Data Visualization Recommender
Added Apr 2, 2026
About This Prompt
This prompt helps analysts and data professionals select the most effective chart types for their specific data and audience. Rather than defaulting to bar charts and line graphs, this prompt provides expert-level guidance on visual encoding, axis mapping, and common pitfalls for each recommendation. It considers the nature of the data, the analytical question being asked, and who will be viewing the visualization. The output includes actionable specifications that can be directly implemented in tools like Tableau, Power BI, or Python libraries, making it a practical bridge between raw data and compelling visual stories.
Variables to Customize
[DATASET_DESCRIPTION]
Overview of your data including columns, types, and row count
Example: Monthly sales data with columns: date, region (4 regions), product_category (8 categories), revenue, units_sold. 960 rows over 2 years.
[ANALYSIS_QUESTION]
The specific question your visualization should answer
Example: Which product categories are growing fastest across regions?
[AUDIENCE]
Who will view this visualization and their data literacy level
Example: Executive leadership team in a quarterly business review
[VIZ_TOOL]
The visualization tool or library you plan to use
Example: Tableau Desktop
Tips for Best Results
- Be specific about data types and cardinality so the recommendations account for visual clutter
- Mention if your data has significant outliers or missing values as this affects chart choice
- Include whether this is for a live presentation, printed report, or interactive dashboard since each format favors different chart types
Example Output
**Recommendation 1: Small Multiples Line Chart (Sparkline Grid)** Why: With 8 categories across 4 regions, a single chart would be cluttered. Small multiples let you compare growth trends at a glance. - X-axis: Month (time), Y-axis: Revenue, Facet: Region (columns) x Category (rows) - Color: Single color per panel with a reference line for overall average - Pitfall: Ensure all panels share the same Y-axis scale to avoid misleading comparisons...