A major problem with current strategy development is in the predictability of the system. Complex adaptive systems are notoriously difficult to predict, primarily because
- There are often an overwhelming number of variables to consider.
- The dynamics of the systems create emergent behaviors, which cannot be predicted from decomposing the system into individual components.
- The dynamics of the systems are generally poorly understood.
Fundamentally, many business strategies focus on the desired outcome. They then create a plan to achieve that desired outcome. The problem with this, of course, is that the desired outcome is just one of a large number of possible futures. From simple statistics, the probability that the future is not the desired outcome is much higher than the probability that the future is the desired outcome.
To make matters worse, the activities used to achieve the strategic goal will change the system, and those changes will be, by the very nature of complex systems, difficult or impossible to predict. In the end, the chances that a strategy will "succeed" are dramatically lower than the chances the strategy will "fail." This conclusion is supported by both Duncan Watts (in his book "Everything is Obvious, Once You Know the Answer") and Michael Mauboussin (in his book, "Think Twice").
If, then, current approaches to strategy are more likely to fail than succeed, how do we change the approach to improve performance?
There are two critical aspects to "strategy" in complex environments: robustness and agility. Shifting focus to these concepts during strategy development, instead of a particular "future configuration," improves the chances of success.
Robustness is feature persistence across structural change in the system. This means that a robust behavior will exist regardless of dramatic environmental changes, or changes in the system. For example, a robust service will continue to function if, say, 80 percent of the components are removed. Or, in a technology environment, the system will continue to exhibit the behavior if the hosting model changes.
Agility, contrary to popular usage, is not "doing stuff faster," it's a measure of the organism's ability to adapt to changing environments. It's not a "speed of execution" measure, it's a "rate of change" measure. In a biological system, it's the mutation rate that governs how quickly the organism evolves. There is also the concept of an energy boundary for an organism (or organization). If you push adaptation/mutation (or agility) beyond that energy boundary, you will, as David Krakauer says, "remove the brakes from the car," and the organism will never stabilize on an optimum - it will just keep mutating rather than stabilizing.
The key to a strategy, then, is to balance robustness and agility. To some extent, you can balance lack of robustness with increased agility: if the environment changes in a way that you can't continue to inherently provide a critical feature, it might be ok, provided you have the organizational ability to mutate to meet new conditions. The trouble appears when an organization can no longer provide critical features, and cannot adapt rapidly enough to the new environment. That's a disruptive change, and can lead to organizational failure.
The other thing to remember is that robustness comes with a price. If your organization focuses on cost optimization, efficiency metrics, and high degrees of specialization, you are by definition creating a fragile (i.e., not robust) structure. That'll make you a great competitor, provided your environment doesn't change - and the one thing we know from experience is that the environment will always change. More about this in the next post.
No comments:
Post a Comment