Only 2% of users were navigating past the first tab in our 3-tab recommendations section. While the design was space efficient, it was just not working. We dropped the tabbed structure and rethought how recommendations were surfaced.

Users were scrolling vertically far more than horizontally, which is typical user behaviour. We made the switch knowing it would increase page length, but with ~70% of users already reaching the bottom of the cart, the tradeoff was acceptable.

Despite low tab engagement, the Favorites tab had the highest ATC rate. So we moved those suggestions upfront on the page and reinforced education around other Favourites entry points in the cart.

There was no existing space for monetised SKUs on the cart, and mixing it with personalised recommendations would conflate two different intents. To avoid that, we carved out a dedicated section for it to keep the two distinct.

We noticed that a few of our competitors were auto-adding products to cart, which is a dark pattern, and received backlash for good reason. We wanted our recommendations to feel helpful, not manipulative, so the suggested items are always kept visually distinct from what's already in the cart.

While the brief was just for the cart page, this feature needed to reflect across Checkout, Addresses, Order Confirmation, and Order Tracking pages. We expanded scope deliberately to build a consistent end-to-end experience.

The Order Savings Summary section was being built to drive affordability perception, and no hard business metric was attached to it. This made it difficult to get engineering bandwidth, which directly shaped how much time we spent designing and how lean we kept the development ask.

The cart is seen by all users. Adding new sections and reorganising a high-intent functional page carried real risk. Testing each addition independently, and ramping recommendations gradually, gave us the confidence to roll out everything.