DAData

AI Data Analysis Prompt Generator

Build a structured AI prompt for metrics analysis, segmentation, anomaly checks, charts, and decision-ready insights.

Data Analysis Prompt

Act as a rigorous data analysis assistant.

Analysis question:
Which new AI tool categories drive repeat visits and related-tool clicks?

Decision owner:
Product owner planning Batch 3

Available data:
Page views, copy button events, related-tool clicks, Search Console impressions, locale, tool category, publish date.

Target output:
Analysis plan

Requirements:
1. Start by confirming grain, time range, filters, event definitions, and metric logic.
2. Propose primary metrics, supporting metrics, segments, and possible comparison groups.
3. Run data-quality checks first: missingness, duplicates, spikes, sample size, sampling, and tracking changes.
4. Separate descriptive findings, correlations, and causal hypotheses.
5. Recommend charts or tables that answer the question and explain what each should reveal.
6. Return insights with confidence, risks, validation steps, and the decision they can support.

Use this structure:
- Questions to confirm first
- Analysis plan
- Metrics and segmentation table
- Data-quality checks
- Chart recommendations
- Insight template
- Next experiment or data request

How to use Data Analysis Prompt

Step 1

Name the decision before asking for metrics.

Step 2

Describe data grain, filters, time range, and known limitations.

Step 3

Ask for sanity checks before accepting charts or conclusions.

Example

Sample input

Analysis question: Which new AI tool categories drive repeat visits and related-tool clicks?
Available data: Page views, copy button events, related-tool clicks, Search Console impressions, locale, tool category, publish date.
Decision owner: Product owner planning Batch 3
Output format: Analysis plan

Result preview

Act as a rigorous data analysis assistant.

Analysis question:
Which new AI tool categories drive repeat visits and related-tool clicks?

Decision owner:
Product owner planning Batch 3

Available data:
Page views, copy button events, related-tool clicks, Search Console impressions, locale, tool category, publish date.

Target output:
Analysis plan

Requirements:
1. Start by confirming grain, time range, filters, event definitions, and metric logic.
2. Propose primary metrics, supporting metrics, segments, and possible comparison groups.
3. Run data-quality checks first: missingness, duplicates, spikes, sample size, sampling, and tracking changes.
4. Separate descriptive findings, correlations, and causal hypotheses.
5. Recommend charts or tables that answer the question and explain what each should reveal.
6. Return insights with confidence, risks, validation steps, and the decision they can support.

Use this structure:
- Questions to confirm first
- Analysis plan
- Metrics and segmentation table
- Data-quality checks
- Chart recommendations
- Insight template
- Next experiment or data request

FAQ

Does this analyze my dataset directly?

No. It creates a strong prompt and analysis checklist for an AI assistant or analyst workflow.

Why describe data limitations?

Limits such as missing events, sampling, or short time windows prevent confident conclusions from weak evidence.