Storytelling with Data & AI

Data Doesn't Speak for Itself

AI can analyse a dataset in seconds and produce charts, trends, and summaries. But a spreadsheet has never changed anyone's mind. Stories do.

Why This Skill Is AI-Proof

AI can tell you that customer churn increased 15% last quarter. It can generate a chart. It can even suggest possible causes. What it can't do is:
• •Know which stakeholder cares about which metric
• •Frame the data in a way that drives a specific decision
• •Build narrative tension that makes people pay attention
• •Connect data to emotion and urgency

That's storytelling. And it's increasingly what separates good analysts from great ones.

The Data Story Framework

1. Start with the audience What do they care about? What decision are they facing? What would make them act? 2. Lead with the insight, not the data

•Bad: "Q3 churn was 12.3%, up from 10.7% in Q2."

•Good: "We're losing customers faster than we're gaining them — and it's accelerating."

3. Use the And-But-Therefore structure

•And (context): We launched three new features AND customer usage increased 20%.

•But (tension): BUT churn also increased 15%, concentrated in mid-market accounts.

•Therefore (action): THEREFORE we need to investigate whether new features are creating complexity that drives mid-market customers away.

4. One chart, one message Every visualisation should make exactly one point. If you need to explain what a chart means, the chart isn't working. 5. End with the decision Data stories should end with a clear ask: here's what I recommend, here's what I need from you, here's the decision we need to make.

Using AI to Build Better Data Stories

AI is your research assistant, not your storyteller:
• •Use AI to find patterns in data
• •Use AI to generate chart options
• •Use AI to identify counterarguments
• •Then you craft the narrative that connects it all