PDGMSTalk to us

How to measure if a supervisor is actually making good calls

You measure decision quality by looking at two things together: whether the supervisor knows the right move, and whether their actual behaviour on the floor matches it. The gap between those two is the real measure. Most factories track neither, so they judge supervisors on output and personality instead of on the calls they make.

It feels hard to measure because a decision seems invisible. It happens in someone's head and then the moment passes. But the decision leaves a trail. The escalation that did or did not happen, the rework that followed, the time the line waited. You measure the call by what it produced.

The two numbers that matter

A good call has a known-right answer and a done-right action. So you measure both.

  • The recognition ceiling: does the supervisor know the right action when you describe the situation? You can check this directly with real scenarios from your floor.
  • The behavior floor: what does their actual work data show they did when that situation came up for real?

The distance between the two is the Diagnostic Gap Model, and it is the most honest measure of decision quality you have. The hub guide on why supervisors escalate defines it in full.

Why output and reviews mislead you

Output rewards the supervisor who happens to run an easy line and punishes the one who runs a hard one, regardless of their calls. Annual reviews capture impressions, not decisions. Neither tells you whether this person makes the right call when the situation is unfamiliar, which is the only thing you actually need to know.

What to track instead

  1. Escalation rate, by type. How often does a supervisor send a call up that they could have made? Falling rate means rising judgment.
  2. Rework traced to decisions. Of the rework on a line, how much came from a wrong call rather than a machine fault? This is your money measure.
  3. Decision speed. How long between a situation arising and the right call being made on the floor.
  4. The recognition-to-behaviour gap, per person. Where each supervisor knows the move but does not make it, versus does not yet know it.

Track these over time, per supervisor. The trend tells you who is improving and where to put help.

How you will know your measure is right

A good measurement system points you at a specific person and a specific call, and tells you whether to build knowledge or build the habit of acting on it. If your measure only gives you a score and no next step, it is not measuring decisions. It is measuring vibes.

Where this fits

This is one piece of building decision quality on your floor. The full picture is in the hub guide on why supervisors escalate and how to fix it. PDGMS HCD is the system that builds and proves it, and the measurement above is exactly what it tracks.

Related questions