Letting suppliers run their own brand storefronts — without the months-long relay.
As the sole designer and researcher, I took it from a blank page to a validated alpha — shipping a metrics dashboard, then a guided page builder, on the way to a one-week publishing cycle.
RoleSole product designer & researcher
CompanyThe Home Depot
Timeline2022–2026
SurfaceInternal B2B web platform
Users200+ supplier brands + internal teams
StatusDashboard shipped · portal in alpha
Fig. 01 — The self-service page builder. Suppliers compose their storefront module by module, with
guidance built into every step.
01The wait
Eight to ten weeks. That's what a supplier waited to change a single line on their own brand page.
Brand pages are the storefronts suppliers run inside homedepot.com. Every change went through a
managed service: the supplier sent content to a sales contact in a spreadsheet, which relayed through
ad operations, SEO, design, and a build team before anything reached a customer. One update touched
eight internal roles, a dozen tools, and six systems, each holding a piece of the truth.
Suppliers provided everything — content, creative, brand direction — and could touch none of the
platform that published it. The bottleneck was the relay itself, not any one team.
8–10wks
To publish one change
8
Internal teams involved
6
Disconnected systems
1×/ qtr
Updates allowed per supplier
02What I found
Discovery started outside the building — a competitive scan of self-service publishing tools, to see
what good looked like. Then inside it: I interviewed the teams who ran the managed service and mapped
the whole thing into a current-state journey. One session reframed the product. A supplier mentioned
they ran their own Amazon storefront, so I had them walk me through it screen by screen — and there
was the model, already working in their hands.
Two findings set the direction. Every update needed eight teams to coordinate, with no one accountable
end to end — that's why publishing took months. And once a page went live, suppliers had no way to see
how it performed. I started with the visibility gap: it was the highest-risk, and the fastest to prove.
03The design
Suppliers were pouring real investment into these pages and couldn't see how any of it performed. That
was the program's biggest risk, so it's where I started.
The performance data existed; it just wasn't reachable. Even internal teams booked a monthly call to
pull basic numbers. I shipped a metrics dashboard that put traffic, orders, conversion, and gross
demand one click away — plain-language glossary for every term, one-click export. It's in daily
internal use today.
Fig. 02 — The metrics dashboard. Shipped; in daily internal use.
Then the harder half: the builder itself. Instead of routing content through eight teams, a supplier
assembles the storefront directly — module by module, with best-practice guidance, asset specs, and
accessibility prompts built into each step. Draft-saving and a live preview mean nothing ships until
they're ready. The know-how that lived across eight people now lives in the interface.
04Outcome
The dashboard shipped. The builder is validated. A months-long relay is on track to become a one-week,
self-serve job.
The dashboard is in daily use by the internal teams. The builder went through two rounds of alpha
testing — five brands over roughly ten weeks — which proved the self-service model with real
suppliers. The cycle it's built to hit is one week, down from the
eight-to-ten-week average.
The last round closed with a UMUX-Lite usability score of 85.6 out of 100 — a strong result for a
workflow suppliers had never run themselves before.
“
In my other systems, I need IT support to make a change. Here, I can do it all myself. I'm not a
designer or a developer — it's really not that tricky.
— Supplier, alpha testing
I've since moved to another team. The program continues toward general availability on the foundation
I built: the research, a shipped dashboard, and a validated self-service model ready for the next
phase.
05Where it's headed
The validated MVP opened a bigger picture. Two directions I prototyped for what comes next:
Concept Manage a full set of pages, with a live desktop and mobile preview before publishing.Concept A richer, microsite-style storefront — more expressive than a templated page, still self-served. Click to view full.
06What I'd change
If I were starting over, I'd move faster. Ship the smallest useful version sooner and let supplier
feedback pull the rest forward, instead of trying to get more of it right up front. The signal we got
in alpha is exactly what I'd want driving the work from week one.
