Deborah Foley Portfolio Deborah Foley Portfolio

Amazon Developer

Case Study 02 · Amazon Developer — the input side

The engine behind the storefront

The B2B/SaaS platform where developers build, submit, test, publish, monetize, and analyze their apps for our app stores — across our 1P Alexa and 3P partner devices, plus Amazon.com and the Amazon Mobile native apps — and the multi-year effort to rebuild it into a self-service, developer-facing product held to a consumer-grade standard.

The consumer-facing app store is the output; the Amazon Developer portal is the input that powers it — a complex application used by everyone from small-business developers to Sparq VIP partners like Disney and Netflix. I led the design strategy, the research that made the case for a full rebuild, and the design direction — spotting the pain points and opportunities from years in developer portals — while a senior designer drove much of the hands-on execution under my direction and critique.

Outcome

Baseline-to-redesign usability testing showed high satisfaction with the monetization and submission workflows, and the long-requested analytics shipped as part of the redesign.

2023 – 2026

Amazon Developer

Head of UX, Amazon Appstore → 1P/2P/3P Enablement

As Head of UX for the Appstore, I owned the consumer and developer experiences — and, for the first three years, marketing too. Requirements came from every direction: product, the Tablet and Fire TV device teams, compliance, our “Cape” business-development partners, and the app developers themselves — who also build for Google and Apple, and could walk if we couldn’t automate what they needed. The constant challenge was how to prioritize initiatives across all these competing requests — from so many different partners and stakeholders.

Click any image to enlarge.

At a glance

Role

Head of UX, Amazon Appstore — strategic design leader and people manager who led UX for Amazon Developer and the Amazon Appstore on Alexa devices (including Appstore commerce), leading the design org for over three years; and, with AI, building production code myself

Scope

The Amazon Developer platform — the developer-facing (input) side; B2B/SaaS across 1P Alexa and 3P partner devices, Amazon.com and Amazon Mobile

Also leading

Concurrently led the consumer Appstore across devices (including its brand & marketing), plus the systems and tooling behind it — Fire TV & Fire Tablet design systems, internal enterprise tools for Fire TV (1P enablement), AI tooling for design-system automation, and 3P manufacturer HIG & support

Skills

Design strategy · Design systems · Platform HIG · AI design tooling · Information architecture · Stakeholder prioritization · People management

A note on my background: I’ve been lucky to have a genuinely wide-ranging career and portfolio — consumer-facing, award-winning products in banking; Chase’s more innovative work building data-aggregation and API developer portals; intranets; and enterprise work for a global law firm. At the Appstore I ran both sides: the consumer commerce store across many devices and applications, with its complex information architecture, and — what you’ll see here — the Amazon Developer application, where third-party developers come to build and submit their apps. Throughout, I’ve been as comfortable on the technical and program-management side — scoping and building the project plans that break huge initiatives into deliverables — as I am with ideation and design critique. Add a front-end background with real design-systems expertise — and a growing comfort with vibe coding and the automation tools that speed up workflows and delivery — and the Appstore was a rare fit: it used almost every hat I’ve worn, from strategy to leading a team across multiple orgs to maturing our design-ops practice. Consumer-facing or developer-facing; product, enterprise, and systems design; a design-and-program hybrid with real technical depth — whichever label a role leads with, the range is the point, and the proof is in these case studies rather than any single title.

On confidentiality. This case shows the case for the rebuild. The Redex design process and the finished redesigns were done under NDA and aren’t public yet; they’re described and shown abstracted behind the password. Reach out for access.

The Two-Sided Marketplace

One submission becomes the storefront

Every listing and subscription a customer sees begins as a form a developer fills here. I led the strategy and teams on the designs across the whole chain — the developer’s input, the automation that processes it, and the output: how it appears on the store and on the device.

This is a different kind of design than consumer UI. It sits closer to enterprise and service design — complex, systems-level thinking and digital transformation, where much of the job is defining how a submission gets automated from the backend out to the store. Service blueprints and journey maps did a lot of the heavy lifting, mapping what each form actually has to do, and the monetization side drew on service design for payment flows.

Developer · Input

Store Listing

The form where an app developer or marketer fills in the metadata and target markets and uploads the imagery.

Developer store-listing form
→
Customer · Output

Subscription detail page · Fire TV

It renders as the customer-facing detail page — its job is to persuade, then let the customer buy or subscribe and finish in the payment flow.

Fire TV subscription detail page
Developer · Input

Customer Preview form

Part of the new subscription form — where the app developer chooses the tiers (basic, premium, with or without ads) and the terms (monthly or yearly).

Developer customer preview form
→
Customer · Output

FTV subscription option selection

Those tiers and terms become the subscription options a customer selects on Fire TV before they subscribe.

Fire TV subscription option selection
The Starting State · 2023–2024

A decade of dev-led, feature-parity work

The portal had been built the way dev portals usually are: feature by feature, with siloed teams bolting each new request onto a form and no one holding the whole. We made incremental improvements, but the pattern had a cost — and it was time to name it.

Concept proposal — dashboard direction for the rebuilt developer portal
Concept proposal
The problem

Feature parity, no self-service, no single owner

Chasing competitor parity, with no self-service and no one accountable for the experience end to end, left the app with a tangled information architecture, poor discoverability, and long, difficult workflows that badly needed simplifying.

What I did

Argue for the overhaul — but earn it through the developers

I began making the case that the entire application needed a major information-architecture overhaul — the initiative we called Redex. That was a multi-year bet, and it had to start by proving why it was needed: research, a workshop, and a business case for stakeholders. Meanwhile, we kept prioritizing and delivering what developers actually wanted — the friction fixes and features they were asking for.

The nuance on parity. Chasing it wasn’t always wrong. Most of these developers had already submitted to Apple and Google, so matching familiar patterns gave them an easy learning curve and let them expand to our audience with no extra steps — and sometimes a Google-like flow genuinely helped them submit faster, with less friction. The skill was competitive analysis: knowing when to match, and when we could do better rather than chase the tail. Simplification was one of those calls — Google’s portal suffered the same “feature-add” creep into a clunky information architecture, and our redesign was the chance not to repeat it.

The current state · Pain points

Years of feature adds — bulky workflows, poor discoverability

Years of chasing competitor parity — one feature bolted on after another — left the app hard to navigate: weak discoverability, and core workflows stretched into long, unintuitive forms that got less efficient with each addition. Three screens made the case on their own.

Current state · Pain point

A marketplace, not a home

The homepage was a link farm out to different businesses — not the landing dashboard a developer should drop into and get straight to work.

Legacy marketplace homepage
Current state · Pain point

A submission form that never ends

App submission had grown into one long workflow, every team’s feature ask bolted on — exactly the kind of flow we needed to simplify.

Legacy app submission form
Current state · Pain point

A multi-release flow shipped as an MVP

The new multi-release flow went out as an MVP and it showed — not intuitive; we wanted to rebuild it end to end, folding in testing and sections for the different people who touch a release. This view is the start of the store-listing list: multi-releases often need edits to a listing’s metadata, so we decoupled the store listing from “create a new release” — letting each listing act as a reusable template a developer or marketer can pick from.

Store-listing list view — decoupling store listing from new-release submission

Delivered · Reports (2024)

The analytics developers kept asking for

Amid all that, we still delivered the thing developers asked for most. Until Reports, they had no analytics at all — just PDF downloads of their numbers. Reports (2024) gave them a real analytics UI, to positive reviews: it pulled developers back into the app — where we could surface features they were missing and unblock friction — and it earned the credibility to argue for the full rebuild that came next.

Reports · Landing analytics

Executive summary

The landing view of Reports — an at-a-glance overview of the KPIs our research showed developers most wanted first: installs, revenue, active devices, and rating.

Reports landing analytics
Reports · App health

App health

The quality of an app on the store — so a developer can act before a rating slips.

Reports app health
Reports · Targeting

Targeting & devices

Where installs land across the device family — so a developer can target the surfaces that matter most for their app and reach larger audiences.

Reports targeting and devices
We weren’t satisfied — these were exactly the problems a rebuild had to solve. So we kicked off Redex, our initiative to win a major revamp: run the workshop, build the business case, and show stakeholders the value of rethinking workflows, brand, and information architecture — not another year of parity.

The Redex deep dive — password protected

One initiative, start to finish

Amazon Developer is really one story: Redex, the multi-year initiative to win and design a full rebuild of the developer portal. The case for it is above. The design process that made that case — and the redesigns it’s producing — is available on request. There’s no results section here on purpose: winning the rebuild was the result, and the finished screens aren’t public yet, so the deep dive shows the thinking, not the shipped UI.

Password protected The Redex design process, end to end — this is the path the deep dive walks.

Research & discovery

Usability testing, competitor analysis, DSAT scores.

“Walk the store” workshop

Mapping pain and opportunity with product & research.

Personas & voice of customer

Two developers; the on-video, unfiltered truth.

The IA problem

Nav exercise, synthesis, the brand split exposed.

Concepts & AI prototyping

The Monitor UI, prototyped with AI.

The Figma proposal

Hi-fi mocks that won our scope.

Brand, systems & the call

Unifying the portal; I pitched the third path.

The full design-thinking process → proposal

Research, the “walk the store” workshop, the two personas and the on-video voice of the customer, the information-architecture fix, and the Figma strategy proposal that won our scope — the complete case for the rebuild, step by step.

Mega-menu concept for the developer portal navigation

AI in the work

Where AI entered the practice — the Monitor UI we prototyped with AI, and how I’m bringing AI into delivery through prototypes and concepts. Shown abstracted; one or two pieces are active rebuilds.

Illustrative concept from the AI work

Share Project :

  • Share Project :