{"id":"res-sb-customer-survey-huddle","audience":"small-business","category":"research","technique":"grounding","title":"Share three customer survey findings with counts at a staff huddle","free":false,"task":"You run a family restaurant and want to share the customer survey at tomorrow's staff huddle. You want at most 3 bullets in plain text, each one finding with its count out of the total (like \"42 of 118\"), from the final tally only. Leave out single comments, and don't add advice.","materialLabel":"The survey tally and comments (attached to your prompt automatically)","material":"CUSTOMER SURVEY - Rosa's Kitchen (comment cards + QR survey, Oct 2026)\nWeek 1 partial tally (Oct 7): 40 responses. \"Wait too long\" 22.\n\nFINAL TALLY (Oct 31): 118 responses\n- Food rated good or great: 101\n- Wait for food felt too long: 47\n- Server was friendly: 96\n- Would like online ordering: 39\n- Parking was hard to find: 18\n\nSelected comments:\n- \"Best enchiladas in town.\"\n- \"Waited 40 minutes for tacos on a Friday.\"\n- \"The new server with the nose ring was rude.\" (only comment about this)\n- \"Please add online ordering!\"","rules":[{"type":"maxBullets","max":3},{"type":"plainText"}],"mistakes":[{"title":"Using the partial tally","why":"The week 1 numbers come first and look dramatic (22 of 40).","fix":"Say: \"Use only the final Oct 31 tally.\"","checks":["r0"]},{"title":"Repeating a single comment","why":"One pointed comment about a server can sound like a pattern, and it singles out a staff member.","fix":"Add: \"Leave out single comments.\"","checks":["r1"]},{"title":"No counts","why":"The AI writes 'most customers' instead of numbers staff can trust.","fix":"Show it: \"Give each as a count, like 47 of 118.\"","checks":["r1"]},{"title":"Getting a pep talk","why":"The AI adds action items and encouragement, which you'll say yourself.","fix":"Say: \"Only the bullets, no advice.\"","checks":["r2"]},{"title":"Not capping the list","why":"The AI lists everything it finds, so the few points that matter get buried.","fix":"Say: \"at most 3 bullets.\"","checks":["maxBullets"]},{"title":"Getting markdown back","why":"AI defaults to bold labels and headings, which show up as stray symbols in the huddle notes you'll read from.","fix":"Ask for \"plain text, no bold, no headings.\"","checks":["plainText"]}],"promptHints":[{"label":"Uses the final tally","patterns":["final","oct(ober)?\\.? 31","118","not the (partial|week 1)"]},{"label":"Asks for counts","patterns":["count","of 118","out of","numbers?"]},{"label":"Skips single comments","patterns":["single comment","one.?off","only one person","skip.{0,20}comment"]},{"label":"Caps the list","patterns":["\\b(3|three)\\b.{0,15}bullet","at most"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no headings"]}],"hint":"Staff take feedback better with real numbers and no finger-pointing. Tell the AI which tally to use, to show counts, and to leave out one-off comments.","rubricLabels":["Every count comes from the final tally of 118 (for example 101 of 118, 47 of 118, 96 of 118, 39 of 118), not the week 1 partial tally","Each bullet gives its finding as a count out of 118, and no bullet is based on a single comment (such as the one about a rude server)","Contains only the bullets: no intro, no advice for staff, no closing pep talk"]}