You get:
- stakeholders drawing wrong conclusions from misleading visuals
- dashboard that look professional but communicate poorly
- executive decisions based on truncated axes that exaggerate changes
- color-blind colleagues who can’t read your red-green charts
- reports that hide the true story because the chart type is wrong
But good charts follow rules:
- axes start at zero for bar charts (or clearly marked if not)
- color is accessible (no red-green, sufficient contrast)
- chart type matches data (time → line, comparison → bar, part-to-whole → donut/bar)
- labels are readable and complete
- no 3D effects that distort perception
Without critique, bad charts survive.
This prompt reviews your visualizations for common problems.
Assume the role of a data visualization expert who critiques charts. Your task is to review a chart description and identify misleading or ineffective elements. Generate: 1. CHART TYPE ASSESSMENT - Current chart type - Is it appropriate for the data? (Yes/No) - Better alternative (if any) 2. AXIS CRITIQUE - Y-axis starts at zero? (If not, is it justified and marked?) - Axis labels clear and complete? - Any truncation that exaggerates differences? 3. COLOR & ACCESSIBILITY - Color-blind safe? (Red-green issues?) - Sufficient contrast? - Color used meaningfully (not just decoration)? 4. DATA INTEGRITY - Any cherry-picked time ranges? - Missing context (comparison points, benchmarks)? - 3D effects distorting perception? 5. RECOMMENDED FIXES - Specific changes to make the chart honest and clear INPUTS: Chart description or image description: [DESCRIBE THE CHART — type, axes, colors, data range] Data being displayed (summary): [E.G., "Monthly revenue Jan-Dec 2025"] Audience: [EXECUTIVE / ANALYST / PUBLIC] Purpose: [E.G., "Show growth trend"] RULES: - Assume good intent (chart may be amateur, not malicious) - Flag misleading elements even if common practice (truncated axes are still wrong) - Suggest specific fixes, not just problems - Note when a chart is technically correct but confusing
- Run this on every chart before it goes into a board deck or client report.
- Use it to review dashboards you inherit from previous analysts.
- For color choices, test with a color-blind simulator — this prompt catches common issues.
- Pay special attention to bar charts with non-zero y-axes — they’re the most common deception.
- Ask “what would a critical reviewer say?” — then fix it before they do.
Chart description:
“Bar chart showing revenue by month. Y-axis from 95 to 100. Bars increase from 95 in Jan to 100 in Dec. Colors: red for months below 98, green for months above 98. 3D effect with slight rotation.”
Data being displayed:
Monthly revenue as percentage of target (100 = hitting target)
Audience:
Executive team
Purpose:
Show that we’re improving toward target
This framework improves outcomes by forcing:
- chart type assessment (right tool for the job)
- axis critique (truncation is deception)
- accessibility check (color-blind viewers matter)
- data integrity review (cherry-picking hides truth)
- specific fixes (actionable improvements)
Great visualization critique doesn’t tear down — it builds better charts.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI coding tools, debugging frameworks, and practical strategies for developers and engineers.