Key Takeaways: Communicating Results: Reports, Presentations, and the Art of the Data Story

This is your reference card for Chapter 31. The skills in this chapter are different from the rest of the book — less code, more craft — but they determine whether your analyses have impact or gather dust.

The Threshold Concept

Communication is not the wrapping paper around your analysis — it IS the deliverable.

If your audience does not understand your findings, those findings do not exist. The last mile of data science is communication, and it requires a different skillset than analysis: empathy for your audience, narrative structure, visual clarity, and the discipline to cut.

Key Concepts

  • Finding vs. Insight. A finding is a fact extracted from data ("churn increased 8%"). An insight places the finding in context and suggests action ("churn increased 8%, driven by discount customers — we need onboarding checkpoints"). Always deliver insights, not findings.

  • The Curse of Knowledge. Once you know something, it is hard to imagine not knowing it. This causes data scientists to skip context, use jargon, show too much, and bury the lead.

  • Audience Analysis. The most important question is "Who am I talking to?" Technical audiences want methodology. Managers want implications. Executives want recommendations. The same analysis requires different presentations for each.

  • The Pyramid Principle. Start with the conclusion, then provide supporting evidence, then provide details. This is the opposite of chronological order but far more effective for non-technical audiences.

  • Narrative Arc. Structure data stories as: Situation (context) → Complication (problem) → Resolution (finding) → Call to Action (recommendation).

The Three Audiences

Audience They want to know They care about Format
Technical How you did it Methodology, reproducibility, rigor Notebooks, technical reports
Managerial What you found Business impact, cost, risk Slides, dashboards, memos
Public/Executive Why they should care Big picture, human impact Summaries, blog posts, infographics

Executive Summary Template

  1. Headline — one sentence capturing the key insight
  2. Context — 2-3 sentences on why the analysis was done
  3. Key findings — 3-5 bullet points (insights, not raw findings)
  4. Recommendation — specific, actionable next step
  5. Caveats — honest limitations

Slide Design Quick Reference

  • Use assertion-evidence titles. Not "Q3 Results" but "Q3 revenue exceeded target by 7%, driven by enterprise."
  • One idea per slide. Slides are free; attention is not.
  • Minimize text. Slides support your words; they do not replace them.
  • Annotate charts. Label key data points, add reference lines, mark events.
  • Use progressive disclosure. Build complex charts step by step.
  • End with the call to action, not "Questions?"

Annotation Checklist for Charts

Every chart in a communication should have:

  • [ ] An assertive title that states the insight (not just the topic)
  • [ ] Labeled axes with units
  • [ ] Annotated key data points — the ones that support your message
  • [ ] Reference lines where relevant (targets, benchmarks, thresholds)
  • [ ] Event markers if the data spans a period with significant events
  • [ ] No chartjunk — unnecessary decoration, 3D effects, or gridlines removed

Notebook Narrative vs. Lab Notebook

Lab Notebook Narrative Notebook
Documents your process Tells your analytical story
Cells may run in any order Reads top to bottom
Cryptic variable names Clear, descriptive names
Commented-out experiments Clean, polished code
Minimal Markdown Rich Markdown narration
No conclusions Clear conclusions and limitations

Dashboard Design Principles

  • Lead with the key metric (largest, top of page)
  • Follow the inverted pyramid (summary → trends → details)
  • Limit to 3-6 charts
  • Use consistent colors and scales
  • Include filters for different user needs
  • Add data source and last-updated date

Common Mistakes to Avoid

Mistake Fix
Data dump (showing everything) Cut to what supports your message
Burying the lead Start with the conclusion
Jargon without translation Use plain language; translate every number
Chart without context Add assertive titles and annotations
False precision Round to meaningful precision
Ignoring uncertainty Communicate ranges and confidence honestly
Cherry-picking Report all relevant findings, including contradictory ones

Writing Tips

  • Use active voice ("We found" not "It was found")
  • Lead with the verb ("Sales dropped 12%" not "There was a 12% drop")
  • One idea per paragraph
  • Translate every number into something meaningful
  • Cut mercilessly — every sentence must earn its place
  • Use formatting — bold key findings, use bullets, add headers

The Communication Checklist

Before delivering any data communication:

  • [ ] I know my audience and what they need to decide
  • [ ] I lead with the insight, not the methodology
  • [ ] I have answered "so what?"
  • [ ] I have included a call to action
  • [ ] Every chart has an assertive title and annotations
  • [ ] Numbers are rounded to meaningful precision
  • [ ] Uncertainty is communicated honestly
  • [ ] I have not cherry-picked findings
  • [ ] Limitations are disclosed
  • [ ] The document can be understood without me present

What You Should Be Able to Do Now

  • [ ] Distinguish between findings and insights
  • [ ] Identify your audience and adapt your communication
  • [ ] Structure a data story with situation, complication, resolution, and call to action
  • [ ] Write a one-page executive summary
  • [ ] Design slides using assertion-evidence titles
  • [ ] Build a narrative notebook that reads as a coherent story
  • [ ] Annotate charts so they communicate without explanation
  • [ ] Evaluate data communications from others for clarity and honesty
  • [ ] Cut ruthlessly — knowing what to leave out is as important as knowing what to include

The Sentence That Guides All Communication

Before writing, presenting, or building anything, ask yourself:

"If my audience walks away remembering only one thing, what should it be?"

Design your entire communication around the answer to that question.

You are ready for Chapter 32, where the focus shifts from how to communicate data to the ethical responsibilities that come with analyzing data about people. Communication is powerful — and with that power comes the obligation to be honest, fair, and thoughtful about the impact of your work.