Why are some firms better than others at recombining ideas? Often, the capabilities involved are only understood in broad terms, rather than being analysed in detail. To help fill this gap, we wanted to disentangle two capabilities that are often seen as one: recombinant creation and recombinant reuse.
Recombinant creation means combining technologies that have never been combined before. It requires ‘broad’ capabilities: the ability to think laterally and make new connections across technological areas. If successful, it generates new technological combinations for the firm.
Recombinant reuse means refining and improving existing combinations to solve new problems, find new applications and identify new contexts where previous knowledge can be used. This is more about ‘deep’ capabilities, as the firm ‘digs down’ into its existing portfolio of recombinations to find new ways to use them.
To help answer the question of why some firms innovate better than others, we wanted to discover what drives these two recombinant capabilities. We focused on two aspects of firms’ organization: the network of collaborative ties that links inventors together, and the diversity of a firm’s technological knowledge base.
Networks of Collaboration
Experts such as inventors always specialize in one area. So a firm’s capability in recombinant creation depends on them exchanging knowledge through a network of collaboration. Often, this network isn’t reflected in the firm’s official organization chart, but in less formal relationships.
When the network is integrated, inventors can interact and collaborate closely with one another and draw on their deep knowledge of existing recombinations. They also feel a stronger sense of shared ownership and identity, so they perceive the value in everybody’s ideas. But when the network is fragmented across unconnected groups and inventors, there are fewer opportunities to work together, knowledge is isolated and opportunities remain unexplored. So our first hypothesis is that the more collaborative integration a firm has, the better it will be at recombinant reuse.
However, in an integrated collaborative network, inventors may be so focused on the value of existing ideas that they become negative towards radical new ideas, seeing them as unnecessary or risky. They may get locked into established approaches or ‘the way we do things around here’, preventing them from seeing new ways to combine ideas. So our second hypothesis is that the more collaborative integration a firm has, the worse it will be at recombinant creation.
Another factor in firms’ innovation capability is the diversity of its knowledge – in other words, whether its knowledge is spread over many areas or concentrated in a few.
When knowledge is diverse, it is harder for inventors to understand each other because they lack shared technological ‘languages’ or ‘thought worlds’. But, as we’ve seen, recombinant reuse depends on inventors being able to exchange knowledge freely and easily. So our third hypothesis is that the more diverse a firm’s knowledge, the worse it will be at recombinant reuse.
The upside to knowledge diversity is that it allows individual inventors to make new connections – the flashes of inspiration that would be more difficult if everyone shared the same specialization. The more ways of thinking there are in an organization, the better its chance of generating radical new ideas. Also, because knowledge diversity hinders the exchange of knowledge, inventors will be less likely to choose recombinant reuse. So our fourth hypothesis is that the more diverse a firm’s knowledge, the better it will be at recombinant creation.
Collaborative Integration and Knowledge Diversity Together
What about the interplay of these two factors? We reasoned that while close integration would help recombinant reuse, this effect would be reduced by knowledge diversity. The reason for this is that although inventors can share knowledge more easily when networks are integrated, the actual knowledge being shared is very diverse, which hinders communication and makes it harder for them to collaborate. So our fifth hypothesis is that the positive link between collaborative integration and recombinant reuse will be reduced by knowledge diversity.
Logically, the converse should also be true. As we’ve seen, integrated collaboration networks make recombinant creation less likely, because inventors focus on what they know. But this is harder to do when knowledge is more diverse. Also, inventors are more likely to find new angles when they all specialize in different areas. So our sixth and final hypothesis is that the negative link between collaborative integration and recombinant creation capability will be reduced by knowledge diversity.
Testing the Theory
To test our hypotheses, we carried out a study of 126 global semiconductor firms between 1984 and 2003. We focused on this industry because technological recombination is so important in it, and because firms vary greatly with respect to their organization of innovation activities. We used patent data to gain insights into collaborative integration and knowledge diversity, including such aspects as whether inventors had collaborated on an innovation, and whether it was based on a new or existing recombination.
Our results confirmed our first four hypotheses. Having an integrated network increases a firm’s ability to innovate through recombinant reuse, but reduces its capability in recombinant creation. Conversely, when a firm’s inventors are specialized in diverse technologies, it is worse at recombinant reuse but better at recombinant creation. So the drivers of recombinant creation generally hinder recombinant reuse, and vice versa.
Our fifth hypothesis was not confirmed, but the sixth was. This suggests that when firms have both an integrated network and diverse knowledge, they may be able to innovate through both recombinant creation and recombinant reuse.
Strategy scholars know that firms perform differently because they have different capabilities. What is less clear is why there are such persistent differences in capability between firms. Our work highlights the importance of two aspects of the firm’s informal organization: its knowledge base and the way its inventors collaborate. Informal structures are hard to change, which is why firms’ innovative capabilities tend to stay the same.
Sometimes, scholars define recombination too broadly. Our study shows the importance of identifying different types of recombination. We hope that the distinction between recombinant reuse and recombinant creation will be useful to researchers in the future.
Our work can also help managers by directing their attention away from formal structures and towards the informal side of their organizations. Changing informal structures may be difficult or time-consuming, but our analysis shows that improving recombinant capabilities can make a significant difference to the innovative capability of a firm.
"Where do firms’ recombinant capabilities come from? Intra-organizational networks, knowledge, and firms’ ability to innovate by technological recombination », published in Strategic Management Journal.