The Orchestrator’s Scorecard

How do you measure whether an AI workflow pays off?

The Orchestrator’s Scorecard measures one live AI workflow using five operating numbers: hours before, hours now, fully loaded labor cost, AI running cost, and correction rate. Enter the real numbers to get a red-or-green monthly value verdict. The calculation stays in your browser.

1 · Hours before.
People-hours this workflow consumed per month, before AI.

2 · Hours now.
People-hours per month now — including reviewing and fixing AI output. Count honestly.

3 · Cost per hour.
Fully-loaded hourly cost of the people doing it, in your currency.

4 · AI running cost.
Per month: tokens, licenses, vendors, and the people who babysit it.

5 · Correction rate.
Of the outputs the AI produces, what share needs a human to correct it? (%)

The verdict appears when all five boxes have numbers. If you don’t know a number, that finding is the scorecard working — go measure it first.

A red verdict is not a failed workflow — it’s the first honest measurement it ever had. Diane’s rule: no initiative graduates from pilot until this card is green two months running. Not sure which workflow to score first? Find your stage in 2 minutes — the assessment points at the layer that’s leaking.