Data Analysis
Prompts.
Transform raw data into business intelligence. Use AI to explore datasets, identify patterns, test hypotheses, and generate actionable recommendations.
Data Exploration Fundamentals
Prompting for data analysis turns AI into an analytical partner, extracting trends, anomalies, and recommendations from complex datasets via structured prompts. Data analysts, BI professionals, and business leaders use these prompts to interpret CSVs, reports, and visualizations.
Comprehensive Dataset Overview Prompt
Analyze this dataset [UPLOAD CSV or PASTE SUMMARY]: 1. **Structure**: Rows/cols/types/missing values (%) 2. **Distributions**: Skewness/outliers per numeric col 3. **Correlations**: Top 5 pairs (strength/direction) 4. **Segments**: Natural groupings/clusters 5. **Quality Issues**: Priorities + fixes Provide: Summary table + 3 priority questions to ask next
Exploratory Data Analysis (EDA) Prompts
Quick EDA Template
EDA for [DATA DESCRIPTION]: - Key stats: mean/median/mode/IQR per column - Visual suggestions: 3 chart types - First insights: Hypotheses to test - Cleaning plan: Step-by-step - Missing data strategy
Time Series Analysis
Analyze time series: [PASTE DATA or "Monthly sales 2020-2025"] 1. Trends: Growth/seasonality/cycles 2. Anomalies: Dates + magnitudes 3. Forecasts: Next 6 periods (method/confidence) 4. Drivers: Likely correlations 5. Interventions: When/impact? Provide: Line chart description + forecast table
Trend and Pattern Recognition
Correlation Deep Dive
Find strongest relationships in [DATA]: Create table: | Var1 | Var2 | Corr | P-value | Causal plausibility | Analysis: - Visualize top 3 as scatter + regression - Business implications ranked - Confounders to consider
Statistical Inference Prompts
Hypothesis Testing Guide
Test hypothesis: "[HYPOTHESIS]" on [DATA] 1. Statistical test recommended 2. Null/alternative clearly stated 3. P-value + interpretation 4. Effect size/practical significance 5. Conclusion + confidence level
A/B Test Analyzer
Analyze A/B test results: Control: [METRIC=N, MEAN=VAL, SD=VAL] Variant: [SAME] Determine: Winner? CI? Power? Subgroups? Recommendations
Customer Segmentation Prompts
Customer/Product Clustering
Cluster analysis for [DATA: customer metrics]: 1. Natural segments (3-5) 2. Profiles: Traits + size (%) 3. Value differences (LTV/churn) 4. Strategies per segment 5. Validation metrics (silhouette score) Output: Segment comparison matrix
Actionable Insight Generation
Executive Summary Engine
From analysis [PASTE RESULTS], create executive brief: 1. The story (headline insight) 2. Key evidence (3 bullets) 3. Actions (prioritized table: Impact/Effort/Owner) 4. Risks/watch items 5. Next dataset to acquire
KPI Impact Calculator
Quantify business impact: Finding: [INSIGHT] Metrics affected: Revenue/churn/LTV Magnitude: $$ range Leverage points: Experiments to run ROI estimate: Year 1/2/3
Data Analysis FAQ
Data-analysis prompting, FAQ
Can ChatGPT or Claude do data analysis?
Yes. Both can analyze data you paste or upload, cleaning, summarizing, finding trends, running calculations, and explaining results in plain language. With code execution and file tools (ChatGPT's data analysis, Claude with files), they can process CSVs and spreadsheets, build charts, and run statistics. For sensitive data, check your tool's privacy settings first.
What's the best prompt for analyzing a dataset?
Give the AI context and a clear goal: describe the columns, state the question ('Which region grew fastest quarter over quarter?'), and ask for a specific output such as a table, ranked list, or short summary with the key numbers. Ask it to show its steps or the formula used so you can verify, and request caveats about data quality.
How do I avoid wrong numbers in AI data analysis?
Provide the actual data instead of relying on the model's memory, ask it to compute step by step or use its code/data tool rather than estimating, and request the exact figures and method. Spot-check a few results manually, and ask the AI to flag assumptions, missing values, and outliers that could skew the analysis.
Which AI is best for data analysis?
For spreadsheet/CSV work and charts, ChatGPT's data analysis (code execution) and Claude with file uploads are both excellent, while Gemini integrates well with Google Sheets. The best choice depends on where your data lives and whether you need code execution, for most tabular analysis, an AI with a code/data tool gives the most reliable numbers.