Data color is a different discipline
Marketing color is about emotion and attention. Data color is about accuracy and clarity. In a dashboard, a poor color choice doesn't just look bad—it can make someone misread a metric and make a wrong decision. That raises the bar considerably.
The good news: data visualization has well-established color patterns. You just need to pick the right type for the right data.
Three kinds of data color
1) Categorical (qualitative)
For distinct, unordered categories (e.g., product lines, regions). You need colors that are clearly different from one another but equal in visual weight—no category should look "more important" because it's brighter.
- Cap your categorical set at about 6–8 colors. Beyond that, humans can't reliably tell series apart.
- Keep saturation and lightness roughly consistent across the set.
- If you have more than 8 categories, group them or use direct labels instead of more colors.
2) Sequential
For ordered data that goes from low to high (e.g., a heatmap of traffic). Use a single hue that ramps in lightness—light for low values, dark for high. The brightness ordering does the communicating, so it survives even grayscale printing.
3) Diverging
For data with a meaningful midpoint (e.g., profit vs. loss, above/below target). Use two contrasting hues that meet at a neutral middle. Blue-to-red through light gray is the classic example.
Semantic status colors: handle with care
Success-green, warning-amber, error-red are nearly universal, but they carry two big risks:
- Color-blindness. Red-green deficiency affects roughly 8% of men. Never encode status with color alone—pair it with an icon, a label, or a shape. A red dot and a green dot are identical to many users; a red "✕" and a green "✓" are not.
- Reuse. If green means "success," don't also use that same green as a decorative accent in charts. Mixed meaning erodes trust in the color.
Contrast and the dashboard context
Dashboards are read for hours, often on imperfect monitors. A few rules keep them comfortable:
- Use calm, low-saturation neutrals for the canvas so data colors stand out (the classic 60-30-10 split applies—see our 60-30-10 rule guide).
- Ensure chart labels, axes, and table text meet 4.5:1 contrast against their background.
- On dark dashboards, slightly desaturate chart colors so they don't vibrate against the dark surface.
Tables: the overlooked surface
Tables are mostly text, so color's job is structure, not decoration. Use subtle row separation (a faint border or zebra striping at low contrast), reserve strong color only for status cells and the occasional highlighted column, and keep numeric columns right-aligned with consistent formatting. The goal is scannability.
A starting point
Begin with a palette designed for data density rather than marketing flair. Our Dashboard palettes and Finance / Business palettes use restrained neutrals and clear accents that work well as a base for charts, tables, and status systems. Layer your categorical and semantic colors on top of that foundation, always pairing color with a non-color cue, and your dashboard will stay both beautiful and trustworthy.