{"id":"res-sales-case-study-numbers-check","audience":"sales","category":"fact-checking","technique":"verification","title":"Verify a customer case study draft against the customer's data","free":false,"task":"Marketing drafted a case study about your customer Pell Street Clinics, and you must confirm the numbers before the customer approves it. You want only the numbers the customer's data contradicts, as numbered lines in the form \"Draft says X; data says Y\", under 70 words total, plain text. Skip numbers that match and claims the data doesn't cover, and don't rewrite the draft.","materialLabel":"The customer's data summary and the case study draft (attached to your prompt automatically)","material":"CUSTOMER DATA SUMMARY - Pell Street Clinics (from their ops manager, Oct 2026)\nLocations live on our booking platform: 9\nMonths live: 14\nNo-show rate before: 18% / after: 11%\nOnline bookings as share of all bookings: 64%\nFront-desk phone calls per week: before 2,100 / after 1,450\nStaff trained: 85\n\n-----\nCASE STUDY DRAFT - \"How Pell Street Clinics Cut No-Shows\"\nPell Street Clinics has run our booking platform across all 9 of its locations for 14 months. Since going live, no-shows have dropped from 18% to 8%. Nearly two thirds of appointments are now booked online. Front-desk phone calls fell by more than half, freeing staff for patients in the waiting room. More than 100 staff members were trained in the first month. \"Our patients love it,\" said operations manager Dana Wells.","rules":[{"type":"maxWords","max":70},{"type":"plainText"}],"mistakes":[{"title":"Asking for edits","why":"The AI polishes the story and never checks the numbers.","fix":"Say: \"List only numbers the data contradicts.\"","checks":["r2"]},{"title":"Not doing the math","why":"\"More than half\" sounds plausible unless you work out 2,100 to 1,450.","fix":"Add: \"Check percentages and words like 'half' against the raw numbers.\"","checks":["r0"]},{"title":"Flagging everything","why":"The AI questions the quote and matching numbers to look thorough.","fix":"Say: \"Skip numbers that match and claims the data doesn't cover.\"","checks":["r1"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total word limit and it often lands 10-20% over.","fix":"Cap each line: \"each line under 15 words.\"","checks":["maxWords"]},{"title":"Getting markdown back","why":"AI defaults to bold labels and headings, which show up as stray symbols in an email to marketing.","fix":"Ask for \"plain text, no bold, no headings.\"","checks":["plainText"]}],"promptHints":[{"label":"Asks for contradictions only","patterns":["contradict","(don't|do not|doesn't) match","only.{0,20}(wrong|errors)"]},{"label":"Gives the line format","patterns":["draft says","data says","\\bX\\b.{0,20}\\bY\\b"]},{"label":"Asks to check the math","patterns":["math","calculat","percent","half"]},{"label":"Sets a length limit","patterns":["\\d+\\s*words","one (short )?sentence","word (limit|count)"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no headings"]}],"hint":"Case studies blur numbers into phrases like 'more than half'. Tell the AI to check phrases against raw numbers and list only what the data contradicts.","rubricLabels":["Flags all three contradictions: no-shows after (draft: 8%; data: 11%), phone calls (draft: fell by more than half; data: 2,100 to 1,450, about 31%), and staff trained (draft: more than 100; data: 85)","Does not flag numbers that match (9 locations, 14 months, 18% before, nearly two thirds online) or claims the data doesn't cover (the quote, 'in the first month')","Outputs only numbered lines in the form 'Draft says X; data says Y': no intro, no rewritten draft, no closing note"]}