data-storytelling

name	Data Storytelling
description	Build data narratives using three-act structure and audience-first framing to turn analysis into compelling arguments that drive decisions.

Purpose

Transform data analysis into a complete, coherent narrative. A collection of charts with captions is a report. A data story says "here's what happened, here's why it matters, and here's what we should do." Data stories have arguments, not just observations.

Steps

1. Identify the audience and their decision context (executive, manager, analyst, general)
2. Write the Big Idea in one sentence: "[Audience] should [action] because [evidence from data]"
3. If you cannot write the Big Idea sentence, the analysis is not done yet. Go back to data analysis.
4. Structure findings into three acts:
   • Setup: Establish context and baseline ("here's how things were")
   • Conflict: Reveal the surprise, problem, or gap ("but then this happened")
   • Resolution: Deliver insight and recommendation ("this means X, we should do Y")
5. For each finding, choose a chart type based on story intent:
   • Compare: bar chart (horizontal if many labels)
   • Trend: line chart
   • Distribution: histogram or box plot
   • Part-to-whole: stacked bar or pie (max 5 slices)
   • Relationship: scatter plot
6. Title every chart as an insight statement, not a label ("Revenue grew 34%" not "Revenue by Quarter")
7. Add annotations: callouts on key data points, footnotes for methodology and caveats
8. Create the deliverable: Word document, PowerPoint deck, or Excel workbook depending on audience
9. Validate: read top to bottom. Does it tell a story with setup, tension, and resolution?

Audience Matching

Audience	Time Budget	What They Need	Deliverable
Executive	30 seconds	Headline + action recommendation	PowerPoint (max 5 slides)
Manager	2 minutes	Context + options + trade-offs	Word report with summary
Analyst	Unlimited	Full data + methodology + caveats	Excel workbook with detail
General	1 minute	Simple story, familiar visuals	Word doc with annotated charts

Knaflic Method (Storytelling with Data)

1. Understand context: who is the audience? What do they need to do?
2. Choose an effective visual: match story intent to chart type
3. Eliminate clutter: remove everything that is not data or supporting story
4. Focus attention: use color, size, and position to direct the eye
5. Tell a story: connect visuals with narrative text

Duarte Contrast Pattern

Use "What Is" vs. "What Could Be" when the story needs to motivate action:

• What Is: "Today, we process 500 support tickets per day"
• What Could Be: "With the new model, we could process 2000 with the same team"

Quality Checks

Before delivering, verify:

• Big Idea exists as one sentence
• Three-act structure is identifiable (setup, conflict, resolution)
• Every chart title is an insight statement, not a label
• Same category uses the same color across all visuals
• Every chart cites its data source
• Each chart conveys its point in 3 seconds
• Hero chart has at least one callout annotation
• Date range and "as of" timestamp included

Guidelines

• Default to explanatory mode (guide the viewer) for executives and general audiences
• Use exploratory mode (let the viewer discover) only for analysts
• No chart collection dumps. Every visual must serve the argument.
• If a chart does not support the Big Idea, remove it
• Annotations carry the argument; the chart is evidence
• Always end with a recommendation, not just "here's what happened"