Orange Buy Now buttons site-wide: +5% clicks, sales inconclusive

Flat / inconclusiveSite-wide

Orange Buy Now buttons across the whole site lifted engagement by 5%, but the effect on sales was not proven, so we count this test as inconclusive.

GSM.nl, a Dutch phone and accessories store, ran this test, and it was published as a VWO case study.

Engagement here meant any click on the page.

+5% clicks, sales inconclusive
  • Engagement (any click on the page): +5% for orange buttons, reported as statistically significant.
  • Sales: up, but not statistically significant. This is why we treat the test as inconclusive.
  • The test period was short, and no sample size was shared.

The change

Control (A)
header
nav
hero
Featured phone
Buy Now, original styleREPLACED
Variant (B)
header
nav
hero
Featured phone
Buy Now, orangeNEW

Every Buy Now button on the site is restyled orange.

Simplified mockup drawn by LiftArchive from the shared screenshots.

Control

Every Buy Now button on the site, on home, product, catalog and offer pages, used the original style.

Variant

Alternate stylesheets restyled every Buy Now button at once. Three versions were tested: text links, green buttons, and red (in practice orange) buttons.

Why it likely came out this way

Our reading of the result, not the practitioner's claim.

  • A bright, contrasting button is easier to spot on pages with many products, so more people clicked.
  • A click is not a purchase. Price, shipping costs and discounts weigh more on buying a phone than button color.
  • The short test period likely left too little data to detect a small change in sales.

Caveats: Counted as inconclusive: only the click lift was significant, and the sales result was not. The test ran for a short period and no sample size was shared.

Who ran it

GSM.nl ran this test for GSM.nl and published the results (VWO case study). The numbers are theirs; the summary and commentary on this page are ours.

Read the original case study on GSM.nl →

How to run this test on your store

Hypothesis
If we make our buy buttons a high-contrast color across the site, then more visitors will click through and buy because the next step is easier to see.
Primary metric
Conversion rate (orders per visitor)
Guardrails
  • Click-through rate on buy buttons
  • Revenue per visitor
  • Bounce rate
Who should test it
Stores with many buy buttons per page whose buttons blend into the design. Run it long enough to measure sales, not just clicks. If your buttons already stand out, expect little change.

Questions

Does button color affect conversion rate?

It can change clicks, but the effect on sales is often small and hard to prove.

Here orange buttons lifted clicks 5%, while the sales gain was not significant.

How do I test a button style on every page at once?

Change the button style in a stylesheet and split traffic between stylesheets.

GSM.nl did this so every Buy Now button changed together.

Similar tests