Capstone: Build a Data Story from Scratch
Apply everything you've learned. Take raw data and transform it into a compelling data story with clear visualizations, thoughtful design, and a presentation that drives action.
Premium Course Content
This lesson is part of a premium course. Upgrade to Pro to unlock all premium courses and content.
- Access all premium courses
- 1000+ AI skills included
- New content added weekly
From Numbers to Narrative
Everything you’ve learned converges here. Chart selection. Design principles. Storytelling. Dashboards. Color and accessibility. Presentation skills.
In this capstone, you’ll work through a complete data visualization project—from raw data to a finished presentation. We’ll use a realistic dataset and walk through every decision, showing how the techniques from each lesson apply.
The Dataset: E-Commerce Performance
You’ve been asked to present quarterly performance to the leadership team of an e-commerce company. Here’s the raw data:
Monthly Revenue (2025):
Jan: $420K | Feb: $445K | Mar: $510K
Apr: $485K | May: $520K | Jun: $580K
Jul: $540K | Aug: $560K | Sep: $610K
Oct: $680K | Nov: $920K | Dec: $1.1M
Revenue by Channel:
Organic search: 35% | Paid ads: 28% | Email: 20%
Social media: 10% | Direct: 7%
Customer Metrics:
New customers/month: 2,400 → 4,100 (Jan → Dec)
Repeat purchase rate: 42% → 38% (declining)
Average order value: $85 → $72 (declining)
Customer acquisition cost: $32 → $45 (increasing)
Customer lifetime value: $340 → $290 (declining)
Product Categories:
Electronics: 40% of revenue (growing)
Home & Garden: 25% (stable)
Clothing: 20% (declining)
Accessories: 15% (growing)
Customer Satisfaction:
NPS: 62 → 48 (Jan → Dec)
Returns rate: 4.2% → 7.8% (increasing)
Support tickets/month: 800 → 2,100 (increasing)
Look at all that data. There are multiple stories hiding in these numbers. Let’s find the most important one.
Stage 1: Data Exploration (Finding the Story)
Before making any charts, ask AI to help identify the story:
Here's quarterly data for an e-commerce company.
Help me find the most compelling story for the
leadership team.
[paste all the data above]
Identify:
1. The 3 most important patterns or trends
2. Any contradictions that would surprise leadership
3. The single most important story to tell
4. What this data suggests the company should do
AI identifies the key story: “Revenue is growing, but the business is getting less healthy.”
Revenue doubled from $420K to $1.1M—impressive. But underneath:
- Repeat purchase rate is falling (38% vs. 42%)
- Average order value is dropping ($72 vs. $85)
- Customer acquisition cost is rising ($45 vs. $32)
- Customer lifetime value is declining ($290 vs. $340)
- NPS dropped from 62 to 48
- Returns nearly doubled
The company is growing by spending more to acquire customers who spend less and are less satisfied. This isn’t sustainable.
That’s the story. Revenue growth is masking deteriorating fundamentals.
Stage 2: Chart Selection (Lesson 2)
Now select the right charts for each part of the story:
I'm telling the story: "Revenue is growing but the
business is getting less healthy."
Narrative arc:
1. Context: Revenue doubled this year (good news)
2. Tension: But customer metrics are declining
3. Insight: We're buying growth at the expense of quality
4. Action: Recommendations
For each narrative beat, recommend the best chart type
and explain why. I'll have 6-8 charts maximum.
AI recommends:
| Story Beat | Chart | Reasoning |
|---|---|---|
| Revenue growth | Line chart (monthly) | Shows the impressive upward trend |
| The revenue breakdown | Stacked area by channel | Shows growing reliance on paid ads |
| Customer health decline | Small multiples (4 metrics) | Shows simultaneous decline across metrics |
| AOV vs. CAC | Dual line chart | Shows the scissors—AOV dropping while CAC rises |
| NPS + Returns | Combo chart | Links satisfaction to tangible impact |
| Revenue vs. LTV trend | Diverging bar chart | Contrasts short-term revenue with long-term value |
| Channel ROI | Horizontal bar | Ranks channels by actual profitability |
| Recommendation impact | Projection chart | Shows what happens if trends continue vs. if we act |
Stage 3: Design and Build (Lesson 3)
Now design each chart following the principles:
For each chart in our story, create a specification:
1. Exact title (insight-driven, not topic-driven)
2. Color scheme (using our palette: blue primary, orange accent, gray neutral)
3. What to emphasize vs. de-emphasize
4. Annotations needed
5. Labels and axis configuration
6. Any elements to remove for clarity
Remember: colorblind-safe, high contrast, minimal clutter.
AI produces specifications like:
Chart 1 - Revenue Line:
- Title: “Revenue Doubled in 2025—But at What Cost?”
- Single blue line, y-axis starting at $0
- Annotate November (“Black Friday: $920K”) and December (“Holiday peak: $1.1M”)
- Light gray gridlines, direct labels on Jan and Dec values only
Chart 4 - AOV vs. CAC:
- Title: “We’re Paying More to Attract Customers Who Spend Less”
- Two lines: blue (AOV declining) and orange (CAC rising)
- Highlight the crossover implication with a shaded “danger zone”
- Annotate the gap: “In January, we spent $32 to acquire a customer spending $85. Now we spend $45 for a customer spending $72.”
That annotation is the chart’s power moment. The numbers in context tell the story more powerfully than the lines alone.
Stage 4: Storytelling and Sequence (Lesson 4)
Write the narrative bridges between charts:
Write the presentation script for my 6-chart data story.
For each chart:
1. What I say BEFORE showing it (setup)
2. What I say WHILE showing it (key point)
3. What I say AFTER (transition to next chart)
**Quick check:** Before moving on, can you recall the key concept we just covered? Try to explain it in your own words before continuing.
Story flow:
Chart 1: Revenue doubled (context)
Chart 2: Customer metrics declining (tension)
Chart 3: AOV vs CAC scissors (deepening tension)
Chart 4: NPS and returns (evidence)
Chart 5: Channel profitability (insight)
Chart 6: Projected scenarios (action)
Tone: Direct, confident, concerned but not alarmist.
Audience: CEO, CFO, VP Marketing, VP Operations.
Total time: 10 minutes.
AI writes the script. For example, the transition from Chart 1 to Chart 2:
“That growth is real and it’s impressive. But when I dug into the customer metrics behind those revenue numbers, I found a pattern that concerns me. Let me show you what’s happening under the surface.”
This transition creates anticipation. The audience knows something’s coming. They’re leaning in.
Stage 5: Accessibility Check (Lesson 6)
Before finalizing:
Review my 6-chart presentation for accessibility:
Charts use: blue (#2563EB), orange (#F97316), gray (#6B7280)
Background: white
Font sizes: titles 24px, labels 14px, annotations 12px
Check:
1. Colorblind safety (deuteranopia simulation)
2. Contrast ratios for all text
3. Do any charts rely on color alone to convey information?
4. Write alt-text for each chart
AI identifies: the 12px annotation text has marginal contrast against white—suggest increasing to 14px or darkening to #4B5563. All other elements pass. Each chart has secondary encoding (labels, annotations) in addition to color. AI generates alt-text for all six charts.
Stage 6: Prepare for Questions (Lesson 7)
I'm presenting to CEO, CFO, VP Marketing, VP Operations.
My recommendation: Shift 20% of paid ad budget to
retention marketing (email, loyalty program, customer
experience improvements).
Anticipated questions:
1. CEO: "What's the revenue risk of reducing paid spend?"
2. CFO: "What's the ROI timeline for retention investment?"
3. VP Marketing: "Won't we lose new customer volume?"
4. VP Ops: "What specifically is causing the returns increase?"
Help me prepare concise answers with supporting data
from our dataset. Also suggest 3 questions I might
not anticipate.
AI helps you prepare for the toughest questions, including ones you hadn’t considered: “Could the AOV decline be seasonal?” (Answer: no, it’s trending down across all months, not just holiday discount periods) and “Are competitors seeing the same NPS decline?” (Answer: worth investigating—need competitor benchmarking data).
The Complete Presentation
Here’s the final flow:
Slide 1: Hook “Revenue doubled this year. But I need to show you what’s happening underneath the numbers.”
Slide 2: Revenue Growth (Chart 1) Clean line chart. Revenue from $420K to $1.1M. Impressive.
Slide 3: Customer Health Dashboard (Chart 2) Four small multiples showing decline: repeat rate, AOV, LTV, and NPS. All going the wrong direction.
Slide 4: The Cost of Growth (Chart 3) AOV vs. CAC chart with the scissors pattern. The annotation drives the point home.
Slide 5: Customer Experience (Chart 4) NPS decline with returns increase. “Our customers are telling us something.”
Slide 6: Channel Reality (Chart 5) Horizontal bars showing actual profitability by channel. Email is 4x more profitable than paid ads. But we’re spending 40% more on paid and neglecting email.
Slide 7: Two Futures (Chart 6) Projection chart. Scenario A: continue current trajectory (revenue grows but profitability craters). Scenario B: rebalance toward retention (slightly slower revenue growth but dramatically better profitability).
Slide 8: Recommendation Three specific actions with projected impact and timeline.
Your Challenge
Build your own capstone. Use real data from your work or organization:
- Explore the data with AI—find the most compelling story
- Select 5-7 charts using the chart decision framework
- Design each chart with insight titles, emphasis, and minimal clutter
- Sequence them into a narrative with context → tension → insight → action
- Check accessibility (color, contrast, alt-text)
- Prepare for questions your audience will ask
- Present to someone—or practice presenting to yourself, timed
This exercise integrates every skill from the course. It takes 2-3 hours for a complete data story, and the result is a presentation that’s more effective than most professionals produce.
What You’ve Learned in This Course
| Lesson | Skill | Key Principle |
|---|---|---|
| 1 | Visualization fundamentals | Every chart needs one clear message |
| 2 | Chart selection | Match chart type to the relationship you’re showing |
| 3 | Design | Titles are insights, emphasis guides attention, clutter hides data |
| 4 | Storytelling | Context → Tension → Insight → Action |
| 5 | Dashboards | Design for decisions, not data; hierarchy matters |
| 6 | Color & accessibility | Never rely on color alone; design for everyone |
| 7 | Interactivity & presenting | Default views for most; prepare for questions |
| 8 | Capstone | Everything together: from raw data to finished story |
Key Takeaways
- Start with the story, not the charts—explore data to find the insight first
- Select charts that serve the narrative, not charts that show the most data
- Design every element with purpose: titles, colors, emphasis, annotations
- Sequence visualizations to build a narrative arc
- Accessibility isn’t optional—8% of your audience may have color vision deficiency
- Prepare for questions to build credibility and depth
- The goal isn’t beautiful charts—it’s data that drives decisions
Congratulations on completing Tell Stories with Data and AI. You now have the skills to transform any dataset into a compelling visual story. Go make data visible.
Knowledge Check
Complete the quiz above first
Lesson completed!