{"id":"dat-mkt-landing-test-min-sample","audience":"marketing","category":"decisions","technique":"context","title":"Pick a landing page winner while respecting a minimum sample","free":false,"task":"You're picking which landing page to ship. Your team needs at least 1,000 visitors before calling a variant, and conversion = sign-ups ÷ unique visitors. Write exactly 3 numbered lines, one per variant, each as Variant: sign-ups ÷ visitors = X% (or not enough data), then one unnumbered last line: Ship: [variant]. Plain text, under 60 words.","materialLabel":"Landing page test results (attached to your prompt automatically)","material":"LANDING PAGE TEST - \"Free trial\" page (Oct 27 - Nov 5)\n\nVariant | Unique visitors | Sign-ups | Notes\nA (current) | 2,000 | 120 | control\nB (short form) | 2,100 | 147 | name + email only\nC (video hero) | 900 | 90 | added Nov 2\n\nNotes:\n- Design loves C: \"10% conversion! Ship it!\"\n- Page views (not unique) were about 30% higher than unique visitors for all three.\n- Paid traffic was split evenly across variants from Nov 2 on.\n- Dev can ship whichever we pick by Friday.\n- Next test idea: testimonial carousel.\n- Legal approved all three pages.\n- Sign-ups are free trial starts.\n- Test tool is SplitLane.\n- Shared with the team in the usual folder.","rules":[{"type":"numberedItems","count":3},{"type":"plainText"},{"type":"maxWords","max":60}],"mistakes":[{"title":"Not giving a minimum sample","why":"The AI crowns C on 900 visitors, like Design.","fix":"Add: \"Need at least 1,000 visitors to call a variant.\"","checks":["r1","r2"]},{"title":"Not defining conversion","why":"The AI may use page views instead of unique visitors.","fix":"Add: \"Conversion = sign-ups ÷ unique visitors.\"","checks":["r0"]},{"title":"Not asking to see the work","why":"Bare rates hide the sample size.","fix":"Add a line to copy: \"1. A: 120 ÷ 2,000 = 6%\"","checks":["numberedItems","r0"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total word limit and it often lands over it.","fix":"Cap each line: \"each line under 10 words.\"","checks":["maxWords"]},{"title":"Forgetting where it goes","why":"AI loves markdown tables and bold, which show up as clutter in a marketing channel.","fix":"Add: \"Plain text, no bold, headings or tables.\"","checks":["plainText"]}],"promptHints":[{"label":"Gives a minimum sample","patterns":["1,?000","minimum|at least|enough data"]},{"label":"Defines conversion","patterns":["unique visitors","sign.?ups ÷"]},{"label":"Asks to show the math","patterns":["show (the |your )?(math|work)","x \\$|\\bx\\b.{0,10}=","calculation","step"]},{"label":"Sets a word limit","patterns":["\\d+\\s*words","word (limit|count)","under \\d+"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no (headings|tables)"]}],"hint":"A small sample can look like a big win. Give the AI your minimum and how conversion is measured.","rubricLabels":["A is 120 ÷ 2,000 = 6% and B is 147 ÷ 2,100 = 7%","C is marked not enough data (900 visitors, under 1,000), not called a 10% winner","The last line ships B, with nothing before the first numbered line"]}