Same-category recommendations on PDPs: -9.9% CR, +19.7% AOV
Test run byClean CommitCRO agency
Commentary byLiftArchiveOur analysis, not theirsAdding a block of other products from the same category to product pages cut conversion rate 9.9% while average order value rose 19.7% and revenue per visitor 7.9%.
Clean Commit ran this test for an anonymized Shopify store and published it in its A/B test library.
We count it as inconclusive: no probability was computed, and the revenue gain looks driven by a few large desktop orders.
- Average order value (the target): +19.7%, but mostly from desktop, where a few large orders likely drove it.
- Conversion rate: -9.9% and orders -10.1%. Negative: fewer shoppers bought.
- Revenue per visitor: +7.9% overall, so the net effect on sales is unclear.
- Mobile, with 7,391 of the 9,093 visitors: conversion rate -5.1% and revenue per visitor -4.7%. Negative.
The change
The variant adds a block of other products from the same category.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
Product pages with no product recommendations.
Variant
A recommendations block showing other products from the same category as the one being viewed.
Why it likely came out this way
Our reading of the result, not the practitioner's claim.
- Products from the same category are alternatives, not add-ons, so they may have pulled shoppers into comparing instead of buying.
- More choice on the page can delay a decision, which fits the drop in conversion.
- Desktop order value rose 153.5% on 1,702 visitors, which is the pattern a handful of big orders produces, not a steady shift.
Caveats: Mixed and directional only: 9,093 visitors and no probability computed. The revenue gain leans on a few large desktop orders, while mobile, most of the traffic, went negative.
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 show related products on product pages, then average order value will rise because shoppers see items to add to their order.
- Primary metric
- Revenue per visitor
- Guardrails
- Conversion rate
- Mobile revenue per visitor
- Orders
- Who should test it
- Stores where shoppers often buy several items from one category, like apparel basics or consumables. If same-category items are substitutes, test complementary products instead, since alternatives may cost you sales.
Questions
Do product recommendations increase average order value?
They can, but here order value rose while conversion fell.
Judge the test on revenue per visitor, not order value alone.
Should I recommend similar products or complementary ones?
Similar products compete with the one being viewed and can cause second-guessing.
Complementary items are more likely to be added alongside it.
How can a few big orders skew a test?
On small samples, a handful of large orders can swing average order value a lot.
Check the median order or remove outliers before trusting an AOV lift.
Similar tests
@DaveDiederen@DaveDiederen@DaveDiederen
Blend CommerceBlend CommerceBlend Commerce