Measure behavior, not attendance.

Ask a stalling program how its people were prepared for the change and you will often be handed a completion report. Ninety-four percent of staff completed the training. It looks like evidence of readiness. It is evidence of attendance, and attendance and readiness are not the same thing.

Training completion measures that a person was present, or at least that a module registered as finished. It does not measure whether they can do the new work, whether they will do it when no one is watching, or whether anything about how they operate has actually changed. A program that mistakes the first for the second has measured its own activity and called it progress.

The comfort of the completion number

Completion metrics are seductive because they are clean, high, and easy to produce. They turn the messy question of readiness into a crisp percentage that looks good on a status slide. They let a program report that preparation is done and move on.

But the number answers a question no one should care about. Nobody wants people who completed the training. They want people who work the new way. The completion report is silent on that, and its silence is comfortable, because the alternative, measuring whether behavior actually changed, is harder and often less flattering.

A ninety-four percent completion rate tells you people attended. It tells you nothing about whether a single one of them now works differently.

What behavior measurement looks like

Measuring behavior means looking at what people actually do after the training, in the flow of real work. Are they using the new process on live cases, or quietly reverting to the old one once the classroom closes? When a real situation arises, do they apply the new approach or the habit they had before? Can they perform the task without the expert sitting beside them?

These are observable. They require more effort than pulling a completion report, watching real work, sampling real cases, asking managers what they actually see their teams doing, but they measure the thing that matters. A program that instruments behavior knows whether its people are ready. A program that instruments attendance only knows they showed up.

Readiness is desire and ability

Prosci's model is useful here because it separates the components that completion collapses. Readiness is not one thing; it is at least desire and ability, the willingness to work the new way and the capability to do it. Training, at best, touches ability, and often only its surface. It does almost nothing for desire, which comes from understanding why the change matters and believing it will hold.

A completion report cannot see either component clearly. A person can complete every module and still lack the desire to change and the genuine ability to perform. Measuring behavior catches what completion misses, because behavior is where desire and ability either combine into action or fail to.

The cost of the false signal

The real damage of the attendance metric is that it gives false confidence at exactly the wrong moment. A program sees high completion, concludes its people are ready, and proceeds to go-live and cutover on that basis. Then adoption comes in low, and everyone is surprised, because the readiness number said the people were prepared.

They were not prepared. They were present. The program measured the wrong thing and made irreversible decisions on the strength of it. The surprise at go-live was engineered months earlier, the moment attendance was accepted as a proxy for readiness.

Watch the work, not the roster

When you plan how to prepare people for a change, plan how you will know it worked, and make sure that measure looks at behavior in real work rather than attendance in a classroom. Completion reports have their place as a hygiene check. They are not evidence of readiness, and treating them as such is one of the quiet, common ways a well-run program walks confidently into a stall.