{"id":"dat-pm-cohort-retention-import","audience":"product-managers","category":"data","technique":"context","title":"Compare cohort retention after removing a bulk import","free":false,"task":"You're comparing month-1 retention for the Aug, Sep and Oct signup cohorts. The Sep cohort includes 200 accounts from a partner bulk import that never logged in; remove them (the export doesn't flag them). Write exactly 3 numbered lines, each as Cohort: active in month 2 ÷ signups = X%, then one unnumbered last line: Best: [cohort]. Plain text, under 60 words.","materialLabel":"Cohort export (attached to your prompt automatically)","material":"MONTH-1 RETENTION BY SIGNUP COHORT (from Trackwise)\n\nCohort | Signups | Active in month 2\nAug | 1,000 | 400\nSep | 1,400 | 540\nOct | 1,100 | 418\n\nNotes:\n- Sep: new onboarding checklist launched Sep 3.\n- Oct had a 2-day login outage (Oct 14-15).\n- Partner (Brightline Schools) sent a bulk account import in September.\n- Dashboard shows Sep at 39% and calls the checklist \"a flop.\"\n- Growth team wants to roll the checklist out to mobile in Q1.\n- Trackwise counts \"active\" as at least one session.\n- A cohort is everyone who signed up in that calendar month.\n- Month-1 retention = still active in the month after signup.\n- Retention review with the growth team is Tuesday.\n- Brightline is our biggest school partner.","rules":[{"type":"numberedItems","count":3},{"type":"plainText"},{"type":"maxWords","max":60}],"mistakes":[{"title":"Not mentioning the import","why":"The export doesn't flag those accounts, so Sep looks worst.","fix":"Add: \"Remove 200 imported accounts from Sep's signups.\"","checks":["r0","r2"]},{"title":"Not asking to see the work","why":"Bare rates hide what was removed.","fix":"Add a line to copy: \"2. Sep: 540 ÷ 1,200 = 45%\"","checks":["numberedItems","r1"]},{"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 product review doc.","fix":"Add: \"Plain text, no bold, headings or tables.\"","checks":["plainText"]}],"promptHints":[{"label":"Removes the import","patterns":["import","200","never logged in"]},{"label":"Asks to show the math","patterns":["show (the |your )?(math|work)","x \\$|\\bx\\b.{0,10}=","calculation","step"]},{"label":"Asks for the best cohort","patterns":["best","highest"]},{"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":"One cohort has accounts that were never real users. Tell the AI to remove them and show each rate.","rubricLabels":["Sep is 540 ÷ 1,200 = 45% (200 imported accounts removed), not 39%","Aug is 400 ÷ 1,000 = 40% and Oct is 418 ÷ 1,100 = 38%","The last line names Sep as best, with nothing before the first numbered line"]}