Overview
Where we came in
A specialist outdoor equipment retailer selling technical layers and boots. The team had planned a discount campaign to lift a flat product page. We asked to run interviews first, and eight sessions in it was obvious that price was not the blocker. People were not price-sensitive, they were size-uncertain, and returns data agreed.
The baseline problem
- Return rate on footwear was 27%, and 68% of those returns were logged as a fit issue.
- Interviewees repeatedly opened the size guide, then closed it without deciding, because it gave centimetres and they thought in their usual brand size.
- The size guide opened in a modal that lost the product page scroll position on mobile.
- Reviews contained good fit advice, but it was buried on page three of the review list.
Hypothesis
What we expected, and why
Written before the test was built. Because we observed X, we expect Y, measured by Z.
Because customers cannot resolve size uncertainty without leaving the page, we expect that surfacing fit guidance from real reviews next to the size selector will increase add-to-cart rate and reduce fit-related returns.
Experiment plan
How the test was set up
Sample size and stop date were both fixed before launch, so no result here was called early.
Two-arm split test on all footwear and technical layer product pages. Control kept the modal size guide. Variant added an inline fit summary next to the size selector, generated from structured review data, plus a brand comparison table that stayed on the page.
- Primary metric
- Add-to-cart rate on product pages in the test set
- Traffic
- ~31,600 product page sessions per arm
- Duration
- 24 days
- Guardrails
- Fit-related return rate over 90 days, the real point of the exercise
- Product page load time
- Revenue per visitor, so a higher add rate on cheaper items is caught
Results
What actually moved
| Metric | Control | Variant | Change |
|---|---|---|---|
| Add-to-cart rate | 7.74% | 8.99% | +16.2% |
| Size guide interactions | 18.2% | 9.1% | -9.1pp |
| Fit-related returns, 90 days | 18.4% | 13.7% | -4.7pp |
| Revenue per visitor | £6.41 | £7.28 | +13.6% |
| Product page LCP | 2.1s | 2.2s | +0.1s |
Size guide interactions fell, which reads like a loss until you notice why: people no longer needed to open it. The returns figure is the one that mattered commercially, and it kept improving for the full 90-day window.
What changed
Shipped to production
- Inline fit summary sits beside the size selector, built from structured review responses rather than marketing copy.
- A brand comparison table renders on the page instead of inside a modal.
- Fit-related reviews are pinned to the top of the review list.
- The discount campaign was cancelled and the budget moved to photography for the same category.
Next experiments
Already in the queue
- 01Test a short fit questionnaire that recommends a size directly.
- 02Test extending the inline fit pattern to the technical layers category.
- 03Test a free returns message on the product page against the current checkout-only placement.