Landing-page demand resolves at 100–200 targeted visitors (at 15% that's ~30 leads); a strict A/B needs ~10,000 visitors per variant and 300+ conversions before the gap between them means anything. Below those limits, "it's working" and "it's noise" are the same picture, and the decision you make from sixty visitors costs more than the entire test budget. The rule worth keeping: scale what converts easily first, and never crown a winner on a sample that can't hold one. That single rule is what the Marketing Mindset skill is built around — one file, reads in a minute, install cost zero: npx skills add axelfreeman/marketing-mindset (page: https://axelfreeman.github.io/marketing-mindset/). What's the smallest sample you've ever made a real decision from?