data•5 prompts
Data Analysis AI Prompts
Data analysis is about asking the right questions, not just running queries. These prompts help you extract meaningful insights and communicate findings clearly.
1. Dataset Explorer
Best: ChatGPTI have a dataset with these columns: [list columns and types]. Sample rows: [paste 5-10 rows]. Tell me: what questions this data can answer, potential data quality issues you notice, which columns are likely correlated, recommended analysis approaches, and 3 non-obvious insights I should look for.
Include data types and sample values. ChatGPT is good at suggesting analytical approaches from data structure.
2. Statistical Test Selector
Best: ClaudeI want to test whether [hypothesis]. My data: [describe data - sample size, distribution, type of variables]. Help me choose the right statistical test by explaining: which test to use and why, the assumptions I need to check, how to interpret the results, and what p-value/effect size would be meaningful in this context.
Be specific about your sample size and data distribution. Claude explains statistical reasoning clearly.
3. Dashboard Specification
Best: ChatGPTDesign a dashboard for [audience - e.g., "marketing team"] that answers: [list 3-5 key questions]. For each metric, specify: the visualization type and why, the data source/calculation, comparison benchmarks, and what action each metric should trigger. Limit to 6-8 visualizations max.
Specify the audience and their decision-making context. Fewer metrics with clear actions beat comprehensive metrics.
4. Anomaly Investigation
Best: ClaudeI noticed [describe anomaly - e.g., "sales dropped 30% last Tuesday"]. Data context: [relevant details]. Help me investigate by: listing the most likely causes ranked by probability, suggesting what data to check for each cause, identifying what would confirm or rule out each hypothesis, and recommending an analysis workflow.
Include as much context as possible - seasonal patterns, recent changes, external events.
5. Executive Summary from Data
Best: ClaudeTurn these findings into a 1-page executive summary for [audience]: [paste raw findings/numbers]. Structure: headline finding (1 sentence), key metrics with context (what changed and why it matters), implications for business decisions, and recommended next steps. Use plain language, no jargon.
Specify who the audience is and what decisions they need to make. Claude is strong at translating data into business language.
Try These Prompts with Multiple AIs
Get responses from Claude, ChatGPT, Gemini, and more—all at once. Compare which AI handles these prompts best.
Try Council Free