Cart Drawer Cross-Sell Case Study: Matching Accessories Cut Revenue per Visitor 12%
[ SUMMARIZE WITH AI ]
[ FREE CRO TEARDOWN ]
Find the 3 biggest revenue leaks on your store.
Every day a conversion leak goes unfixed, you're paying for traffic that doesn't buy. Get a 5-minute Loom through your PDP, cart, and checkout, with mockups of the fixes. No pitch.
Get My TeardownThis cart drawer cross-sell lost. Both versions. We ended it after 15 days and shipped nothing.
A premium bag brand sells straps and strap extenders next to its bags. A lot of customers already buy the two together. So the idea looked safe: put the right straps in the cart drawer, right before checkout, and let the order grow.
It did not work. Basket size held, but fewer people bought. For shoppers who opened the cart drawer, the best version trailed control by 10.3% in revenue per visitor, and the other by 12.4%.
We publish losing tests too. This cart drawer cross-sell result changed how we plan cross-sells for every brand with a high-ticket main product.
The Bet: Straps People Already Buy With the Bag
This was not a random "you may also like" row. We pulled the brand's last 30 days of orders and ranked every strap by how often it was bought in the same order as each bag. About one in three orders already had a bag and a strap together.
The rules were strict:
- Never show a strap that is sold out or already in the cart
- When a shopper adds a strap, the next one steps in
- Bags with too little order history get filled up with the best-selling straps
The cross-sell was as relevant as we could make it. That is what makes the result useful.
Two Placements, Same Picks
Control plus two variations in Intelligems, sitewide in the cart drawer, desktop and mobile, for 15 days.
| Version | What shoppers saw in the cart drawer |
|---|---|
| Control | The existing drawer, including the theme's single "pair with" product on some bags |
| Variation 1 | One "Made to go with your bag" row above the subtotal: the top 3 straps for the whole cart |
| Variation 2 | A "Made to go together" row under each bag in the cart: that bag's top 3 straps |
Each card had a small image, the title, the price and an add button. The full cart page stayed the same for everyone, so the test only measured the drawer.
Same Basket, Fewer Buyers
| Shoppers who opened the cart drawer | Revenue per visitor vs control | Probability to beat control | Orders |
|---|---|---|---|
| Control | baseline | - | 344 |
| Variation 1 (one row for the cart) | -12.4% | 11% | 309 |
| Variation 2 (row under each bag) | -10.3% | not reported | 312 |
The loss did not come from smaller orders. The people who bought spent about the same. The loss came from fewer people buying at all.
Both placements lost by a similar amount. When two different layouts of the same idea both lose, the layout is rarely the problem. The idea, at that moment, is.
Why a Relevant Cross-Sell Can Still Cost You Orders
Our read: an extra choice next to the checkout button slows down a bag purchase.
A premium bag is a considered buy. By the time someone opens the cart drawer, they have picked a color, a size and a price they are fine with. They are one click from paying. A row of straps reopens the decision: which strap, which color, do I need one at all? Some people stop to think, and some of them leave.
That is the trap with cart drawer cross-sells on high-ticket products. "Customers buy these together" is true. It does not mean they want to decide both at the same moment.
What We Test Next
The straps did not stop being a good pairing. We moved the offer to moments where it does not compete with checkout:
- After purchase: offer the matching strap once the bag order is placed, when the main decision is done
- As a gift: a strap as a popup gift offer, so it adds value instead of adding a choice
If you sell one main product with add-ons, test the cross-sell placement before you test the picks. For more on what usually works in the drawer, see our Shopify cart drawer optimization guide.
Cart Drawer Cross-Sell Questions
Do cart drawer cross-sells increase revenue?
Not always. In this test a well-targeted cross-sell kept basket size the same but lowered the number of people who bought. Revenue per visitor fell 10.3% to 12.4% for shoppers who opened the drawer. Test it before you roll it out.
Why would a relevant product recommendation lower conversion?
It adds a decision right before checkout. On a considered, high-ticket purchase, a new choice at the last step makes some shoppers pause, and some of them do not come back.
Is a row under each product better than one row for the whole cart?
In this test, neither won. The row under each bag trailed by 10.3% and the single row by 12.4%. Both layouts losing points to the timing of the offer, not its layout.
How long did the test run?
15 days, with 309 to 344 orders per version among shoppers who opened the cart drawer. The single-row version had an 11% chance of beating control, so we ended the test and shipped nothing.
Where should accessories be offered instead?
Our next tests move them after the purchase, as a post-purchase offer, and into a gift popup. Both keep the add-on away from the checkout decision.