Tuesday, June 12, 2007

Plans, Goals, Forecasts, and Estimates

This entry on the Overcoming Bias blog got me thinking.

Bosses just want the number.

I see a lot of confusion between planning, estimating, goal-setting, and forecasting.

Forecasting as often used in business is simply giving someone the number that represents business performance. Forecasting accurately is impossible, except maybe in the very short or very long term. Nonetheless, you still at some point need to come up with a number. Forecasting is best done using probabilistic techniques and reporting events as probabilities. Simply predicting a number, even if you are right is of little value in most circumstances.

Can you imagine a weather forecaster simply saying, "Rain"? Without a percentage and a timing, it's simply not a real forecast.

Likewise when you estimate business performance, wouldn't it be better to say, given the manpower and marketing budget, there is a 10% chance that we can sell 10,000 units, and a 90% chance that we can sell 8,000 units. Wouldn't that provide more information than just 10,000 units. It is a richer way of expressing a "forecast" and helps put reality into the process of planning for contingencies.

Ah, but goal-setting is important and helps people keep focused. That is true, but goals are not good forecasts. Also the goal setting process is usually based on an arbitrary standard, and negotiated without good information. There's a lot of push, pull, and sandbagging that occurs. Any resemblance of a goal to a forecast is unintentional.

Plans set up a sequence of events or actions that can be taken towards achieving those goals. Plans like forecasts are never accurate, but they do help clarify your path. Plans can and should include resources, cash, timing, contingencies, and risks. There is usually a number that comes out of it, but that number is never a good forecast. Never. No never.

I have two references to point you towards:
The first is Why Can't You Just Give Me The Number? An Executive's Guide to Using Probabilistic Thinking to Manage Risk and to Make Better Decisions by Pat Leach. If you care at all about the topic in this entry, buy the book. In fact, buy a few and give them to your bosses and colleagues. Pat explains in straightforward terms why probabilistic approaches are better. (Disclaimer: Pat and I worked together at Texaco. Chevron does a lot of business with his current employer, Decision Strategies.)

The other is The Strategy Paradox: Why committing to success leads to failure (and what to do about it) by Michael Raynor. In particular, chapter 5, "The Limits of Forecasting" is really good. Other parts of the book didn't quite work for me, but that chapter should be required reading for anyone in business.

No comments: