In a complex adaptive system, local events and interactions among the “agents,” whether ants, trees, or people, can cascade and reshape the entire system—a property called emergence. The system’s new structure then influences the individual agents, resulting in further changes to the overall system. Thus the system continually evolves in hard-to-predict ways through a cycle of local interactions, emergence, and feedback. In nature we see this play out when ants of some species, for example, although individually following simple behavioral rules, collectively create “supercolonies” of several hundred million ants covering more than a square kilometer of territory. In business we see workers and management, through their local actions and interactions, shape the overall structure, behavior, and performance of a firm. In both spheres these emergent outcomes influence individuals and create new contexts for their interactions. Whether we look at team dynamics, the evolution of strategies, or the behavior of markets, the pattern of local interactions, emergence, and feedback is apparent.
Complex adaptive systems are often nested in broader systems. A population is a CAS nested in a natural ecosystem, which itself is nested in the broader biological environment. A company is a CAS nested in a business ecosystem, which is nested in the broad societal environment. Complexity therefore exists at multiple levels, not just within organizational boundaries; and at each level there is tension between what is good for an individual agent and what is good for the larger system.
What does this mean for business leaders?
First, they need to be realistic about what they can predict and control, what they can shape collaboratively, and what is beyond the reach of managerial influence. In particular, they need to expect that unpredictable and even extreme emergent outcomes will cascade from actions at the lower levels. A clear example is the financial crisis of 2007–2008, during which risk created by subprime lending in the U.S. real estate market spread catastrophically throughout the global financial system.
Second, they need to look beyond what their firms own or control, monitoring and addressing complexity outside their firms. CEOs must ensure that their companies contribute positively to the system while receiving benefits sufficient to justify participation. Companies that fail to create value for key stakeholders in the broader system will eventually be marginalized (similarly, business ecosystems that do not provide benefits to their members will experience defections). Consider Sony, which brought out its first e-reader three years before Amazon’s but lost decisively to the Kindle and withdrew from the market in 2014. Because it failed to provide a compelling value proposition that would mobilize key components of the publishing ecosystem—authors and publishers—it could offer only 800 titles when its e-reader launched. In contrast, Amazon initially sacrificed profits, selling e-books for less than what it paid to publishers. It also invested in digital rights management to spur the growth of its ecosystem. With the support of other stakeholders, it launched with 88,000 e-books ready for download.
Third, leaders must embrace the inconvenient truth that attempts to directly control agents at lower levels of the system often create counterintuitive outcomes at higher levels, such as the stagnation of a strategy or the collapse of an ecosystem. They must avoid relying on simplistic causal models and trying only to directly manage individual behavior, and instead seek to shape the context for that behavior. For example, questions or simple rules aimed at fostering autonomy and cooperation and leveraging employees’ initiative can be more effective than top-down control in shaping collective behavior.
We have identified six principles that can help make complex adaptive systems in business robust. (For a related examination of complex adaptive systems in nature, see Fragile Dominion, by Simon Levin.) They derive from our study of features that distinguish dynamic systems that persist from those that collapse or decline. The first three principles are structural; they deal with the design of the system and are broadly seen in nature. The second three are primarily managerial; they deal with the application of the intelligence and intentionality provided by humans. Their key features have been observed in a wide variety of managed systems, from fisheries to global climate control agreements.
A few caveats: Each principle confers a cost and has an optimal level of application; leaders must carefully calibrate how aggressively to implement each one. Furthermore, the principles are in tension with one another; emphasizing one may require deemphasizing another. Leaders must consider how to balance the principles collectively rather than treat the application of each as a unitary goal. For clarity we will illustrate the principles one at a time, but robust systems typically exhibit many or all of their characteristics simultaneously.
Is your strategy aligned with your risk environment? Use the diagnostic tool here.