Data Insight Summarizer

I'll paste raw analysis output: numbers, tables, chart descriptions, or query results. Turn it into a summary an executive would act on.

Structure:

**Headline** — one sentence. The single most important thing this data says. Not "revenue trends this quarter" but "revenue grew 22% quarter over quarter, driven by expansion in the SMB segment."

**What changed** — bullet list of the specific changes with the magnitude. "Signups up 40% MoM," not "signups increased significantly." Include the comparison period.

**Why it likely changed** — 2-4 hypotheses ranked by fit. Mark each as: strong evidence, plausible, or speculative.

**What to do about it** — 2-3 concrete actions with owners. If the answer is "nothing yet, we need more data," say so.

**What we DON'T know** — the important open questions this analysis doesn't answer.

**Confidence** — one line saying how much to trust the top-level conclusion.

Rules:
- Never invent numbers. Only use what's in the input.
- Never smooth over uncertainty. Executives prefer "we think X, but this could be noise from Y" over false confidence.
- If the data is too thin to draw conclusions, say so and stop.
- No "further analysis needed" as the entire recommendation — be specific about what analysis.

## How to use it

Paste this prompt, then paste your analysis (numbers, a table, a chart description). Send the summary to leadership; save the full analysis in the appendix.

## Why it works

Analysts often present the analysis and let the exec derive the insight. That's how good work goes unread. Making the "so what" explicit and calibrated moves the decision without hiding the uncertainty.