Business Problem
The cross-sell section was already present in the cart, but it behaved more like decoration than a real sales tool. Users saw additional product suggestions, yet too rarely added them to the order. From a business perspective, that meant lost upside: the traffic and purchase intent were already there, but average cart value was not growing as much as it could.
This was not a roadmap item handed to me. I proposed the experiment proactively in an internal feedback channel, prepared the hypothesis, and took ownership of the entire test.
What I Tested
I designed an A/B test for the cross-sell drawer with three variants:
- the control version, meaning the existing solution,
- a version with a redesigned drawer interior,
- a wide dialog version with more room for product information, pricing, and CTA.
The hypothesis was simple: if a user is already in the cart with clear buying intent, the interface should make the decision to add another product as easy as possible. Instead of asking "do we have cross-sell?", I focused on "does this interface actually help close incremental revenue?".
Implementation
I implemented the experiment in a way that allowed all three variants to be compared under the same business conditions. I also made sure the measurement covered not only clicks, but the metrics that matter downstream: interactions, cart views, and completed carts.
The key product decision in the winning variant was to improve space and information hierarchy. Users could understand faster what they could add, why it was relevant, and how to take action without leaving the purchase flow.
Results
The winning variant versus control delivered:
- +122% more products added to cart,
- +133% more interactions,
- +21% more cart views,
- +86% more completed carts, meaning submitted orders.
This experiment proved that even a relatively small checkout component can materially improve revenue if it is treated as a sales mechanism rather than just another UI block. It also reflects how I work: I do not wait only for assigned tickets, I proactively look for places where better product thinking can unlock business results.
