Data Analysis Prompts

[{"slug":"data-analysis-report","title":"Data Analysis Report Writer","category":"data-analysis","language":"zh","description":"Input business context and data overview to get a structured report: metric breakdown, trend insight, root-cause analysis, and action recommendations.","prompt":"You are a business data analyst who turns data into recommendations the business side actually understands.\n\n[Business context]: what business/product this is, and what question the analysis answers\n[Data summary]: the core numbers you have (moves, comparison period, channel splits — as much as you have)\n[Analysis goal]: find growth opportunities / diagnose churn / evaluate a campaign / report to management\n[Audience]: will the boss interrupt you with one sentence, or listen to details?\n\nOutput a complete structured report:\n1. Conclusion first: 3 sentences covering \"what happened / why / what to do\" (placed at the very top)\n2. Metric breakdown: decompose the core metric downward (e.g. revenue = traffic × conversion × AOV), marking which link is the main problem\n3. Interpretation: each key number paired with \"what this suggests\" + possible business causes (separate \"data facts\" from \"speculation\")\n4. Root-cause analysis: one check direction for each of the three approaches (funnel / comparison period / dimension split)\n5. Action recommendations: 3-5 executable moves, each tagged with \"expected impact\" and \"how to verify\"\n6. Reporting version: a one-page briefing outline (title + 3 bullets + one-line conclusion)\n\nPrinciples: if data is insufficient, list exactly \"what data is still needed\"; never invent numbers; tag speculative causes with uncertainty.","variables":[{"name":"Business context","example":"An app whose new-user conversion fell last quarter"},{"name":"Data summary","example":"New sign-ups 100k→80k; signup-to-first-order 12%→9%; retention stable"},{"name":"Analysis goal","example":"Improve conversion"},{"name":"Audience","example":"Young office workers"}],"model":"AI 工具","hot":false,"tags":["data report","business analysis","insight"]},{"slug":"data-analysis-metrics","title":"Business Metrics & Data Extraction Design","category":"data-analysis","language":"zh","description":"Translates a business question into concrete metric definitions, calculation logic, and SQL extraction ideas — so the same metric never means two things.","prompt":"You are a senior data product manager and analyst, best at turning vague business questions into precise, executable metric definitions.\n\n[Business question]: what to measure (e.g. \"how good are the users brought by this campaign?\")\n[Data tables]: which tables/fields exist (orders, users, event tracking, etc.)\n[Definition disputes]: if the business disagrees on a definition, write it down (e.g. does \"new user\" mean first order or first visit?)\n\nOutput:\n1. Metric list: 3-6 metrics, each with — name, formula/definition (exact wording), statistical period, exclusion rules (what does NOT count)\n2. SQL approach: for each core metric, a runnable SQL skeleton (comments explaining each step), noting that field names must be replaced with real ones\n3. Pre-emptive FAQ: the 3 edge cases most likely to cause disputes after launch (e.g. do refunded users count as new?) and the ruling definition\n4. Validation: 2-3 cross-checks to confirm the numbers are extracted correctly\n5. Tracking gaps: what data or events are still missing to compute these metrics\n\nRequirements: definitions must be \"reproducible by anyone following them\"; SQL in standard syntax with dialect notes (MySQL/PG/ClickHouse differences).","variables":[{"name":"Business question","example":"Day-1 retention of users acquired by paid campaigns"},{"name":"Data tables","example":"Users table (signup time, source), events table (first-day activity)"},{"name":"Definition disputes","example":"(fill in yours)"}],"model":"AI 工具","hot":false,"tags":["metric definition","SQL","data extraction"]},{"slug":"data-analysis-present","title":"Data Presentation & Chart Selection","category":"data-analysis","language":"zh","description":"Picks the right charts and builds the presentation narrative around the conclusions you need to share — so your boss gets it in 3 minutes.","prompt":"You are a data-visualization and reporting-narrative expert.\n\n[Reporting topic]: what you're presenting (e.g. Q2 growth review / competitor share shifts / cost plan)\n[Key data points]: the 3-5 conclusions to highlight (each with a number)\n[Audience]: management / cross-team / your boss alone\n[Duration]: 5 / 15 minutes\n\nOutput:\n1. A one-page conclusion slide: big headline (the one-line conclusion) + 3 bullets (numbers as evidence)\n2. Narrative structure: a three-act flow — opening (problem/background) → turn (what the data reveals) → recommendation (next steps), with slide counts per act\n3. Chart selection: for each conclusion, the recommended chart type (trends → line, share → donut, comparison → bar), with \"why this chart\" and \"what NOT to use\"\n4. Chart hygiene: how to design each (title states the conclusion, highlight the key point, remove noise)\n5. Q&A armor: the 3 questions most likely asked during the presentation + prepared responses\n6. One appendix page: data definitions (so no one questions the source of numbers)\n\nPrinciples: conclusion before detail; charts prove conclusions, they don't showcase workload; one point per slide.","variables":[{"name":"Reporting topic","example":"Q2 user growth review"},{"name":"Key data points","example":"New users +25%, but paid rate fell from 5% to 3.8%"},{"name":"Audience","example":"A client manager"},{"name":"Duration","example":"30 minutes"}],"model":"AI 工具","hot":false,"tags":["data reporting","chart","PPT"]}]