Module 11: Data Storytelling and Communication

"The greatest value of a picture is when it forces us to notice what we never expected to see." — John Tukey

Learning Objectives

By the end of this module, you will be able to:

• Structure a data narrative with beginning, middle, and end
• Choose visualizations that match your audience and message
• Design dashboards that guide users to insights
• Write clear, concise data reports for different stakeholders
• Present findings confidently to non-technical audiences
• Avoid the common pitfalls that undermine credibility

Why Storytelling Matters More Than Analysis

The best analysis fails if no one acts on it. Jargon-heavy stats slides lose to one clear sentence: what’s wrong, how big it is, what to do Monday.

1. The Narrative Arc of Data

Three beats:

Example: E-Commerce Analysis Story

Context: Revenue up 8% — looks healthy. Conflict: Repeat rate fell from 42% to 31%; acquisition is expensive vs retention. Resolution: Post-purchase journey in the 60-day window, loyalty test, homepage recs A/B — each with rough cost and upside.

Template for Structuring Your Story

2. Choosing Visualizations for Your Audience

The Audience Matrix

Chart Selection Guide

The One-Chart Rule

If the audience remembers one image, which makes the conflict obvious? Often it’s two metrics on linked axes (e.g. revenue up, margin down).

Key Example: Dual-axis bar + line: revenue growth with shrinking margin — the title is the insight.

[object Object], matplotlib.pyplot ,[object Object], plt

months = [,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],]
revenue = [,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],]
margin = [,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],, ,[object Object],]

fig, ax1 = plt.subplots(figsize=(,[object Object],, ,[object Object],))
ax1.bar(months, revenue, color=,[object Object],, alpha=,[object Object],, label=,[object Object],)
ax1.set_ylabel(,[object Object],, color=,[object Object],)

ax2 = ax1.twinx()
ax2.plot(months, margin, color=,[object Object],, marker=,[object Object],, linewidth=,[object Object],, label=,[object Object],)
ax2.set_ylabel(,[object Object],, color=,[object Object],)
ax1.set_title(,[object Object],)

fig.tight_layout()
plt.show()

3. Dashboard Design Principles

The 5-Second Rule

Main message readable in one glance — KPI row or hero chart first.

Layout Principles

Dashboard Anti-Patterns

• Rainbow defaults for categories
• Duplicate metrics that disagree
• Filters that don’t apply to the whole page
• Chartjunk 3D

KPI Card Design

Metric, comparison (YoY, vs target), sparkline or delta arrow, footnote on definition.

4. Writing Data Reports

Report Structure

Executive path first; methodology and appendix last.

Executive Summary (1 paragraph)

Situation, biggest finding, recommended action, rough impact. No jargon wall.

Background & Methodology

Data sources, time window, definitions, known gaps — short and honest.

Key Findings (3-5, ordered by importance)

Finding 1: [Statement, not question]

One declarative sentence, supporting number or visual, business implication.

Finding 2: [Statement]

Same pattern — most important findings first.

Recommendations

Numbered actions with owners, effort, and expected effect (even ranges).

Risks & Caveats

What could be wrong with the data or logic? Builds trust.

Appendix

Extra charts, SQL, data dictionary.

Writing Tips for Analysts

• Lead with verbs: “Revenue grew…” not “An analysis was conducted…”
• Replace “significant” with how big and for whom
• Each figure needs a caption that states the takeaway

5. Presenting to Stakeholders

The Pyramid Principle

Answer first, then reasons. Executives want the recommendation before the journey.

Presentation Structure (10-15 minutes)

Handling Questions

“I don’t know, I’ll follow up by [time]” beats bluffing. Park deep dives: “Slide 12 in appendix.”

6. Common Pitfalls

Data Storytelling Anti-Patterns

Practice Exercises

Exercise 1: Chart Redesign (Beginner)

Take a cluttered chart; one paragraph on what you removed and why.

Exercise 2: One-Page Report (Intermediate)

Executive summary + 2 findings + 3 recommendations on one page.

Exercise 3: Dashboard Wireframe (Intermediate)

Sketch F-layout with 5-second rule in mind.

Exercise 4: Presentation Dry Run (Advanced)

Record yourself; cut anything before the first recommendation.

Exercise 5: The Persuasion Challenge (Advanced)

Same data, two audiences (exec vs peer) — adjust depth and charts.

Mini-Project: Complete Data Story

Real or public data: narrative arc, one hero visual, written report, 10-minute spoken version, appendix with methods.

Key Takeaways

1. Structure: context → tension → resolution.
2. Audience sets density — execs need fewer, sharper charts.
3. Titles and captions carry the argument.
4. Pyramid your spoken delivery: answer first.
5. Credibility includes caveats.

Resources for Further Learning

• Storytelling with Data — Cole Nussbaumer Knaflic
• Presentation Zen — Garr Reynolds
• The Pyramid Principle — Barbara Minto

Next up: Module 12 — Capstone Project Guide — packaging everything into a portfolio piece.