The Lean manufacturing, PPT Plus,
"Back-to-Basics" e-Training Package 
Lean Manufacturing Solutions

Increase the effectiveness of your
Lean Manufacturing Training Program
Lean Manufacturing Simulation Game 

There Is No 'Magic Model'

The elusive "Magic Model"... Many have searched for that perfect Mathematical equation that could just be plugged right into their business application, left alone to do its job, and just keep churning out great forecasts time after time after time. It does sound alluring.

Unfortunately, many widely used models make a fatal erroneous assumption—that the future will be much like the past. That premise quickly unravels when, in the real world, the future is almost guaranteed to never look like a carbon copy of history. And it sure doesn't look like the examples in the textbooks. Promotions are rarely the same dates, for the same products, at the same discount, or under the same economic conditions.

Also consider the fact that many forecasts fall short of desired accuracy not because the annual quantity was off, but because the timing was.

Relying on the model or system alone will not help your forecast accuracy at all, but it does give you a built-in excuse. When the forecast goes haywire, you can always blame the model. Or point out that another system should have been purchased. Neither of these actions help any­thing at all, but there are things that you can do, regardless of the system or method you use. But it will take time, understanding, and patience.

This Method Isn't a 'Quick Fix'

Using these methods probably won't save you any time in forecast preparation. It certainly isn't as quick as letting a model alone to do the work with little or no human intervention. But it may ultimately save you a lot of time when you stop to consider how much time you now spend researching why the forecast was off—and then explaining it to management. I promise you that if the accuracy of the forecast is within just a couple points to actual, you probably won't have to spend as much time explaining.

This approach is low-tech, doesn't rely on a formula as long as your arm, and honestly, has been rejected by many because "it isn't scientific enough." Unfortunately, the mindset of many who consider that forecasting is strictly statistical has been set for a very long time, indeed.

Instead of scientific theory, how about some results? As I was preparing to write this paper, I received the March, 1994, and first quarter results. At the aggregate level for March, the forecast was within ONE-TENTH of ONE PERCENT, and the first quarter total miss was 7600 units on 3.5 million. At the individual product group level, with forecast accuracy measured 90 days in advance of the actual demand, the forecast accuracy on the main product groups (individually) was within two to five percent. Honest.

To be Continued