A/B testing tools use heat maps to show where users click on web pages. Fitness apps use heat maps to show workout intensity across body parts. Financial analysts use heat maps to show correlation strengths across asset classes. Heat maps have become ubiquitous because they communicate density and intensity patterns that other chart types struggle to show. But heat maps are also frequently misused, creating confusion where they should create clarity.
Heat maps work by encoding values as color intensity. Darker or more saturated colors indicate higher values; lighter or less saturated colors indicate lower values. The viewer reads the pattern rather than individual values, identifying hotspots and coldspots intuitively. This pattern recognition is heat maps' primary strength—and their primary limitation. They're excellent for showing where patterns exist but poor for showing precise values.
The most common heat map mistake is inappropriate color scale selection. Rainbow colormaps—the full visible spectrum from red through violet—are frequently used but fundamentally misleading. The rainbow scale creates artificial boundaries between colors that don't correspond to data boundaries, implies ordering where none exists, and performs terribly for colorblind viewers. Sequential scales (light to dark in a single hue) or diverging scales (light to saturated in two hues meeting at a meaningful midpoint) serve most heat map purposes better.
Categorical heat maps—showing categorical data encoded by color—create even more color scale challenges. Using distinct hues for different categories with saturation representing magnitude is technically possible but cognitively demanding. Viewers must simultaneously decode hue categories and saturation magnitudes, then map both to the data dimensions they represent. When your data has two dimensions worth of categorical information, consider whether a small multiples approach (a heat map for each category) might communicate more clearly.
Geographic heat maps (choropleth maps) apply heat map principles to geographic regions. They're powerful for showing regional patterns in data—election results, climate measurements, population density—but require careful attention to map projection and region sizing. Large regions that dominate visual space don't necessarily dominate the phenomenon being measured; small regions with high values can be visually lost. Consider whether area-adjusted encoding might communicate more accurately.
Matrix heat maps arrange data in rows and columns with cell color representing values. These are common in genomics (showing gene expression patterns), correlation matrices (showing variable relationships), and calendar heat maps (showing daily values across months). The effectiveness of matrix heat maps depends heavily on row and column ordering—similar values should appear near each other to reveal patterns, which may require clustering or manual reordering.
Accessibility requires ensuring heat map information is available through non-color channels. Patterns visible in heat maps may be completely invisible to colorblind viewers using inappropriate color scales. Always pair heat maps with direct data access—hover tooltips, data tables, or annotations—rather than relying solely on color encoding. The goal is making patterns both visible in the color display and accessible through alternative means.
The right context makes heat maps sing. A user behavior heat map showing where visitors click on a website layout immediately reveals usability problems: are people clicking where you expect them to click? Are important navigation elements being ignored? A climate heat map showing temperature anomalies across a map tells a climate story in ways that tables of numbers never could. When your data has strong spatial or pattern-based stories to tell, heat maps may be exactly the right tool.