Data Analysis Report Prompt Framework

Help me scope a data analysis project.

DATASET: [DESCRIBE YOUR DATA SOURCE]
TIME PERIOD: [START DATE] to [END DATE]
BUSINESS CONTEXT: [WHY THIS ANALYSIS MATTERS]

Please help me define:

1. PRIMARY QUESTIONS
What are the 3-5 most important questions this analysis should answer?

2. KEY METRICS
What metrics matter most for answering these questions?

3. SEGMENTATION
How should we break down the data (by time, category, geography, etc.)?

4. BENCHMARKS
What should we compare against (prior periods, targets, industry standards)?

5. STAKEHOLDER NEEDS
What decisions will this analysis inform?

OUTPUT:
A focused analysis brief I can use to guide the actual analysis.

Review this dataset for data quality issues.

[PASTE SAMPLE DATA OR DESCRIBE STRUCTURE]

CHECK FOR:
1. Missing values - What percentage is missing? Which fields?
2. Inconsistent formats - Dates, currencies, categories
3. Outliers - Values that seem implausible
4. Duplicates - Repeated records that might skew analysis
5. Data type issues - Numbers stored as text, etc.

FOR EACH ISSUE FOUND:
- Describe the problem
- Quantify the impact (how many records affected)
- Recommend a fix (imputation, removal, standardization)

OUTPUT:
Data quality report with prioritized cleanup recommendations.

Summarize this dataset with key metrics.

[PASTE DATA OR DESCRIBE]

PROVIDE:

1. VOLUME METRICS
- Total records/transactions
- Records by category
- Time distribution (daily, weekly, monthly counts)

2. VALUE METRICS
- Totals, averages, medians for key numerical fields
- Min/max ranges
- Standard deviations where relevant

3. DISTRIBUTION ANALYSIS
- Top 10 values by frequency for categorical fields
- Percentile breakdowns for numerical fields
- Concentration analysis (e.g., top 20% of customers = X% of revenue)

4. GROWTH METRICS
- Period-over-period changes
- Compound growth rates
- Acceleration/deceleration indicators

FORMAT:
Use tables where appropriate. Include both raw numbers and percentages.

Analyze this data for meaningful patterns.

[PASTE DATA OR SUMMARY FROM PREVIOUS STEP]

LOOK FOR:

1. TIME-BASED PATTERNS
- Trends: Is the metric growing, declining, or flat?
- Seasonality: Weekly, monthly, or quarterly cycles?
- One-time events: Spikes or drops that correspond to specific dates?

2. CORRELATIONS
- Which variables move together?
- Leading indicators: Does X predict Y with a time lag?
- Inverse relationships: When A goes up, does B go down?

3. SEGMENT PATTERNS
- Which segments are growing fastest?
- Which segments are underperforming?
- Are there segments with different behavior patterns?

4. CAUSATION HYPOTHESES
- What might be causing the patterns you see?
- What external factors could explain changes?
- What needs further investigation?

OUTPUT:
Pattern analysis with confidence levels. Distinguish between strong patterns and tentative observations.

Recommend visualizations for this data analysis.

DATA: [DESCRIBE YOUR DATASET AND KEY FINDINGS]
AUDIENCE: [WHO WILL SEE THESE - executives, analysts, operations?]
MEDIUM: [Presentation, dashboard, written report?]

FOR EACH INSIGHT, RECOMMEND:

1. CHART TYPE
- Why this visualization is best for this data
- Alternatives if the first choice is not possible

2. SPECIFICATIONS
- X and Y axes
- Color coding or grouping
- Annotations to highlight

3. IMPLEMENTATION
- Provide Python (matplotlib/seaborn) or JavaScript (Chart.js) code
- Include data formatting needed

PRIORITIZE:
- Charts that show the most important insights
- Visualizations that work well together as a dashboard
- Charts appropriate for the audience's data literacy

Identify anomalies and outliers in this data.

[PASTE DATA OR DESCRIBE]

DETECTION METHODS:
1. Statistical outliers (beyond 2-3 standard deviations)
2. Comparison to historical norms
3. Comparison to segment averages
4. Business rule violations

FOR EACH ANOMALY:
- What is the anomalous value?
- How far from expected is it?
- When did it occur?
- What else happened at that time?

CLASSIFICATION:
- Data quality issue (likely an error)
- One-time event (explainable by external factor)
- Emerging trend (early signal of change)
- Investigation needed (unclear cause)

OUTPUT:
Prioritized anomaly report with recommended actions for each.

Transform this data analysis into an executive report.

ANALYSIS FINDINGS:
[PASTE YOUR KEY FINDINGS FROM PREVIOUS STEPS]

AUDIENCE: [ROLE - CEO, CFO, Board, Operations Manager]
DECISION CONTEXT: [WHAT DECISIONS WILL THIS INFORM?]

REPORT STRUCTURE:

1. EXECUTIVE SUMMARY (3-5 sentences)
- The single most important finding
- Key supporting evidence
- Recommended action

2. SITUATION OVERVIEW
- Business context and why this analysis matters
- Time period and scope
- Data sources and methodology note

3. KEY FINDINGS (3-5 max)
- Finding statement (one sentence)
- Supporting data (specific numbers)
- Business implication
- Confidence level

4. DETAILED ANALYSIS
- Trends and patterns with visualizations
- Segment breakdowns
- Comparisons to benchmarks

5. RECOMMENDATIONS
- Prioritized by impact and feasibility
- Resource requirements
- Success metrics

6. RISKS AND LIMITATIONS
- Data quality caveats
- Analysis limitations
- Alternative interpretations

TONE: [Confident but not overreaching. Data-driven. Action-oriented.]