Identity Verification at Amazon
Lead Sr. UX Designer
Q2 2023 – Q4 2023
I redesigned identity verification across Amazon, bringing it in-house so customers never have to leave Amazon's environment. The new system maintained 80% completion rates while reducing Pharmacy cancellations from 10% to 2%.
The Problem
Proving your identity on Amazon used to be frustrating. Whether buying alcohol, filling a prescription, or recovering an account, customers had to leave Amazon's familiar environment and navigate through unfamiliar third-party systems — sometimes even updating their app just to verify who they were.
Behind the scenes, teams were struggling too. Long integration timelines, no control over the interface, and verification requests duplicated across different services meant we couldn't give customers the experience they deserved.
We saw an opportunity to bring identity verification in-house, making it as seamless as any other Amazon experience.
Get ready | Select ID type & state | Scan / upload ID | Verification in progress | Outcome | Complete order | |
|---|---|---|---|---|---|---|
| Feelings | Anxious and annoyed about getting the ID | Frustrated by the process, unsure they're doing it correctly | Hesitant and anxious about getting a good photo | Nervous and tense while waiting, hoping verification works | Relieved if successful, frustrated if errors occur | Relieved and confident for future orders |
😤 | 🤔 | 😰 | 🤞 | 😌😫 | 😊 | |
| Actions |
|
|
|
|
|
|
| Pain Points / User Needs | Pain Points: Surprise at ID requirement, Privacy worries, Physical effort, Concern about time Needs: Clear explanation, Checklist, Privacy assurance, Estimated time prominently | Pain Points: Uncertainty about valid IDs, Fear of choosing wrong Needs: Clear guidance, Quick FAQ | Pain Points: Shaky hands, Bad lighting, Physical limitations Needs: Step-by-step guidance, Lighting tips, Accessibility, Auto-adjust for low light | Pain Points: Anxiety waiting, Uncertainty about time, Fear of failure Needs: Real-time status updates, Clear data handling policy | Pain Points: Errors, Missing recovery steps Needs: Clear error fixes, Alternative paths, Explanation of failure | Pain Points: Remembering process, Future delay concerns Needs: Confirmation, Clear next steps, One-click future verification |
| Opportunities |
|
|
|
|
|
|
Creating a Unified Solution
I created a verification pattern that teams could easily plug into their services, with brand matching, accessibility, language support, and built-in security.
The flow is simple: customers enter from their main task, choose their country and ID type, allow camera access, and send their ID. The system supports over 200 types of government IDs worldwide, but customers only see what's relevant to their region and task.
Whether using auto-capture for instant verification or uploading a saved image, they're guided every step of the way and quickly returned to what they were doing.
The unified verification flow — simple, guided, and consistent across all Amazon services
Design Decisions
Branded where it matters, consistent where it counts
Accessibility built in, not bolted on
Guidance over error handling
Contextual hints and plain-language error messages guide customers through the process
Photo Capture Design
How We Validated
I relied on existing research — including past usability studies on the third-party experience — to inform the design. We knew where people struggled: unfamiliar interfaces, repeated permission prompts, and confusing error states.
We designed with those known pain points in mind and relied on post-launch metrics to validate. The data confirmed we'd matched the old system's 80% completion rate while dramatically improving the experience — and the Pharmacy cancellation drop from 10% to 2% told us the friction was genuinely gone.
Results & Impact
Completion rate
Matched the old third-party system's completion rate — but now with an experience we could control, iterate on, and improve.
Rx cancellations
For Pharmacy customers, ID verification for controlled substances went from a major hurdle to a minor step.
Reusable platform
Instead of weeks of custom development, teams could quickly plug into our system and know their customers would get a secure, reliable experience.
200+ ID types
Global coverage across government IDs worldwide, with customers only seeing what's relevant to their region and task.
"This makes what should be easier, actually easier for patients."
— SVP, Health
"Great to see us bring this UI in-house to deliver a better CX and open up our options, congrats team!!"
— VP, WW Buyer Risk Prevention
Looking Forward
Teams can now quickly add verification that's secure, accessible, and familiar to customers. Whether you're picking up medication, buying age-restricted items, or recovering your account, it all works the same way.
The system gives teams flexibility to customize what information to collect, which documents to accept, region-specific guidelines, verification attempt limits, capture methods, and allowed devices.
What I Learned
Control enables quality
Specific beats generic
Design for the platform, not the feature
More case studies
Care Coordination at Amazon Health
Lead Sr. UX Designer & Researcher
I led design strategy for Amazon Health’s first cross-service care coordination platform. Presented the work to SVP and VP leadership across three organizations and helped launch a system that connected thousands of people in just 4.5 months.
View case study
Designing COMET at Amazon Care
Lead UX Designer
Created the first comprehensive process map of clinical operations, revealing invisible chaos across 8+ teams and driving strategic platform decisions that scaled Amazon Care from 100K to 1M members.
View case study
Kindle Library Experience
Principal UX Designer & Design Lead
I led a platform-wide redesign introducing smart badges, automatic read/unread tracking, adaptive filters, and clear notifications to help readers organize and track hundreds of books. The improvements boosted the iOS app rating to 5 stars and helped millions focus on their next great read.
View case study