Week 6: Evaluate, measure, refine, compound

Growth that is not measured fades. Evaluate turns the first five weeks into a system that keeps compounding, and tells you honestly what worked.

Rescan

Run the S.C.A.L.E. app again on the same page and compare it with your Week 1 scores. In the app, Momentum is the change between your monthly scans, so it appears once you have two. Which scores moved? Which checks now pass?

A tracked set of questions

Pick ten questions real customers ask: your five from Week 1 plus five new ones. Every month, ask each in ChatGPT and Perplexity and record whether the business is named and which sources are cited. Keep the wording identical each month so the results are comparable.

Be honest about timing. AI answers change slowly and vary from one run to the next. Six weeks can move the app's scores; it may not yet move what an AI says. Report what moved and what has not yet, rather than claiming success early.

The monthly routine

1. Rescan on the first of the month.
2. Re-ask the ten questions and log the results.
3. Log new reviews, citations and where inquiries came from.
4. Pick the next fix, then start the cycle again at Scan.

That loop is the point of the fifth stage: each month's Evaluate becomes the next month's Scan, and small improvements compound.

Your capstone

Put it together as one plan for the business: the baseline and the blockers, the positioning and front-page rewrites, the schema, FAQ and llms.txt, the citation map and 90-day plan, and the monthly routine. When you have it, mark this week complete and claim your certificate below.