Shirt finder on the homepage, content section removed: +12.18% sales
Commentary byLiftArchiveOur analysis, not theirsAdding a shirt finder to the homepage and removing a content-heavy middle section raised purchases by 12.18% for Thomas Pink.
Practicology ran this test for the London shirtmaker, and it was published as a VWO case study.
Both changes shipped together, so the lift can't be credited to the finder alone.
- Purchases (the primary goal): +12.18%.
- Revenue: +11.6%.
- Revenue per conversion: +14.2%.
- The test ran for 30 days on 136,000 visitors. No confidence level was shared.
The change
Shirt finder added and content section removed, pushing products up
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The homepage had a content-heavy middle section above the product images and links.
Variant
A shirt finder let visitors pick size, colour and style right on the homepage. The content-heavy middle section was removed, which moved products higher up the page.
Why it likely won
Our reading of the result, not the practitioner's claim.
- The finder copies the in-store habit of asking staff for a shirt in a given size, colour and style.
- Many Thomas Pink buyers were repeat customers who already knew what they wanted, so a faster path to it helped.
- Removing the content block cut distraction and brought products closer to the top.
Caveats: Two changes were bundled, so the source itself says the effect of each can't be separated. The test ran 30 days on 136,000 visitors, but significance was not stated.
Who ran it
Practicology ran this test for Thomas Pink 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 Practicology →
How to run this test on your store
- Hypothesis
- If we add a product finder on the homepage for the key choices shoppers make, then more visitors will buy because they reach the right product faster.
- Primary metric
- Orders per visitor
- Guardrails
- Revenue per visitor
- Finder usage and drop-off
- Return rate
- Who should test it
- Stores whose products vary on a few clear attributes, like size, fit and colour, with many repeat buyers. It fits less well for stores where shoppers mainly browse for inspiration.
Questions
Do product finder tools increase sales?
They can when shoppers know what they want, such as a shirt size and colour.
In this test the finder shipped with a layout change, so its own share of the lift is unknown.
Should I remove content from my homepage?
If a content block pushes products down and gets little use, it's worth testing without it. Check your scroll and click data first.
How do I measure each change separately?
Test them one at a time, or run a multivariate test with each change on its own and both together.
Similar tests
@DaveDiederen
Convertibles@DaveDiederen@DaveDiederen@DaveDiederenConvertibles