a range of techniques in graph theory, used to analyse network structure. Over the last fifty years, methods for analysing social networks have developed rapidly. There are many valuable sources on social network analysis and a range of (mostly free) network analysis software.

This chapter looks at SNA techniques and three ways in which they are used in innovation studies. The first uses node-based SNA metrics to analyse the positions of individual actors (firms, individuals, teams and so on) in a network, and how these metrics are related to the nodes’ characteristics, behaviour or performance. The second is detecting subgroups, or clusters, with certain characteristics within a network, for example, groups in which people have dense networks within technology domains that are closely related to each other or important patents in the evolution of a technology. The third is exploring the implications of global networks for a given purpose. For example, which network structures are the most appropriate for knowledge diffusion? Which networks deter creativity? In this third category, the focus is on global network characteristics, like density, small-world coefficient and centralisation. Finally, some central debates in social network theory have implications for innovation networks, and these are covered in the last section of this chapter.

Network Data Sources in Innovation Studies

Various data sources can be used to construct and analyse innovation networks at the inter-organisational, intra-organisational and market levels. Commonly used data sources for inter-organisational networks are strategic alliance databases, patent data, data on project partnerships, litigations, board interlocks, underwriting syndicates and interviews and surveys. Survey and questionnaires and market research are used as data sources for the diffusion of innovations and user behaviour, but in recent years social media data sources have been increasingly used.

The Bibliothèque Nationale de France (BNF) in Paris is the largest documentation centre in France and contains an estimated 14 million resources, including print material, manuscripts, photographs, maps, coins, medals and multimedia documents. The BNF is organised into fourteen departments across five sites and has 2,400 employees (as of 2013). One of these sites is the Bibliothèque de l'Arsenal, which is devoted to sixteenth- to nineteenth-century French arts, literature and historical documents. The thirteen remaining departments are specialised in widely differing fields. One of the strengths of the BNF is its openness to external collaboration, and it is involved in many networks at national, European and international levels. Examples include participating in joint electronic databases with other libraries in France and abroad, carrying out research alliances that require close coordination and working with partners on a regular basis. One of its projects is the development of Gallica, its electronic library system. Gallica is the result of collaboration between many institutions, including the Ministry of Culture, Centre National du Livre (National Book Centre), Syndicat national de l’édition (Publishers Association), private firms and other electronic resource distributors. Inside France, the BNF is part of a national network of libraries and participates in research programmes and committees, provides expertise in document restoration and preservation, and is involved in the organisation of promotional and digitisation activities.

BNF is a nested network; it is a single node in an inter-organisational network in France and abroad. But BNF is also an inter-organisational network in itself. At this second level of the nested network, the nodes are each of its fourteen departments of which the Bibliothèque de l'Arsenal is one.

The nature of a technology or product system has an important role in the way innovators interact and communicate with each other. The complexity of a product system, its modularity and the nature of the knowledge base largely shape the ways in which producers exchange knowledge and resources and organise production. Therefore understanding how different aspects of a product system influence networks is important in order to align networks effectively with new product development.

A product system can be analysed along three dimensions according to the involvement of different actors. The first is backstage, concerned with the design and manufacturing of products through the networks of producers. The second is the involvement of users in the design of the products, where networks between users and producers are critical. The third dimension is user-only networks, through which knowledge about the product system diffuses and products are adopted. In this chapter, we take a closer look at the first two of these dimensions, namely the relation between the nature of a product system and how it fits with interactions between producers, and interactions between users and producers.

Modularity

Modularity is an important aspect of a product system and has implications for interactions between innovators. The effective coordination of different modules in a product system is vital to avoid scheduling problems and increased costs and to ensure the timely flow of information and resources.

Why are interactions between diverse actors such an important part of innovation? Consider the following news headlines:

Over the last decade, research on networks has increasingly turned to the micro-processes of tie formation and dissolution, as well as the evolution of networks. Tie formation usually depends on individual attributes (personality traits or the specific resource needs of an organisation), dyadic attributes (homophily or physical proximity) and the structure of the local network around an actor. At the actor level, the emphasis is on the ‘agency’ side of network formation, as compared to the long-standing focus on the effect of networks on performance and behaviour. As Poggi's quote at the beginning of this chapter suggests, too much focus on the performance effects of networks can draw our attention away from the role of agency in network formation.

Understanding the tie formation process is important for innovation management in a number of ways. For one thing there can be a gap between what organisations do and the best course of action for a specific objective. Consider an organisation looking for partners with complementary resources and capabilities. How do managers scan the set of possible collaborators? Research shows that physical or social proximity can drive such search processes. But as we have seen in previous chapters, there are myriad contexts in which non-proximate actors can bring significant benefits. Understanding how ties are formed increases awareness about the opportunity costs of established routines in tie formation. Tie formation is the result of search efforts to find the most suitable partners and these efforts may not always be efficient.

At first glance, there is little in common between the Seven Bridges of Königsberg and the New York Training School for Girls. Yet they mark the beginnings of a particular way of seeing what is around us, which is today referred to as the network perspective. In 1736 Leonhard Euler found a mathematical solution to the problem of devising the most efficient way of negotiating the city of Königsberg on foot: the Seven Bridges of Königsberg is often quoted as the first study in graph theory. Graph theory refers to the application of mathematical concepts and theories to a system that is composed of nodes and edges. Euler saw areas of land and bridges as a network of four nodes and seven links respectively (Figure 1.1). Two hundred years later, in a very different context, a leading sociologist, Jacob Moreno, conceived a similar network for the New York Training School for Girls, which had a problem with runaway students. Moreno saw the girls as nodes and explained their runaway patterns by examining influence between them (Figure 1.2). His experimental work on this case is considered the leading study in sociometry today.

These two contributions in strikingly different contexts reveal just how broad a range of phenomena can be regarded as networks.

Most of the time learning in alliances is difficult, frustrating and necessitates effective management of knowledge transfer. The case of New United Motor Manufacturing, Inc. (NUMMI) shows how General Motors learned about the Japanese lean production system through knowledge transferred from Toyota to NUMMI, and then from NUMMI to GM (Inkpen, 2008). The NUMMI automobile manufacturing plant was a joint venture between GM and Toyota, operating between 1984 and 2010 in Fremont, California. Initially, GM faced many challenges in establishing effective knowledge channels to bring external knowledge back into the firm. For one thing, there was strong resistance to a joint venture with Toyota from GM engineers, who felt their own capabilities were under threat. GM-based advisors who were appointed to NUMMI were insufficiently prepared to take the knowledge back to GM and implement what they had learned at NUMMI. Few people in GM were open to the new knowledge they brought back because it challenged the firm's existing operational routines. Faced with these problems GM set up new channels to improve knowledge transfer. Advisor training and rotation programmes were implemented, in which all advisors were assigned GM tutors and regular visits to NUMMI were organised. These processes were facilitated by the formation of a Technical Liaison Office, which over time became the core element of building up alliance learning capabilities in GM.

The case of NUMMI highlights how intra-organisational networks can be aligned during the formation of an alliance. There are many barriers to learning from alliance partners: resistance inside the organisation to new knowledge, difficulties in communicating with a partner, concerns about the leakage of valuable knowledge, problems related to heterogeneous knowledge endowments, and so on.

One of the uses of the network perspective is managing organisational change, where success usually rests on the participation and commitment of people in the organisation. A change project is usually initiated at the management level but people's opinions, beliefs and attitudes about the change largely form and mature through informal networks. Some people in these informal networks are better positioned to diffuse information, bridge management and employees or persuade others in favour of change. Network analysis also reveals cohesive subgroups, in which people tend to share common attitudes. After a reorganisation, the change in formal networks is often rapid but the change in informal networks can be slow to adjust. This is important in two ways. First, by understanding how change will alter resource distribution, possible sources of resistance in the organisation can be spotted and are most likely to come from those who will lose status or privileged network positions. Second, it is important to understand how organisational networks will change after the event because formal and informal networks are complementary in the accomplishment of tasks and organisation-level goals (Gulati and Puranam, 2009).

‘Durable, engineered and permanent – those are in opposition to idiosyncratic, delicate, ephemeral.’ These are the words US artist Janet Echelman uses to describe the challenge she encountered when asked to build one of her giant fishnet sculptures on the waterfront at Porto in Portugal (see Figure 7.1). Creating a sculpture that gives the same feeling as the fluidity of a fishnet but could survive ultraviolet rays, pollution, salt, wind and traffic was a challenge that her skills and knowledge as an artist were not sufficient to master. This led to her collaboration with Peter Heppel, an engineer specialised in lightweight structures and aerodynamics. The combination of Echelman's art and Heppel's technical expertise resulted in steel structures 40 metres in diameter that flow like fishnets on the coast at Porto.

Janet Echelman's work is unique but two more general lessons about creativity can be drawn from her story: exposure to distant worlds and the recombination of distant elements. Janet Echelman had no expertise in fishing. A Fulbright scholarship took her to India, where she had the idea of creating giant fishnet sculptures as she watched fishermen on the beach. However, she did not have the technical knowledge to combine two opposing worlds – ‘durable, engineered and permanent’ versus ‘idiosyncratic, delicate and ephemeral’. This led to her collaboration with engineer Peter Heppel, in which these two contrasting worlds were recombined.

In Chapter 6, we focused on inter-organisational relations in search processes. This chapter deals with creativity in search.