Skip to content

Test archive / Product page

Does a redesigned mobile product page layout lift conversion?

The mobile product page for a teeth whitening kit used the original layout. We tested 4 new layouts that moved the button up and introduced the product step by step. Conversion moved 1.7% lower over 11 days, short of our confidence bar.

Not significant Mobile 11 days Not called

SURFACE
Product page
DEVICE
Mobile
VERTICAL
Ecommerce (B2C)
PRIMARY METRIC
Conversion rate
VARIATIONS
4

The hypothesis

By designing a new page layout from what earlier tests taught us, we will match what visitors need on the product page, resulting in more add to carts and orders.

The first rounds of tests on this page showed which images and messages worked. We used them to plan a full redesign of the mobile layout.

The main layout introduced the product in order: image, name, reviews, details, price and discount, a clear button, and a message to ease worries before buying.

What we changed

We tested 4 new mobile product page layouts against the original.

Variation A moved the button above the fold and introduced the product step by step. Variations B and C changed how kit options were shown, as cards or a drop down. Variation D used a splash page layout. B and C were paused before the end.

CONTROL
Control wireframe
VAR A: ABOVE THE FOLD
Var A: Above the fold wireframe
  • Button moved above the fold
  • Product introduced step by step
  • Message to ease worries added
VAR B: PRODUCT OPTIONS
Var B: Product options wireframe
  • Kit options shown as cards
  • Base kit and add-ons called out
VAR C: PRODUCT DROP DOWN
Var C: Product drop down wireframe
  • Kit options in a drop down
  • Base kit called out
VAR D: SPLASH PAGE
Var D: Splash page wireframe
  • Full width product image
  • Reviews, name and button below it

The result

Conversion moved 1.7% lower over 11 days, too small to call.

CONVERSION RATE
-1.7%
REVENUE PER VISITOR
about -9%

Conversion moved 1.7% lower and revenue per visitor about 9% lower. The main layout, Variation A, was almost flat against the control, and the result never reached our confidence bar.

We paused the test after 11 days so the mobile checkout test could go live without overlap.

What we would do next

Break the redesign into smaller parts and test each.

Each layout made many changes at once. Some likely helped and others hurt, so they cancelled out. Moving the button above the fold rarely loses, so we ran a follow-up that added the changes from Variation A 1 at a time.

Let us find where your conversion opportunities are.

Request an evaluation