Measuring whether your crawl fix actually worked

People make crawl changes and declare victory the same afternoon. Crawling moves on a slow clock. Monitor it properly:

— Record a baseline BEFORE changes: total requests, 200 share, non-200 share, avg response time, junk-URL hit %.
— Make ONE category of change at a time, or you'll never know what moved the needle.
— Wait at least 2–4 weeks. Crawl Stats lags and Google adjusts gradually.
— Compare the same metrics. The win looks like: junk hits down, 200 share up, money pages crawled more often.
— Cross-check in logs, since Crawl Stats is sampled.

Actually crawl behavior changes over weeks, not hours — Google re-learns your site's patterns slowly.

If you changed five things and waited one day, you measured nothing. Baseline, one change, wait, compare.