Matching straps cross-sell in the cart drawer: -10% to -12% revenue
Test run byConvertiblesCRO agency
Commentary byLiftArchiveOur analysis, not theirsShowing straps that customers often buy with each bag in the cart drawer lowered revenue per visitor by 10.3% to 12.4%.
Convertibles ran this test for a premium bag brand for 15 days.
Basket size held steady, but fewer shoppers completed a purchase, so nothing was shipped.
- Version 1 (one row for the cart): -12.4% revenue per visitor, with an 11% probability to beat control. A clear loss.
- Version 2 (row under each bag): -10.3% revenue per visitor. Also a loss.
- Order value stayed about the same. The drop came from fewer people buying.
- Orders per version ranged from 309 to 344 among shoppers who opened the drawer.
The change
A row of the top 3 straps bought with each bag was added to the drawer.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The existing cart drawer, which on some bags showed the theme's single 'pair with' product.
Variant
Version 1 added one row above the subtotal with the top 3 straps for the whole cart. Version 2 added a row of each bag's top 3 straps right under that bag. Picks came from 30 days of orders, ranked by how often each strap was bought with each bag.
Why it likely lost
Our reading of the result, not the practitioner's claim.
- A premium bag is a considered purchase. A row of straps next to checkout reopens the decision: which strap, which color, do I need one.
- Both layouts lost by similar amounts, which points to the timing of the offer rather than its design.
- Products that are often bought together are not always ones shoppers want to decide on at the same moment.
Caveats: The test ran 15 days with about 309 to 344 orders per version, which is a fairly small sample. Only the cart drawer was tested; the full cart page stayed the same.
Who ran it
Convertibles ran this test for a premium bag brand and published the results (Convertibles case study). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on Convertibles →
How to run this test on your store
- Hypothesis
- If we show accessories that are often bought with the main product in the cart drawer, then revenue per visitor will rise because shoppers add a relevant item before checkout.
- Primary metric
- Revenue per visitor among shoppers who open the cart
- Guardrails
- Conversion rate
- Average order value
- Checkout start rate
- Who should test it
- Stores with low-priced main products and quick add-ons may see a different result. For high-ticket, considered purchases, test placing the accessory after purchase or as a gift before putting it next to checkout.
Questions
Do cart drawer cross-sells increase revenue?
Not always. In this test a well-targeted cross-sell kept order value the same but lowered the number of buyers, so revenue per visitor fell.
Why would a relevant recommendation hurt conversion?
It adds a new choice at the last step.
On an expensive purchase, some shoppers pause to think and do not come back.
Is a row under each product better than one row for the cart?
Neither won here. Both layouts lost by a similar amount.
Similar tests
@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon@oliverkenyon