AI Hints
Home›Data Science Tutorial›Storytelling and Tools
Data Science · Lesson 27 of 101

Storytelling with Data

From findings to action.

⏱ 2 min read▶ Video at 46:38🎓 Free
Watch this lesson in the Data Science courseWatch on YouTube ↗Full Data Science course

The biggest difference between an average analyst and a great one isn't technical. It's that a great one gets things to happen. Numbers alone rarely change anyone's mind. A clear story does.

Storytelling with Data
From the video: Storytelling with Data
  • Great analysts make things happen
  • Numbers alone don't persuade
  • Stories do

The pyramid

Executives don't want your journey. They want the answer. So flip your usual order. Start with the recommendation. Then give three reasons, then the evidence. Keep the details in an appendix, for anyone who asks. If they only read one line, they should get the point.

  1. Recommendationwhat to do
  2. Reasonsthree points
  3. Evidenceone chart each
  4. Appendixthe details
The pyramid
From the video: The pyramid

A one page brief

Here's a real example. Recommendation, move forty thousand dollars of marketing from the bottom cafes to loyalty for top tier customers. Why? The top quarter of customers bring in half of the value. Their lifetime value is highest. And card holders leave less in every cohort. Next step, run a four week pilot in two cities. That's clear, specific, and easy to approve.

Recommendshift $40k to top-tier loyalty
Whytop quarter brings half of value
Evidencecard holders leave less
Next stepfour-week pilot

Talking to non-experts

Replace jargon with plain words. Say, we're fairly sure, in place of, p is less than point zero five. Round your numbers. Forty percent is better than thirty nine point seven. Admit what you don't know, because it builds trust. And always end with the decision that you need from them.

Plain wordsno p-values in meetings
Round numbersforty percent
State uncertaintyit builds trust
Ask for a decisionalways

Summary

Lead with the answer, keep it short, and ask for a decision. Your analysis can now change the business. Next, how to give people a tool, so they don't have to keep asking you.

Key takeaways
  • Answer first
  • Three reasons
  • End with a decision
Next up: Dashboards — from one report to a living tool