0 to 1 service design for omnichannel retail
Took in-store pickup from a broken reservation program to a channel live in 1,000+ stores, by designing the digital flow and the in-store operation as one service.
Revenue
Picked-up dollars plus attached in-store sales.
Fulfillment rate
Share of pickup orders successfully picked up.
Stores
Pickup-enabled across the U.S. and Canada.
The Problem
Most reservations never became sales.
American Eagle and Aerie's legacy program, Reserve, Try, Buy, let customers reserve items without paying upfront. With no commitment and unclear pickup logistics, reserved items piled up in stockrooms.
The fix was a real Buy Online Pickup In-Store (BOPIS) system.
- abandonment rate
- 69%
- abandonment rate
- uncollected units sitting in stockrooms
- 223K+
- uncollected units sitting in stockrooms
What Shipped
Customers always know when, where, and how to pick up.
Pickup options at checkout, honest timing, and an update at every step, so nobody has to contact support or second-guess when their order will be ready.
Leading Discovery
I led experience discovery, starting with one question: what was Reserve, Try, Buy telling us?
the research
Went into stores to observe real people shop.
I partnered with UX research, joined store shopalongs, and interviewed customers and store associates directly.
the pace
Started building before the study wrapped.
My designers and I ideated and prototyped in parallel with the research, taking small steps so we could deliver value sooner.
Dual-Track Mapping
Digital and physical are one experience, so I mapped them as one.
Stockroom logistics, associate workflows, and the moment a digital order meets a physical store all shaped the design. Dual-track story mapping let me plan both together.
- Digital Browse: on-screen user stories
- Store Pickup: physical fulfillment for customers and associates
The Service Blueprint
Reserve, Try, Buy broke where the layers met.
I turned the story maps into a five-layer service blueprint. Customers could reserve items, but the backend systems and store operations weren't built to deliver a reliable pickup. The blueprint became the shared map for fixing all five layers at once.
Prototype Testing
Customers would rather pick up sooner than pick up closer.
I put working prototypes in front of customers throughout the design, and brought the store teams in early so they could feel the full flow. One usability study I ran with my researcher changed the core logic.
The system defaulted customers to the closest store, but the closest store often wasn't the one that could fulfill the order fastest.
The finding gave my product partner the evidence to change the default with confidence. I reframed the logic from "nearest store" to "fastest available pickup," with a clear prompt to trade a few extra miles for same-day speed.
- 01
Routed to the store that can fill it fastest
↳ more same-day pickups
- 02
A few extra miles, traded for same-day speed
↳ lower fulfillment costs, with orders going where the stock is
Pilot & Scale
A prototype can only simulate so much of a physical logistics problem, and the holidays were coming. So I pushed to ship something real, learn from it, and build on it.
pilot
Went live in four Pittsburgh stores.
With the design validated through iterative prototype testing, American Eagle and Aerie launched an early rollout pilot in the Pittsburgh area.
Four stores went live
store associates
Fixed the pickup alerts associates were missing.
Associates' pickup notifications weren't cutting through, so I drove a fix with the internal systems team.
Fix shipped with internal systems
customers
Answered the questions customers actually asked.
People wanted to know how long orders would be held and what happened if they never came, so I built those answers into the flow.
Hold times and no-shows explained in the flow
store ops
Every fix built more trust.
As problems got solved, store operations increasingly let us move faster.
The path from 4 stores to 1,000+
Post-Launch Evolution
Research kept running after launch, and it shaped what came next.
After launch, I set up a continuous research loop to keep learning from real pickups.
what it enabled
- In-Store Mode
- Curbside Pickup
Where It Fell Short
The MVP kept working, so the full experience never got scoped.
I wanted a customer to pick up what was ready today and have the rest shipped home, all in one checkout. Instead, the remaining items ship to store.
It works, but it isn't the experience I was chasing. It's the first thing I'd fund if I went back.
What I learned
The screen is only half the experience.
I came in thinking about flows and interfaces. I left thinking about stockrooms, associate workflows, and the moment a customer walks through the door. Omnichannel means designing all of it.
Answer the anxious questions first.
Research kept surfacing the same grey-area fears: who can pick up my order, where do I go in the store. Answering those upfront is what took pickup from 69% abandonment to an 84% fulfillment rate.
A pilot catches what a prototype can't.
The Pittsburgh pilot surfaced edge cases no wireframe could have predicted. I'd never skip a physical pilot again.
What I'd do next
Close the loop between pickup and the digital hand-off.
There's signal in the physical pickup moment that could shape the confirmation and notification flows.
Understand the other 16%.
An 84% fulfillment rate is strong. Is the rest a store operations issue, a communication gap, or a digital flow setting the wrong expectation?
Make the ready-for-pickup alert smarter.
Factor in store hours, estimated wait times, and hold windows so customers always show up at the right moment.
see it live
The pickup flow is live on ae.com.