Complexity has become “the new normal” for industrial sectors as diverse as life sciences and oil drilling, Rutgers University Professor Deborah Dougherty told a May planning meeting for a proposed multi-institution research center with CIMS the lead site. Dougherty suggested several possible research areas for the Open Innovation Management Network, which would be dedicated to advancing the management of technology and innovation. Managing complex innovation would be one research area for the important reasons she details below:
Many kinds of innovation are complex, not simple, and not even complicated. But many managerial models for innovation are based on simplicity, where the relationships between cause and effect are known; the problem is well defined so innovators can drive to solutions; activities are broken down into separate modules and steps; and the cycle time is short so work can be time-paced.
Complicated vs. Complex
Complicated innovations involve many diverse elements, but relations between cause and effect are still known. Complex innovations, on the other hand, comprise many divergent subsystems that interact in unpredictable ways. Complex innovations involve unknown relationships between cause and effect; the innovation problem itself is discovered in the process; activities cannot be decomposed or disentangled; and the process unfolds over many years.
Innovation is predominantly complex in sectors such as bio-pharmaceuticals, life science products for agriculture, new materials, alternate energy systems, and, as we have discovered recently, financial instruments and deep well drilling. Firms in these sectors cannot avoid complex innovation and they cannot shut complexity away in separate skunk works to focus on normal operations.
Complexity also captures more of the actual challenges in many innovations, and reminds us to not sandpaper away the unknowns but rather to embrace them. In short, complexity is the new normal.
When applied to product innovation, complexity suggests three lessons that innovators in many industries can benefit from: building skills to search for interdependencies; developing a new mode of learning — abduction; and managing the processes, not just the outcomes.
1. To create complex new products, innovators must discover a workable configuration of interdependencies among the many subsystems involved in the products—they must shift from knowing about the parts to also knowing how those parts interact. The recent deep well oil disaster and the troubles at Toyota suggest that people were not thinking about interdependencies among their separate systems, did not look for them, and maybe did not know how to look for them.
Complex products are not stand-alone things, because their functionality — and thus their value — depends upon how they interact with all the other systems they are a part of. For example, a new drug is a molecular system that interacts with a disease system to restore the human body system to health, but it also interacts with other biological systems in the body as well as with health care and regulatory systems. Drug innovators must discover all the key interdependencies among these systems that define each new drug. Similarly, the value of a new bio-fuel depends on how it interacts with diverse engines and distribution systems. Simple innovations allow us to bracket off interdependencies into separate steps when we need to develop ways to search for these interactions actively in the design and development process.
2. Learning by abduction provides a systematic way to search for unknown interdependencies. The more familiar processes are induction, or learning from unique circumstances, and deduction, or learning by predicting an outcome and confirming that prediction. Abduction refers to systematically examining a mass of facts and allowing those facts to suggest a theory. Abduction involves reasoning from clues to reach tenable hypotheses and testing those hypotheses in subsequent inquiry.
This process of learning depends on raising good questions, engaging actively in dialogue about the questions and what the answers might mean, and reframing the problem and the theories. Abduction involves anticipation rather than more precise prediction, and is more fluid. We can identify three thrusts that comprise this process, however:
• Reaching out systematically for clues to interdependencies, based on a solid hypothesis about what might be involved for a particular innovation, and following those clues wherever they lead (not drive to answers). Reaching out for clues requires active engagement in the process. Clues are discovered as people carry out the search, and they make sense only as part of the search based on the questions being asked and the ideas being explored.
• Another thrust involves iteratively integrating the clues into possible configurations by interacting across many sets of experts (not honing in on one system). The goal here is to collectively determine if a workable configuration of interdependencies is emerging and if so, whether it points to next steps in the search. Iterative integration is nonlinear since it may seem to cycle “back” over old ground, but in fact it deepens understanding.
• A third thrust is judging whether or not this emerging configuration of interdependencies is worth continuing with (not reaching a clear end). There are no clear answers, but innovators can consider if too many interdependencies remain unknown, and if the questions are getting clearer and the answers more informative.
3. The third lesson is that these abduction thrusts cycle together over time, which means that strategic managers need to focus on the process, MANAGING from Page 14 not only on the outcomes. Managers cannot foresee the future or impose programs of control, and nonlinear systems like complex innovation react to control in ways that are difficult to control. Managers can establish directions and boundaries, and modify these over time by setting constraints, observing outcomes, and tuning processes by altering constraints. For example, they can ask if people are productively engaged in the process of reaching out for clues, rather than working separately or simply dumping in data.
Managers can facilitate iterative integration by ensuring that people work across boundaries. They can foster good judgment by investing in the development of key experiments to clarify interdependencies. Resources must flow into the projects when needed — as with emergency response systems where disasters do not occur on a predictable schedule.
Finally, because the cycle times are so long, projects outlive strategies. Rather than attempt to shuffle projects to fit strategies, companies need repertories of strategies they can shuffle to fit with emerging projects. The whole ecosystem has to engage in abduction.
To Learn More
Complexity has faded in and out of managerial fashion, but it may be here to stay this time because it helps to define key innovation problems. Each of the lessons outlined above is now being explored in various studies, and we will continue to pull these ideas together. Several papers on complexity in bio-pharma are currently in the works and I will be glad to send publication references to those who email me.
Professor, Management and Global Business