Why "Scientific Forecasting" Flops

Last Updated Aug 2, 2007 9:04 AM EDT

Sales forecasting in most firms is a mess. The alternative -- scientific forecasting with a computerized market model -- is a proven solution. Why then is scientific forecasting so rare? The answer is fear, on the part of everyone involved.
  • Sales reps are worried that if forecasting is generated from a computer model, they'll loose what little control they have over setting their sales quotas.
  • Sales managers are afraid that if they can't tweak the forecast, they'll lose the ability to use forecasting to manipulate the behavior of their reps.
  • The marketing group is terrified that scientific forecasting will reveal the fact that their marketing efforts have almost no impact on actual sales.
  • The manufacturing group and the CFO from long experience doesn't trust sales forecasts and thus is afraid that anything that involves input from the sales group will automatically be bogus.
  • The CEO is deathly afraid that scientific forecasting might make it impossible to temporarily goose up the stock price by presenting bold forecasts to the press and investors.
In short, the forecasting dance has become part of the culture of how most corporations work and the idea of replacing it with some sort of propeller-head hoodoo is simply too painful to contemplate. And that's sad, because being able to actually predict what will sell in a fiscal quarter would, in most firms, increase profitability and create the kind of predictability that investors like.

Instead, forecasting in most firms remains a painful process that involves complicated negotiations, political maneuvers, management bluster and, when it comes time to report sales figures to the press, plenty of handwringing, scapegoating and (worse) "creative" accounting.

Unfortunately, moving to a scientific sales forecasting model, in most companies, would involve a top-to-bottom change in the behavior of nearly every organization impacted by the change. And those kind of changes take time - if they ever happen at all. I'm still waiting for the paperless office.

Next... how make forecasting less painful, without hiring a mathematician.