School/Data & Analysis/Data-Driven Decisions
4/4
Wave 510 minintermediate

Data Visualization Prompts

Generate chart specifications and bring your data to life visually.

Data Visualization Prompts

AI can't generate charts directly in most chat interfaces — but it can tell you exactly what to build, write the code to create it, and even generate chart specs you can paste into tools.

Choosing the Right Chart Type

"I want to visualize [describe your data and what story you want to tell]. What chart type should I use and why? Consider:

- The data type (time series, categorical, proportional)

- The audience (technical vs executive)

- The message (comparison, trend, composition, distribution)"

Quick Chart Reference

You Want to Show...Use This Chart
Trends over timeLine chart
Comparisons between itemsBar chart (horizontal for many items)
Parts of a wholePie chart (2-5 segments only) or stacked bar
Relationships between variablesScatter plot
Distribution of valuesHistogram or box plot
Geographic patternsMap or heat map
Process flowSankey diagram or funnel
Multiple metrics at onceDashboard with small multiples

Detailed Chart Specifications

"Create a detailed specification for a chart showing [what]:

- Chart type and rationale

- X-axis: label, scale, range, tick marks

- Y-axis: label, scale, range

- Data series to include (with colors — suggest hex codes)

- Title and subtitle

- Key annotations to highlight (peaks, drops, milestones)

- Legend placement

- Recommended tool to create it (Excel, Google Sheets, Tableau, Chart.js)"

Code-Based Visualizations

If you can run code, AI can generate publication-quality charts:

Python (matplotlib/seaborn)

"Write Python code using matplotlib to create a [chart type] from this data:

[paste data]

Requirements:

- Clean, modern style (use seaborn or a custom theme)

- Proper axis labels and title

- Annotations on key data points

- Color palette: [business professional / bold / minimal]

- Save as PNG at 300 DPI"

Google Sheets Chart Script

"Write a Google Apps Script that creates a [chart type] in my Google Sheet.

The data is in columns A through E, rows 2-50.

Customize: title, axis labels, colors, and legend position."

HTML/JavaScript (for web dashboards)

"Write HTML and JavaScript using Chart.js to create an interactive [chart type].

Data: [paste or describe]

Features: tooltips on hover, responsive sizing, clean design."

Data Storytelling

Charts are only half the equation. The story you tell with them matters more.

"I have this chart showing [describe the chart and data]:

Write a 3-paragraph data story that:

1. Opens with the most surprising or important finding

2. Provides context for why this matters

3. Ends with a clear recommendation or call to action

Audience: [who is reading this]"

Common Visualization Mistakes

Ask AI to audit your chart choices:

"I'm planning to use a [chart type] to show [data]. Evaluate this choice:

- Is this the right chart type?

- What could go wrong or be misleading?

- How could I improve the visualization?

- What common mistakes should I avoid?"

Common pitfalls AI will catch:

  • Pie charts with too many segments (use bar chart instead)
  • Truncated Y-axis that exaggerates differences
  • 3D charts that distort proportions
  • Missing context (no title, no units, no time period)
  • Color choices that are inaccessible to colorblind viewers

Exercises

0/4
Prompt Challenge+15 XP

Describe a dataset you work with and ask AI to recommend the best visualization approach. Then ask for a detailed chart specification. Does the recommendation make sense for your audience?

Hint: Include details about who will see the chart (technical vs executive) — this changes everything about the recommendation.

Quiz+5 XP

What chart type is best for showing how something changes over time?

Quiz+5 XP

Why should you avoid pie charts with more than 5 segments?

Reflection+10 XP

Think about the last presentation or report you created. What chart or visualization would have made the data clearer? Describe the data, the audience, and the chart you would use.

Hint: Consider whether you were showing a trend, a comparison, or a composition. Each has an ideal chart type.