Thursday, September 30, 2010

Is Decision Analysis Dead?

In his book, Rationality for Mortals, Gerd Gigerenzer relates the story:
 A decision theorist from Columbia University was struggling with whether to accept an offer from a rival university or to stay.  His colleague took him aside and said, "Just maximize your expected utility--you always write about doing this."  Exasperated, the decision theorist responded, "Come on, this is serious." 
I happen to love decision analysis and decision theory.  Its a fascinating mental challenge to understand and to apply logical, rational, mathematical principles:  chains of logical inference, Bayesian probabilities, marginal utility theory, etc. Its even more challenging--and perversely fun--to apply these principles to real decisions.  Decision analysis can yield a good result when it is applied to the right kind of problem.  However, it often seems to me that decision analysis is a poor fit for the type of decisions that businesses face today.

We can think of business as a game.  So let's contrast two types of business "games".

Game A.  In Game A, we have a clear goal, and we have or can create clearly defined options.  We have decision criteria which are not only clear but measurable through natural or constructed scales.  Using a clearly defined process and methodology (which includes decision analysis and other problem solving techniques), we can progress toward our goal.

Game B.  We begin with a fuzzy goal concerning a new business or a new product or some strategy.  We're trying to create something new that will allow us to create a viable new business or build a competitive advantage for an existing business.  We have a fluid list of criteria which are not precisely measurable.   We want to discover America, but we don't even know that exists yet.  In this game decision analysis feels like something irrelevant and antiquated.  The last thing we want is to get hung up on structure and mechanics and quantics.  We still want to make a decision by applying criteria to a set of option, but creativity and discovery are more important than precise measurement.

The world is becoming more of a Game B kind of place where creativity is central not just to creative business (media, advertising, consumer products), but to most businesses in the fiercely competitive global economy.  And more and more employees, regardless of job description, are called on to act more like creatives.

Do we need a new type of decision theory or decision analysis that fits our fuzzy Game B world?  If so, what would that be like?  Would it repudiate most or all of decision analysis?  Or would it integrate the two in some way?

There are two objections to decision analysis.  First, the methods are cumbersome and are often less efficient than various heuristics. Second, as discussed above, they don't fit the kinds of problems we face today. They don't match the environment in which we find ourselves.

No comments:

Post a Comment