Data Storytelling: Turning Dashboards into Action

Plants.work · 2026-10-26 · upskilling

A data storytelling structure of context, insight, implication and action, plus how to use AI to draft the narrative without letting it invent causes.

You present the dashboard at the monthly meeting and walk through every chart. Everyone nods, and the meeting ends with nobody taking on any task. The data was correct and the charts looked good, yet no decision was made. What was missing was a "story" connecting the numbers to what people in the room need to do.

What is data storytelling?

Data storytelling means presenting data with a clear narrative: what happened, why it matters and what to do next. It differs from reading charts aloud one by one, because people remember stories better than isolated numbers, and a good story ends with a clear request.

A 4-step story structure

1. Context

Say what we are looking at and what we compare it with, for example "Provincial branch sales this quarter against the target set at the start of the year."

2. Insight

Choose the single most important thing, not everything you saw. For example: "Southern branches are consistently below target, while other regions are on track."

3. Implication (so what)

Translate the numbers into what the audience cares about: the effect on full-year profit, on stock, or on the sales team.

4. Action (now what)

Offer 1 to 3 options and state what you need from the meeting, for example "Approval to pilot a promotion in 5 branches for 4 weeks."

Techniques that let charts tell the story

Conclusion titles: write what you want people to remember as the chart title
Annotations: add short notes on the chart where events happened, such as a campaign launch or a system outage
Highlight selectively: use grey as the base and a strong colour only for the line that is the hero of the story
Cut what is irrelevant: move charts that do not support the main story to an appendix

Adapt the same story to the audience

The same data should be told differently depending on who is listening, because each person makes different decisions.

Senior executives: start with the impact on the annual target and the options needing approval, in no more than 2 to 3 minutes.
Area managers: focus on which branches must do what, and by when.
Store teams: explain the new goal in simple language and what changes in daily work.

Example: a nationwide coffee chain

An analyst finds that southern branch sales fall below target during the rainy season. Instead of opening all 6 dashboard pages, she starts with a single line chart highlighting the South against other regions, annotates the heavy-rain period, then shows that the share of delivery orders at those branches rose over the same period. She proposes piloting more delivery promotions in 5 branches and asks marketing to own it. The meeting ends with a clear owner and a review date.

Numbers tell you what happened. The story tells you what we do next.

Using AI to tell the story

AI is good at drafting. Microsoft Learn documentation says Copilot in Power BI can create a narrative visual that summarises a report in text. Or you can give a summary table to ChatGPT, Claude or Gemini and ask: "Write a 4-step story (context, insight, implication, action) for an executive with 2 minutes."

Watch out for

AI can invent causes. AI sees only numbers, not what happened on the ground. If it writes "sales fell because a competitor opened a new branch", you need evidence.
Correlation is not causation. Two lines moving together do not mean one causes the other.
Do not cherry-pick data that suits the story. If some data contradicts it, mention it; your credibility will go up.

What to do next

Before your next meeting, write your main message in one sentence.
Choose no more than 3 charts that support it.
Add conclusion titles and annotations.
Let AI draft the story, then add causes you have verified.
End with a clear request, naming an owner and a follow-up date.

To shape your dashboard for storytelling, read Designing a KPI Dashboard Executives Can Read in 30 Seconds, and practise using AI for analysis with the course AI for data analysis, for everyone on Plants.work Learn.