Guide chart selection and visualization design based on data type and audience.
## Data Visualization Guide
Create the optimal visualization for [DATA_TYPE]:
**Data Type:** [DATA_TYPE — describe your data]
**Message:** [MESSAGE — what story should the chart tell]
**Audience:** [AUDIENCE — technical/executive/public/mixed]
**Medium:** [MEDIUM — presentation/report/dashboard/social media]
**Chart Selection Decision Tree:**
### What are you comparing?
**Change over time** → Line chart
- Few categories: Standard line
- Many categories: Small multiples
- Show magnitude: Area chart
- Discrete periods: Bar chart
**Part-to-whole** →
- Few categories (≤5): Donut/pie chart
- Many categories: Stacked bar (100%)
- Hierarchical: Treemap
- Two levels: Sunburst
**Comparison** →
- Categories: Horizontal bar chart
- Two variables: Grouped bar
- Rank: Lollipop chart
- Before/after: Slope graph
**Distribution** →
- One variable: Histogram / box plot
- Two variables: Scatter plot
- Density: Violin plot / heatmap
- Over time: Ridgeline plot
**Relationship** →
- Two variables: Scatter plot
- Three variables: Bubble chart
- Network: Node-link diagram
- Correlation matrix: Heatmap
### Design Best Practices
1. **Title:** State the insight, not just the topic ("Sales grew 23% in Q4" not "Q4 Sales")
2. **Color:** Use sparingly — highlight one thing, gray the rest
3. **Labels:** Direct-label data points (avoid legends when possible)
4. **Axis:** Start bar charts at zero, line charts can start elsewhere
5. **Annotations:** Add context for peaks, valleys, and anomalies
6. **Simplify:** Remove gridlines, borders, and decorations that don't add information
**Output:** Chart recommendation with mockup specifications.
**Variables:** Replace [DATA_TYPE], [MESSAGE], [AUDIENCE], [MEDIUM].Free to copy and use. Compatible with GPT-5, Gemini 2.5 Pro, Llama 4.
Replace all [BRACKETED] variables with your specific data and requirements.
Initial release
Sign in and download this prompt to leave a review.