{"id":"data2-ab-test-faq-stakeholder-questions","audience":"product-managers","category":"data","track":"data","technique":"audience","level":"intermediate","title":"Answer stakeholder questions about an unfinished A/B test","free":false,"task":"A stakeholder is asking why you won't call the pricing-page test after 3 days. Write a Q&A with exactly 3 questions and answers, each as 'Q:' then 'A:' on the next line, plain text, under 130 words. Each answer must be no more than 2 sentences, in plain words with no stats jargon, and use the numbers in the notes.","materialLabel":"Test status and the stakeholder's questions (attached to your prompt automatically)","material":"Pricing page test - status after day 3 (Harvest Meal Kits)\n\nPlan (set before launch): run 14 days, covering two full weekends, with at least 5,000 visitors per version.\n\nDay 3 numbers:\n- Version A (current): 1,300 visitors, 52 sign-ups = 4.0%\n- Version B (new layout): 1,280 visitors, 64 sign-ups = 5.0%\n- Total visitors so far: 2,580 of the 10,000 planned (5,000 per version).\n\nWhy we wait:\n- The first 3 days were Tuesday to Thursday; weekend shoppers behave differently.\n- A gap of 1 point on 1,300 visitors is about 12 extra sign-ups, which can easily be luck at this size.\n- Early leads often shrink as more visitors arrive.\n\nStakeholder questions (from VP Elena Ruiz):\n1. B is up 25% already. Why can't we just ship it?\n2. When will we know?\n3. Can we show B to everyone now and keep measuring?\n\nDates: test started Tuesday Nov 4; the 14 days end Monday Nov 17.","rules":[{"type":"plainText"},{"type":"maxWords","max":130}],"mistakes":[{"title":"Caving to the lead","why":"The AI agrees B is winning and says shipping looks fine.","fix":"Add: \"Explain why 3 days is too early: about 12 extra sign-ups, no weekend yet.\"","checks":["r0"]},{"title":"Missing the end date","why":"The AI says 'a few more days' instead of the real date.","fix":"Add: \"Answer 2 uses Monday Nov 17 and the 5,000-per-version goal.\"","checks":["r1"]},{"title":"Piling on jargon","why":"The AI explains p-values and statistical power to the VP.","fix":"Add: \"Plain words only, no stats terms, at most 2 sentences per answer.\"","checks":["r2"]},{"title":"Leaving the format to chance","why":"The AI picks its own length and layout unless told the exact shape.","fix":"Add: \"Plain text, no bold or headings.\"","checks":["plainText"]},{"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 part instead, for example \"each line under 15 words.\"","checks":["maxWords"]}],"promptHints":[{"label":"Names the reader","patterns":["elena","vp","stakeholder"]},{"label":"Asks for Q and A pairs","patterns":["q:","a:","exactly 3"]},{"label":"Uses the numbers","patterns":["4\\.0%","5\\.0%","12 extra","2,?580"]},{"label":"Gives the end date","patterns":["nov 17","november 17","monday"]},{"label":"Rules out jargon","patterns":["no (stats|jargon)","p-value","plain words"]}],"hint":"Answering a pushy stakeholder works when you keep it to the numbers on the page and one short reason each.","mission":"Answer these questions about the A/B test.","musts":["Three Q and A pairs","Max 2 sentences each","No stats jargon"],"rubricLabels":["Answer 1 says 4.0% versus 5.0% after 3 days is too early to trust (about 12 extra sign-ups, only 2,580 of 10,000 planned visitors, no weekend yet)","Answer 2 gives the date Monday Nov 17 (14 days, two full weekends, 5,000 visitors per version)","Answer 3 says that showing B to everyone would leave nothing to compare against (no group still seeing A), and every answer is two sentences or fewer with no terms like p-value, significance or confidence"]}