Sam’s Club: Frequently Ordered Items
Used purchase history and the hooked methodology to shift in-club Sam’s Club shopping onto mobile — a single feature that drove a 3% lift in app GMV and $37.4M annualized.
- $37.4MAnnualized GMV
- 75%Engagement with triggers
- 4%CTR on landing trigger
Sam’s Club members buy the same things, on the same trip, almost every time. The question wasn’t what to recommend — it was when to remind them.
Convert in-club behavior into mobile behavior
Sam’s Club mobile had a clear mandate: take the high-frequency, predictable shopping patterns members already exhibited inside the club and surface them in the app. The business outcome was straightforward — lift GMV through the mobile channel — but the member-side problem was subtler. Existing app behavior was browse-led, while in-club behavior was list-led.
Three buckets, one researcher, one Concord club
Working with our internal research team, we mapped the member journey end to end. I spent time alongside a researcher inside a Concord club, watching how members moved through aisles, what they brought with them, and where the app could earn a place in the routine.
The findings synthesized into three buckets, which became the strategy.
- 01Member engagement
How and when does the app earn a place in the in-club routine? Push notifications were the only realistic channel that respected member intent.
- 02Shopping
Members rebuy the same items. Surface them — pre-grouped, ready to add — rather than asking members to search.
- 03Most-loved membership
Tie the experience to the values that make Sam’s Club membership unique, not generic e-commerce conventions.
Hook the rebuy, not the recommendation
We applied the Hooked model directly: trigger, action, variable reward, investment. The trigger was a timely, product-specific push. The action was a single tap into the app. The variable reward was whatever else the FOI list reminded members they were also out of. The investment was the re-order itself, which deepened the personalization loop for next time.
Design to the memory, not the algorithm.
The InVision pivot
We ran two rounds of moderated InVision testing inside the club itself. The first round dropped members from the notification onto the single featured product. It tested fine but felt narrow — members bought the item and exited.
In round two we changed one thing: the same notification dropped members onto their entire FOI list, with the featured item called out at the top. That single change turned a one-item purchase into a full re-stock.
100% rollout, $37.4M annualized
The feature shipped to 100% of Sam’s Club mobile members. 75% of members who received a trigger engaged with the FOI surface. 4% clicked through on the landing trigger itself. The basket size and re-trip frequency that followed drove a 3% lift in app GMV — $37.4M annualized.
The lesson I still apply: when behavior is high-frequency and predictable, design to the memory.