ChatGPT for Data Analysis

Turn raw data into actionable insights with AI-powered prompts for trend detection, statistical analysis, and executive reporting.

ChatGPT for Data Analysis Overview

ChatGPT accelerates data analysis by turning raw numbers into actionable insights through structured prompts that identify trends, anomalies, and recommendations. Data analysts and marketers use these analytics prompts to interpret datasets, generate visualization ideas, and communicate findings to stakeholders without coding from scratch.

Core Capabilities:

✓ Dataset quality assessment and cleaning recommendations
✓ Anomaly and outlier detection with severity ratings
✓ Correlation analysis and causation checking
✓ Statistical significance evaluation
✓ Visualization type recommendations
✓ Business impact quantification
✓ Actionable recommendation generation

Data Exploration and Cleaning Prompts

Initial Dataset Overview

Analyze this dataset summary: [PASTE METADATA: rows, columns, types, sample values].

Output structured analysis:
1. **Data Quality**: Missing values? Duplicates? Outliers?
2. **Key Distributions**: Skewness patterns, categorical imbalances
3. **Correlation Signals**: Top 3 potential relationships  
4. **Recommended Cleaning**: Specific steps prioritized by impact
5. **Analysis Priority**: Top 3 questions this data can answer

Anomaly Detection

Review these data points for anomalies: [PASTE DATA OR SUMMARY].

Flag:
- Statistical outliers (IQR method)
- Business rule violations 
- Trend breaks
- Seasonal anomalies

For each: Severity (Critical/High/Medium), potential causes, investigation steps.

Statistical Analysis Prompts

Correlation and Causation Checker

Given these correlations: [VARIABLE PAIRS + COEFFS]

Assess:
1. **Statistical significance**: Likely noise vs signal?
2. **Business plausibility**: Makes causal sense?
3. **Confounding variables**: What might explain both?
4. **Actionable experiments**: How to test causality?

Output: Risk-rated recommendations (Safe to act / Needs validation / Likely spurious).

A/B Test Result Interpreter

A/B test results:
Variant A: [METRIC=VALUE, N=SAMPLE_SIZE, CONFIDENCE]
Variant B: [METRIC=VALUE, N=SAMPLE_SIZE, CONFIDENCE]

Analysis:
1. Winner declaration + p-value
2. Practical significance (effect size)
3. Subgroup analysis suggestions
4. Confidence intervals visualization description
5. Implementation recommendation

Visualization Recommendation Engine

Chart Type Selector

Dataset characteristics: [NUMERICAL/CATEGORICAL, DIMENSIONS, TREND/COMPARISON/RANKING GOAL]

Recommend 3 visualization types with:
Chart | Why This Chart | X-Axis | Y-Axis | Color Encoding | Annotations Needed
------|---------------|---------|---------|----------------|------------------
[Row] | [ ] | [ ] | [ ] | [ ] | [ ]

Include accessibility + stakeholder communication considerations.

Actionable Insights Generator

Business Impact Calculator

Key findings: [LIST 3-5 STATISTICALLY SIGNIFICANT RESULTS]

Translate to business impact:
1. **Revenue Impact**: $$ quantification or range
2. **Customer Impact**: Retention/churn/LTV effects  
3. **Operational Levers**: What dials to turn
4. **Quick Wins**: 30-day actions (<$5K budget)
5. **Strategic Implications**: 12-month roadmap shifts

Customer Segmentation Storyteller

Segmentation results: [CLUSTER DESCRIPTIONS + SIZES + KEY TRAITS].

Create customer journey narratives:
Segment 1: "Meet Sarah..." [pain points → current behavior → ideal solution]
Segment 2: etc.

Marketing recommendations per segment (personalization tactics).

Frequently Asked Questions

Analyze this dataset summary: [PASTE METADATA: rows, columns, types, sample values]. Output structured analysis: 1. **Data Quality**: Missing values? Duplicates? Outliers? 2. **Key Distributions**: Skewness patterns, categorical imbalances 3. **Correlation Signals**: Top 3 potential relationships 4. **Recommended Cleaning**: Specific steps prioritized by impact 5. **Analysis Priority**: Top 3 questions this data can answer

Review these data points for anomalies: [PASTE DATA OR SUMMARY]. Flag: - Statistical outliers (IQR method) - Business rule violations - Trend breaks - Seasonal anomalies For each: Severity (Critical/High/Medium), potential causes, investigation steps.

Given these correlations: [VARIABLE PAIRS + COEFFS] Assess: 1. **Statistical significance**: Likely noise vs signal? 2. **Business plausibility**: Makes causal sense? 3. **Confounding variables**: What might explain both? 4. **Actionable experiments**: How to test causality? Output: Risk-rated recommendations (Safe to act / Needs validation / Likely spurious).

A/B test results: Variant A: [METRIC=VALUE, N=SAMPLE_SIZE, CONFIDENCE] Variant B: [METRIC=VALUE, N=SAMPLE_SIZE, CONFIDENCE] Analysis: 1. Winner declaration + p-value 2. Practical significance (effect size) 3. Subgroup analysis suggestions 4. Confidence intervals visualization description 5. Implementation recommendation

Dataset characteristics: [NUMERICAL/CATEGORICAL, DIMENSIONS, TREND/COMPARISON/RANKING GOAL] Recommend 3 visualization types with: Chart | Why This Chart | X-Axis | Y-Axis | Color Encoding | Annotations Needed ------|---------------|---------|---------|----------------|------------------ [Row] | [ ] | [ ] | [ ] | [ ] | [ ] Include accessibility + stakeholder communication considerations.

Key findings: [LIST 3-5 STATISTICALLY SIGNIFICANT RESULTS] Translate to business impact: 1. **Revenue Impact**: $$ quantification or range 2. **Customer Impact**: Retention/churn/LTV effects 3. **Operational Levers**: What dials to turn 4. **Quick Wins**: 30-day actions (<$5K budget) 5. **Strategic Implications**: 12-month roadmap shifts

Segmentation results: [CLUSTER DESCRIPTIONS + SIZES + KEY TRAITS]. Create customer journey narratives: Segment 1: "Meet Sarah..." [pain points → current behavior → ideal solution] Segment 2: etc. Marketing recommendations per segment (personalization tactics).