Data Storytelling

Analyst charts answer "what does the data show?" — storytelling charts answer "what do I want the reader to understand?" This skill is for the latter.

When to invoke

• The user is preparing a report, blog post, pitch, or briefing — not exploring for themselves.
• The audience is non-specialist or time-constrained.
• A single take-away needs to land.
• The user says "storytelling", "narrative", "walk me through", "explainer", or "scrollytelling".

Core principles

1. One chart, one claim. Each visual carries a single headline finding. If you need three claims, build three charts.
2. Write the headline first. The chart title is the argument — e.g. "Inflation fell faster than any G7 peer" — not "CPI by country, 2023-2025".
3. Annotate the evidence. Mark the exact data points that support the headline directly on the chart (arrows, labels, call-outs).
4. Strip everything else. Remove gridlines, legends, and axes that don't carry the story. Default to low ink.
5. Sequence, don't cram. For multi-step arguments, prefer a sequence of simple charts (scrollytelling / animated states) over one dense chart.