Friday, June 29, 2007

A somewhat unified view of decision making: part 2

Basic decision model augmented with the loss of insight over time and the rebuilding of insight through simulation, rehearsals, and other means

My last posting described the basic model, but it left lots of gaps. For one, Actionable Insight isn't static, as we all know. While some insights maintain their validity for years and some perhaps forever, others either lose validity as new knowledge is developed or lose relevance as their need goes away. I'll show that effect as a flow draining out of Actionable Insight into another cloud, for it doesn't matter what happens to it; we just don't have it anymore. Don't forget to click on the image to see a full-sized, pop-up version.

If Actionable Insight can drain away, how do we refresh it?

We no doubt have as many ways of gaining insight as there are people. Trying to make a comprehensive list might be futile, and it doesn't really contribute to my point. I want to highlight one approach that I think helps integrate the various ideas I've been exploring.

Someone once said, "Don't ever go where you have not been before." You could also say "Don't do anything for the first time"; it sounds similar and seems to make as much (or as little) sense.

With possible apologies to Tom Peters (I'll explain that note in a later column), I think there's a lot of merit to those statements. Actors and musicians practice it faithfully; symphony orchestras and theater companies always have rehearsals before they show their work to an audience. You might bring up improv or jazz as counter-examples, but, as says, it takes hours of rehearsal to learn how to improvise.

On a more academic note, Dietrich Dörner has made a career of studying decisions and how people make decisions that go awry. He has advocated simulations as a way for us to learn more rapidly than we can in real life and with less risk.

An organizational simulation is, in many ways, nothing more than what musicians and actors do in rehearsals: an opportunity to go through the important parts in private, with the ability to try things different ways, and with little risk. We can learn what works so that things go more smoothly when we do it for real.

For example, I recently did a set of simulations to explore a marketing program for symphony orchestras. In a relatively short time, I was able to create a simulation model of the essence of the problem Drew McManus was trying to address and to apply multiple solutions to see how each might work. The model isn't good enough (nor, likely, should it be) to tell Drew exactly how many additional tickets he'll sell for Friday's concert, but it should give him better insight into what is key for making his program succeed.

Certainly there are other ways we also use to increase our insight; why am I writing about simulation? It's because of the link to Gary Klein's recognition-primed decision model. Gary's model, simply put, is that we tend to simulate situations quickly in our heads. If our plans of action work out, we put them into place in the real world. If they fail, we try a different approach until we find a mental simulation that works. We can do that really quickly, and that's key to our survival.

But there's a catch. As Barry Richmond once said, our mental models aren't always good enough, and our ability to simulate those mental models in our heads doesn't always suffice, thus leaving us with holes in our feet.

I've come to believe that simulations and rehearsals are important ways we prepare to make good, quick decisions later, when we need to. These, plus all the other analytic and rational tools we can bring to bear, are the means by which we enable ourselves to be good at the recognition-primed decision model. Certainly we can do those simulations in other, more manual ways, but, for many situations, the alternatives are tedious and error-prone. Certainly we can and do learn in other ways, and those are important—I'll even talk about some of them later. What simulation almost uniquely provides is the ability to specify a situation and see the logical ramifications of that situation.

All this has been to support two conclusions:

  1. The recognition-primed model and more rational decision-making models aren't two choices from which you get to select only one. They can work together, the more rational model providing you the stock of insight you need to make faster, better recognition-primed decisions.
  2. Simulations (rehearsals for musicians and actors) are great tools to build insight for those recognition-primed decisions, for both are in fact simulations; they just have different constraints. Thus the structure of the learning you get from the one should fit nicely into the structure of the insight you need for the other.

There's more to come. In future installments, I'll cover how Insight Quality, action learning, and Tom Peters figure in.

If you know someone who is facing tough organizational or business problems, problems that persist, problems that don't lend themselves to quick, intuitive solutions, introduce us. Perhaps I can help.

Labels: ,


Post a Comment

<< Home