JD Sports - My Account
Where this started
Not a brief. A pattern I had to make the case for.
Biweekly CS reports, hundreds of thousands of contacts a month. WISMO — “where is my order” — was always the highest.
The PO agreed it mattered. It wasn't a priority: the business was weighted toward conversion, not cost.
It had been tried before. The old stack made it too costly to fix properly. The 2023 data-breach replatform opened the window.
Agreed success criteria with CS and PM: 5% fewer WISMO contacts. The old “dashboard” was personal details. Nothing else.
Not a content problem. A structural one.
The volume pointed to a self-serve failure. A content fix felt obvious — nobody had questioned the structure itself.
The information customers needed already existed on the old My Account. The problem was structural, not visible.
The assumption nobody tested
PM and CS shared a belief: My Account was a low-engagement surface, and customers would self-serve through the footer link instead. I suspected it was wrong.
The old hierarchy was built on business logic, not customer logic — order tracking sat level with newsletter preferences.
A content fix would have shipped, and missed the real cause.
The scope widened
Worked it with the CS lead first. The real complaint: tracking links dropped customers into a manual lookup, and status rarely moved past “on its way.”
Brought in PM, BA, TPO and tech lead. That's when Narvar surfaced — a third-party vendor already being evaluated for accurate tracking status.
Two workstreams landed together. The IA fixed where to find it. Narvar fixed what it said.
The Narvar work
From a design perspective, I worked with the internal Narvar team to map delivery and returns statuses, and get the wording right for customers. I also shaped the content and links in Narvar's order emails and texts — kept the experience unified end to end.
Making the case with evidence, not opinion
A designer I lead ran the competitor review — ten sites, nearly all led with orders and returns. I reviewed CS data and ran a card sort to test the mental model directly.
The result: order tracking, returns and delivery grouped at the top. Unambiguous.
What it changed: the PM expected the research to confirm the existing structure. It didn't. CS wanted a content fix. The research argued against that too.
The new hierarchy was A/B tested for click-through pre-launch. Contact volume was the real-world result, tracked after launch.
What I explored
Four concepts, three shown as screens, user-tested with 100 customers: a large STATUS banner, cumulative spend and Top Brands, and an early gamification concept.
An acquisition business — I wanted My Account earning engagement value too.
The Wallet Concept
Tested by hypothesis, not a shown screen.
Two audiences: parents topping up through the year with STATUS cash as an incentive, and savvy younger customers saving pre-holiday while avoiding BNPL debt. A gateway into a linked parent-and-child accounts concept, with an LTV idea behind it.
The research: 100 customers via usertesting.com. A split response — some wanted it, others wouldn't hand JD their savings interest. The business reached the same conclusion independently. Easy no.
Even so, I pushed for saved payment methods. Flutter wallets cut time-to-bet — betting's regulated and faster-paced, but I judged the parallel held. The business wanted it too. That's what shipped.
What the research told us
A mixed result. Not a clean win, not a clean rejection.
STATUS cash balance: customers liked seeing what they'd earned. Confused they couldn't spend it online.
Total spend + Top Brands: the spend figure caused unease, a risk of suppressing purchases. Top Brands worked — personalised, purchase-driven.
Gamification: liked the ability to earn free STATUS cash, but didn't think My Account was the right location for it.
I took the customer STATUS spend concerns to the trade director, with the research behind me. The STATUS team held their position — spending in-store was better commercially. They won. Status cash still can't be spent online.
What shipped
Restructured around the card sort findings - not just reordered
Dashboard removed — no perceived value, inconsistent across platforms. Returns folded into Order History, tested via a tree test. Payment methods added — quicker saved-card checkout.
Unified responsive build — one domain, replacing split mobile and desktop URLs.
Engineering, through a reshuffle
The team who'd build this didn't exist yet. JD had just sold its engineering function to a third party — my old contacts were gone.
I brought the new tech lead in early: data, findings and designs, before a single ticket.
My part: two sprints. Then it sat in the engineering backlog. Live within two months regardless.
What we didn't get
An honest trade-off, not a polished win. My Account sits behind “More…” on iOS and Android. I pushed to elevate it.
JD had a public data breach in 2023. The replatform was the security response — every function resourced against it. The app was ugly, but secure; web wasn't.
Web sat inside a design system built by a team of four, me included — accessibility-checked, 98% Axe pass rate. I also started a native app design system. It never shipped. The restructure was scoped to follow web.
Monthly product forums put me in the room with SLT. Most months, the app trade managers pushed for investment and got shut down. Not my call to make — but watching that pattern told me not to spend more effort pushing it.
Fewer contacts, fewer failures
~100,000 WISMO contacts a month, pre-launch. ~12,000 fewer contacts a month after launch — a 12% reduction against a 5% target.
At roughly £3 per contact, confirmed with the CS lead, that's ~£36k a month, or ~£432k a year annualised.
Stable baseline. IA and Narvar landed together — no other material WISMO initiatives shipped in the same window. The result can't be cleanly split between the two, and I'd rather say that than invent a split.
What I'd do differently
The decision to deprioritise the app was right.
What I'd change: I never wrote down why — it just lived in the room.
Next time — document the trade-off. Not to ask permission. So the reasoning outlives the meeting.