In this section, I describe how strategic and institutional managers can work on these problems over time. The institutional innovations in governance structures enable the development, evaluation, and reframing over time of value creating opportunities that map out directions for project (new drug therapies) and knowledge process (new integrations of sciences and technologies to support discovery) innovations.
Strategically managing value creation to anticipate deeper futures
Strategic managers need to work collaboratively across the eco-system to leverage all the resources embedded in learning events that are generated by the product, process, and strategic problem setting and solving cycles. The simple question for strategizing in complex eco-systems is how can we generate value creating opportunities to shape and guide complex innovation projects and processes that emerge over very long periods? I suggest a diverse portfolio of value creating possibilities that direct emerging innovations into actual commercial or publicly valuable applications. The portfolio overall emerges and changes over time as it maps out opportunities in various time periods, and informs the entire eco-system about possibilities. The portfolio affords a deeper look in time because various possibilities work as stepping stones into the future. Strategic thinkers imagine where they can go with what the system is learning, and they shape and redirect that learning with ongoing ideas of value creation. They implement imagined configurations of learning events to see how they might work, to consider what else seems to be going on, and to surface new possibilities. Different groups of agents and agencies would collaborate over different sets of value creating opportunities, but the entire eco-system can generate a portfolio of possibilities.
Formulating Hypotheses by Imagining Configurations of Interdependencies: Strategic managers would consider what certain learning events that arise in innovation efforts (including developing strategies) suggest for a future value creating possibility, and consider how these learning resources could work together to create a viable configuration. Going forward in time, managers would hypothesize how the configuration they imagine will emerge based on current and anticipated learning events, what interdependencies might be involved, and to acquire and deploy those resources. I am describing basic business planning, except that this planning reaches out over time, anticipates emergent changes based on learning, and emerges continually. Imagining a configuration of interdependencies among knowledge resources from learning events includes identifying certain assumptions about what would make configurations a good opportunity, what are people going to learn for value creation by developing the configuration, and how will it help muster the staying power to persist and learn. Different values might include opening a new niche in the market and/or in therapeutics, generating some protection from competition, and providing a long term foothold.
A good exercise to learn about the cycle of abductive learning routines would be to interrogate current business models such as shutting down internal R&D and outsourcing the work, or focusing on block-buster drugs. For example, what configuration of interdependencies among what knowledge resources generates value how by outsourcing R&D? What learning events are this strategy based on, and what are the underlying assumptions to be evaluated and reframed? What learning events give credence to the block-buster business model?
Other examples of possible value creating opportunities for pharmaceuticals include Christensen et al.’s (2009) suggestion that small firms can flourish with diagnostics that identify the specific customer base for particular drugs, and serve small markets and unserved customers. Rather than trumpet breakthroughs in biotechnology, with abductive reasoning one explores what a breakthrough depends on to be useful to innovation, or to effect in some way an important outcome. For example, what does this breakthrough technology depend on to help define a new business opportunity, and bring it into existence? And consider the Bill and Melinda Gates Foundation’s work on developing malaria vaccines and other ways to address this debilitating disease in impoverished societies. Their mapping would consider a few different configurations of resources (including not just drug discovery but also marketing and distribution to areas that are difficult to reach), and then hypothesize what each would achieve and how, and what learning events are needed to gauge progress.
Another type of pharmaceutical value creating opportunity is figuring out what works well and what does not. Scannell et al. (2012) point out that most of the R&D costs are in failures, so they recommend that firms develop a chief dead drug officer to investigate the failures. These failures are learning events. People in this role across the eco-system would set out the major factors responsible for the progressive decline in R&D productivity, and compare different therapeutic areas to explain the differences between them in productivity. Chief dead drug officers can explore the extent to which factors are tractable, such as where the molecular reductionism of rational drug design or the brute force screening of high throughput systems become distractions, and where do they help. Chief dead drug officers can measure the veracity of previous diagnostic forecasting exercises, and examine which clinical test requirements are most costly and least valuable.
Evaluating the Configuration of Interdependencies by Elaborating and Narrowing: These ideas are hypothesized configurations of interdependencies among learning events that might constitute a value creating opportunity that is to be empirically evaluated and reframed. Instead of assuming that outsourcing R&D to universities and biotech firms will automatically work, for example, strategic managers would implement this idea in a particular way (e.g., with a certain university or type of collaboration), and experiment with the hypothesized configurations to generate evaluative knowledge. By elaborating and narrowing around the interdependencies in the configurations, managers and innovators explore the actual interactions to see how and why their hypothesized configuration actually works. They elaborate out around a subset of interdependencies among resources to consider if these interdependencies are central or not, how and why, and what else they can learn. They narrow in on interdependencies that seem stable and useful, and then elaborate out again to see other possibilities. Elaborating and narrowing balances new knowledge with existing insights. For example, managers might narrow in on particular kinds of university collaborations, and elaborate out around how fully and usefully knowledge transfers as assumed, and if not why not. The goal is to evaluate if this configuration of resources can create value based on why managers thought it would, and to explore underlying assumptions in order to learn.
Evaluating hypothesized value creating opportunities combines clock time and event-time pacing. The goal is to make better judgments, not simply better decisions, about why and how this is a good business opportunity. Clock-time pacing questions include: 1) how long does it take us to figure out that we are at a good or bad point; 2) how quickly can we evaluate learning events; 3) how quickly do others provide input to our analyses; 4) how quickly do we identify alternatives and choose among them to take next steps. Event-time pacing evaluative inquiries include: 1) we think we are here, is here good enough for a possible value opportunity; 2) are we able to handle a larger variety of configurations; 3) how much are we willing to pay to explore potential, and 4) are the learning events that emerge getting better and better? Both temporal structures can address strategic issues such as does this opportunity open a new niche, protect us from competition, extend our existing franchise adequately, give us a long term foothold, and allow us to know more about the opportunity as we also generate revenues.
Tests of confirmation are part of evaluating the imagined configuration of interdependencies, but they do not stand alone. Strategic managers need to go beyond confirmation to develop insights into why and how a particular element affects the configuration, and what else may be involved. Since the goal is learning, managers go beyond does the configuration work or not, and consider how and why it works.
Reframing the Imagined Configuration of Interdependencies by Iteratively Integrating: The third strategic abductive learning routine cycles back to the imagined configurations by refining or reframing the interdependencies among knowledge resources that are involved. Participants in the strategic problem setting and solving process critically examine assumptions, deliberate over different perspectives, and bridge possible differences into new shared directions (Ansell 2011). Strategic managers might refine and replace milestones, and develop new performance objectives that reflect new alternatives and consequences learned from evaluation (Grandori 2010). Strategic managers would also rethink how clock-time and event-time pacing combine, and how well the combinations create a rich set of reference points that people can use to anticipate more possibilities further into the future. Reframing might revise the future trajectories of anticipated activities that need to be accomplished, and coordinate attention and effort to carry out those activities. Event-time pacing structures the inherently exploratory searching and helps to constrain the short term nature of clock-time by keeping the future open to emergent possibilities. Clock-time pacing helps to constrain the potentially expansive searching, and marshals the development of resources that can be clocked. Strategic managers in the innovation eco-system identify new future trajectories and eliminate others based on emergent learning.
Institutional innovating to generate collaborative commons
The strategic management problem setting and solving cycles cannot take advantage of emergence unless the eco-system also generates new governance structures that will enable the collaborations among disparate entities that must take place. As already noted, the emergence of new sciences and technologies into viable innovations has always depended on governance structures for collaboration, along with government and private support. However, complex innovation eco-systems require continuous innovation in governance structures, since the kinds of enterprises that require collaboration will continuously emerge and evolve. As Ansell (2011) explains, the objective of many of these collaborative problem setting and solving cycles will be to build up problem solving capabilities and to engage in ongoing problem solving efforts. Participants working on a particular category of strategies (e.g., using genomics, or developing different clinical trials) would identify, develop, and manage over time their particular configuration of rules and relations.
A simple question for institutional innovating in complex eco-systems is how can we work together effectively over time to plan, experiment with, learn from, and revamp value creating opportunities, ways to integrate sciences and technologies to support actual innovation work, and/or specific therapeutic innovations? The institutional challenge is to develop a configuration of interdependencies among rules and relations that form the particular governance structure. I suggest that participants build on the non-market structures that already exist to include many organizations and agencies around diffuse objectives. These existing systems provide models for different kinds of governing structures that enable collective, long term co-evolution of sciences and technologies for innovation. These systems include open innovation (Chesbrough 2003), regional clusters (Gilbert 2012), network innovations (Iansiti and Levien 2004), industry platform systems (Gawer and Cusumano 2002), and the more general ideas about technology trajectories that build on the efforts of many different organizations and innovators (Dosi 1982; Floricel and Dougherty 2007).
For example, “open innovation” emphasizes the idea that knowledge is widely dispersed and that some innovations require multiple actors (Chesbrough 2003). Research demonstrates that open innovation requires participating firms to develop strategies for the long term, build capabilities for absorptive capacity (i.e., abilities to spot, bring in, and use emerging ideas), and build up abilities for long term partnering (Di Minin et al. 2010). There is no simple “outsourcing.” Regional clusters that foster technological innovation depend on diverse sources of knowledge, government policies that foster new technologies rather than allow incumbent firms to dominate, and rich networks of relationships that enable the exchange of tacit knowledge (see summary by Gilbert 2012). Organizations willingly participate because they can tap into supplier networks or access customers. These clusters demonstrate that for complex innovation eco-systems, multiple organizations can co-create a context of mutual learning that enables participants to muster the staying power to persist and learn.
Firms that compete also collaborate over standards setting, because standards, for example in communications systems, enable all firms to keep innovating (Piepenbrink 2015). Participants in standards setting bodies follow rules for IP and appropriation, for responsibilities for reciprocity, and for how changes in standards are continually orchestrated. Network and industry platform systems (Nambisan and Sawhey 2011) depend on the active leadership of large firms that continually upgrade the core technology or architecture, enable ongoing negotiations among participants for IP, and provide market access. In return, the large firms receive continual innovation in components and other network externalities that keep their core technology valuable. Ansell (2011) summarizes studies in the public domain among collaborating agencies around improved policing that depend on a rule of accountable autonomy.
These existing governance structures suggest a variety of rules and relational elements for collaboration that institutional managers can select from and work into configurations of interdependencies that they can hypothesize, evaluate, and reframe over time. These institutional elements include: 1) a problem solving focus: centering on what people need to collaborate over such jointly as developing a kind of drug therapy (e.g., immune therapy for particular cancers; combining drugs to form “cocktails,” or trying out a new business model); 2) heterarchical organizing that enables participation and inclusion with various levels working on different aspects of the problem; 3) leadership by individuals and by large firms or coalitions of firms; 4) self-organizing, where organizations participate voluntarily because they gain value from that participation; 5) co-dependence among participants to reinforce active participation; 6) protected niches for value creation that includes end-to-end with customers, perhaps with disease foundations, clinical research groups; 7) intellectual property protection and development; and 8) articulation of basic rules for partnering and accountability.
Institutional managers in the drug discovery eco-system (and for any other complex innovation problem) can begin by developing collaborative commons around specific pressing problems such as clinical trials for diverse therapies or advancing immunotherapy. Experiments with these and similar problems are already on-going (albeit in localized or one-off modes), and should provide considerable insight for formulating, evaluating, and reframing hypotheses about governance structures that enable ongoing collaboration around specific concrete problems. The difference for complex eco-systems would be that these emerging governance structures would continually be developed, experimented with, and revised based on progress with actual problems. People would not simply hold a big meeting and then go on about their separate work.
As Dougherty (2016) details, cycling through the abductive learning routines for institutional problem setting and solving begins with imagining a configuration of interdependencies among relational elements and rules that can produce the specific kind of collaboration needed to address a certain problem. For example, if the problem is developing improved models for the early evaluation of drug possibilities, participants would include large firms, research hospitals, regulators, and small firms. Leaders need to be appointed, perhaps a coalition with clear rules for IP, participation, and task forces to oversee particular processes. Evaluating digs into the hypothesized interdependencies to examine assumptions and figure out what works and not, and why and how. Over time, experiments generate insights into how to develop and deploy metrics about how well the collaboration is doing, update and revise the problem, arrive at joint decisions and agreements, and foster ongoing participation. Some collaborative commons will be relatively short-lived, while others may continue for years.