{"id":"fpa-cohort-retention-percent-table","audience":"fpa","category":"finance-work","track":"finance","technique":"format","level":"intermediate","title":"Build a retention table from cohort counts with a corrected start","free":false,"task":"You're building a retention view for the customer success lead. Write a prompt for a markdown table with columns Cohort, Start, Q+1, Q+2 and Q+3 showing retention in each Q+ cell as only a whole-number percent like 90% (no counts), and the Start column as a count (use a dash \"-\" where the quarter has not happened yet), followed by one last line starting \"Weakest Q+1:\" naming the cohort and its percent. Use the corrected Q2 start. Under 90 words.","materialLabel":"Source material (attached to your prompt automatically)","material":"CUSTOMER COHORTS - customers still active at the end of each quarter after signup (as of Sep 30, 2026)\n\nQ1 2026 cohort: start 200, Q+1 180, Q+2 160, Q+3 150\nQ2 2026 cohort: start 240, Q+1 215, Q+2 190\nQ3 2026 cohort: start 300, Q+1 270\n\nCorrection from RevOps, Oct 8: the Q2 cohort start was understated by 10 because a reseller batch signed on Jun 28 was loaded late. The Q2 start is 250, not 240. The Q+1 and Q+2 counts already include these customers.\n\nDefinitions: a customer is active if it has a paid invoice in the last 60 days. Q+1 means the end of the first full quarter after the signup quarter. Retention % is active customers divided by the cohort's start.\n\nThe customer success lead will present this in the Nov 12 review and wants to see which cohort is slipping early.","rules":[{"type":"maxWords","max":90}],"promptHints":[{"label":"Names the columns","patterns":["cohort","start","q\\+1"]},{"label":"Uses the correction","patterns":["250","correct"]},{"label":"Handles missing quarters","patterns":["dash","not happened","blank"]},{"label":"Asks for the last line","patterns":["weakest"]},{"label":"Caps length","patterns":["\\d+\\s*words","under \\d+"]}],"mistakes":[{"title":"Using the old Q2 start","why":"The AI divides by 240 and gets 90% and 79%.","fix":"Say: \"Use the corrected Q2 start of 250.\"","checks":["r1","r2"]},{"title":"Filling the future quarters","why":"The AI invents numbers for quarters that haven't happened.","fix":"Add: \"Use a dash where the quarter has not happened.\"","checks":["r0"]},{"title":"Missing the closing line","why":"The AI ends the table with commentary.","fix":"Add: \"Last line: Weakest Q+1: cohort and percent.\"","checks":["r2"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting and runs past word limits.","fix":"Ask for fewer words than the real cap and keep each part short.","checks":["maxWords"]}],"hint":"Corrections hide in the notes. Tell the AI to use the corrected start and how to show quarters that haven't happened.","rubricLabels":["The header is Cohort, Start, Q+1, Q+2, Q+3 and the cells show only percents: the Q1 row is 200, 90%, 80%, 75% and the Q3 row is 300, 90%, -, -","The Q2 row uses the corrected start of 250 and shows 86% for Q+1 and 76% for Q+2","The last line starts \"Weakest Q+1:\" and names the Q2 cohort at 86%"]}