Related category links on collection pages: -0.4% conversion
Test run byClean CommitCRO agency
Commentary byLiftArchiveOur analysis, not theirsAdding links to related categories on collection pages left conversion rate almost unchanged at -0.4%, while average order value fell 9.9%.
Clean Commit ran this test for an anonymized Shopify store.
It ran only 3 days on 6,267 visitors, so the result is inconclusive.
- Conversion rate moved -0.4%, with a 49.7% probability to beat control. That is no difference at all.
- Average order value fell 9.9% and revenue per visitor 9.5%. Both are negative, but too early to trust.
- Add-to-cart rate rose 6.0% and bounce rate fell 5.6%.
- Devices split hard: conversion fell 7.1% on mobile and rose 24.7% on desktop.
The change
The variant adds links to sibling categories on collection pages.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
Collection pages had no links to other categories.
Variant
Collection pages showed links that followed the existing navigation. Main category pages, such as Men or Women, linked to all main categories, and subcategory pages linked to their sibling subcategories.
Why it likely came out this way
Our reading of the result, not the practitioner's claim.
- Links to sibling categories help shoppers who landed in the wrong place, but most visitors may already be where they want to be.
- On mobile, an extra row of links can push products down, which may explain the mobile drop.
- Three days is too short to smooth out daily swings, so the large device split may be noise.
Caveats: Too short to call: 3 days and 6,267 visitors. Results by device pointed in opposite directions.
Who ran it
Clean Commit ran this test for a Shopify store (anonymized) and published the results (Clean Commit A/B test library). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on Clean Commit →
How to run this test on your store
- Hypothesis
- If we add links to related categories on collection pages, then shoppers will explore more and convert more because they can move between collections easily.
- Primary metric
- Conversion rate
- Guardrails
- Average order value
- Revenue per visitor
- Conversion rate by device
- Who should test it
- Stores with several sibling categories that shoppers often compare, like Men and Women or Tops and Bottoms. Run it for at least two full weeks, and check mobile on its own.
Questions
Do related category links help collection page conversion?
Not clearly in this test. Conversion barely moved, and order value fell 9.9% over a short run.
How long should a navigation test run?
At least one to two full weeks, to cover weekday and weekend patterns.
This test ran 3 days, which was too short to call.
Why do mobile and desktop results differ so much?
Extra links take up more of a small screen and can push products down.
On desktop they fit more easily, so the effect can differ by device.
Similar tests
Blend CommerceBlend CommerceBlend CommerceBlend Commerce
GrowthRock