Direct Marketing Magazine - December 1996

Most “initial fill” rates are computed as the ratio:

Thus, stockouts count only if the customer takes a backorder. By this reasoning, substitutions are just as good as having the item the customer wanted, and soldouts or customers who say “forget it” are simply ignored. They don’t count at all.

As many managers recognize, this formula makes service levels look better than they really are. But few companies appear to have quantified the distortion, which often is substantial. The nearby chart illustrates one case, in which about 30% of customers who encountered stockouts bought a substitute item or said they would look elsewhere.

Click on the chart to enlarge it.

As the chart shows, the “initial fill” rate has another, insidious property. As service level gets worse and worse, the “initial fill” report masks the problem even more.

A rule of thumb I have used for gauging true service level is to take the “initial fill” number, subtract it from 100 percent to get the supposed fraction of unserviced customers. Increase that fraction by two-thirds. Thus, for an 85% “initial fill” the implied frac-tion of unserviced customers would be 15%. Increasing that by two-thirds yields 25% of customers encountering stockouts, for a true service level of about 75%.