Skip to content
Bixlow
A/B test idea · Category page

A/B test: show a second image on listing cards

If each listing card lets shoppers see a second image, with the product in use or a close-up, then more people will click through to the product page, because a single studio shot on a white background does not show how it really looks.

By Fernando Blanco Dosil · Updated November 2, 2026

Why it might work

A bag on a white background looks a lot like any other bag on a white background. What makes someone tap the card is picturing themselves with it: seeing it on a shoulder, seeing its size next to a person, seeing the clasp up close.

On desktop many stores already swap the image on hover. On mobile there is no hover, so nobody sees that second image. That is where the test is.

How to set it up

In Bixlow you describe the change to the agent, it builds it on your site and launches it as an A/B test once you approve it.

  1. Check that your products have a second image that adds something: in use, on a person, or a detail. Another studio angle adds little.
  2. On mobile, the variant can be a two-image swipe inside the card, with dots showing there is more.
  3. On desktop, swap the image on hover and also on keyboard focus.
  4. Run it on category listings and split mobile and desktop in the analysis.

What to measure

Metric
Primary metricClicks to a product page from the listing
Metric that must not get worseListing load speed and orders per visitor
Where it is measuredCategory listings, by device
FAQ

Second image on the card

Which image should be the second one?

Whichever answers what the first one doesn't: how it looks worn, how big it is, what the fabric looks like close up. For clothing, almost always the shot on a model.

Won't it slow the page down?

It can, if you load every second image at once. Load them when the card scrolls into view or when the shopper swipes.

Does it work the same on mobile and desktop?

They are different experiences (swiping versus hovering), so read the results separately. It can win on one and not the other.

START WITH YOUR SITE

What would you test first on your site?

Bixlow analyses your site for free and proposes your first hypotheses, ranked by impact.