Color for e-commerce conversion: what the research actually shows

A lot of CRO advice on button colors is folklore dressed as science. Here is what the actual research shows about how color affects e-commerce conversion, and the much more nuanced picture it paints.

E-commerce Conversion Research

Highlights

The famous orange button vs. green button tests that circulate in CRO communities were run on specific sites in specific contexts. Color conversion effects are real but highly context-dependent — the same button color change can increase conversion on one site and decrease it on another with a different brand palette and audience.

Color contrast between a CTA button and its surrounding surface consistently outperforms color hue in A/B tests. A high-contrast button in a suboptimal hue will typically outperform a trend-color button that blends into the page.

Product photography background color has a measurable effect on perceived product quality and price tolerance. Neutral or slightly warm off-whites outperform pure white in A/B tests for premium products; pure white and high-saturation backgrounds perform better for value-positioned products.

What the button color research actually says

The most cited color and conversion research in e-commerce is the Hubspot orange vs. green button test from around 2011, which showed a 21% lift for orange. This finding has been repeated and generalized to the point of becoming design folklore: orange buttons convert better. But the actual research literature is far more qualified. Across multiple studies, the consistent finding is not that any specific hue converts better, but that contrast converts better. A button that visually separates from its surrounding context — through luminance contrast, hue contrast, or both — performs better than one that blends in. The specific hue matters far less than how much it stands out. This means that on a blue-dominant e-commerce site, orange may convert well simply because it is the highest-contrast option available. On the same site with a red brand palette, orange would no longer carry that contrast advantage.

Product photography and background color

While button color gets most of the CRO attention, product photography background color has a more consistent and larger effect on conversion metrics — particularly on perceived quality and price tolerance. The research pattern that emerges is roughly this: pure white backgrounds (#ffffff or close) communicate clarity and value-for-money positioning. They reduce cognitive load and let price comparisons feel straightforward. Slightly warm off-whites (around hsl(40, 10%, 96%)) read as premium without being sterile. They add perceived craft and materiality to products. Colored or textured backgrounds work best for lifestyle positioning — they add context and mood but reduce the directness of value assessment. Retailers that sell premium or artisan goods tend to outperform when they shift from pure white to warm off-white product photography, while value retailers see no benefit or slight negative effects from the same change.

Designing a color system that can be tested

The practical takeaway for e-commerce designers is less about specific color choices and more about building a color system that can be iteratively tested without requiring full design overhauls. This means defining your semantic color tokens with testable variants in mind: primary-action, secondary-action, and tertiary-action as separate tokens rather than hardcoded values. It means building your design system so that swapping a button color token affects every instance simultaneously rather than requiring manual updates. And it means separating brand colors (fixed, identity-defining) from conversion colors (flexible, test-ready) in your token architecture. The brand palette can remain stable while the conversion layer — button hues, badge colors, urgency indicators — gets tested and optimized over time.

Newer issue

A practical color forecasting workflow for product designers

2027-12-10

Older issue

Managing color fatigue in long-running brands: when to refresh and when to hold

2027-12-24