The case of a bad incentive

November 17, 2009

A manufacturing company paid sales managers a bonus for forecast accuracy. Great idea, right?

Except the bonus was paid on the accuracy of the total number of cases projected. It didn’t matter what category, or what SKU. If the sales manager said he was going to sell 100,000 cases, it didn’t matter what was in the cases. This makes determining an accurate forecast as easy as mathematically possible. Good for the sales managers, bad for anyone who needed to use the information.

One customer of the forecast was the production planning department, which needed an accurate forecast at least by product group (e.g. a 24ct package vs. a 48ct package of widgets), and ideally by SKU (e.g. 24ct red vs. 24ct blue). Since the sales managers were not expected to be accurate to that degree, the forecast accuracy incentive was essentially worthless to those who actually needed the forecast to be accurate.

The actual desired behavior was not designed into the incentive program.

Ultimately, the forecast accuracy incentive was ignored anyway. Sales managers were paid even more when they sold more. In fact the prevailing attitude in the sales department was not to meet forecast but to beat it. Anyone who did so was celebrated.

Competing incentives – both money and adulation – drowned out the forecast accuracy incentive.

Bad incentive!