The audit before the audit

Before you score anything, you need to know whether the company can survive being scored. A short protocol for finding out.

Every measurement programme has a political precondition: somebody has to be willing to publish a number that makes them look bad in quarter one.

If that person does not exist, the audit will happen, the deck will be delivered, and the numbers will quietly be renegotiated until they flatter the person who commissioned them. This is not a measurement failure. It is a sponsorship failure, and it is detectable in advance.

The protocol is three questions, asked separately, of three different people.

First: what would we do differently if this score came back low? An answer that describes a communications response rather than an operational one is a warning.

Second: who sees the raw number before it is presented? If the answer is more than two people, the number will be smoothed.

Third: what is the last uncomfortable metric this company published internally? A company with no track record of that has no muscle for this.