Establishing Content Governance: Standardizing Terminology for Amazon One

Context

About eight months into my role as the first UX writer for AWS Physical Stores, I noticed a big problem.

Inconsistent terminology across Amazon One’s product and support content was causing real user confusion. Terms like “sign up” and “enroll” were used interchangeably. “Amazon One ID”, “profile”, and “account” appeared without a clear pattern.

Amazon One is a contactless identity service that lets customers check in or pay by simply hovering their hand over a palm-scanning device — no phone, card, or wallet needed. Being a biometric identity and payment product, trust and clarity are paramount for Amazon One. Inconsistent terminology was more than just a copy issue; it was a brand issue.

At the time, the UX team was preparing for a full website content refresh, including FAQs and help articles. I saw an opportunity to get ahead of any conflicting language within the scope of that project and propose standardized terminology before new work shipped.

01 — Problem statement

For a product built on biometric identity, users needed clarity and reassurance at every touchpoint. However, inconsistent language and unclear guidance made Amazon One feel more technical than trustworthy. Before standardization, inconsistent terminology like "ID," "Account," and "Profile" caused user confusion and hesitation in a critical, high-trust context.

01 — Project goal

To earn trust and drive adoption by establishing a shared lexicon and scalable content system that makes advanced technology feel simple, safe, and human

02 — Design principles

Lead with clarity to earn trust.

Make advanced tech feel human, not mechanical.

Keep guidance simple, approachable, and consistent.

02 — Governance Process

Audit all language for inconsistencies and trust gaps

Host content workshops to align Legal, Product, and Design on shared terminology

Create naming conventions and terminology guidelines for consistency

Build a scalable content system for naming, tone, and states

The email I sent to stakeholders — up to director level — spanning Legal, PR, Marketing, and Product.

It's worth noting,

I wasn’t proposing brand-new terminology. All of these terms (sign up/enroll, Amazon One ID/profile/account) already existed in the product and its support materials. The issue was that they were used interchangeably, without a clear rationale for when to use each. As Amazon One’s features and customer touchpoints expanded, those inconsistencies would only become more visible and more confusing. My goal was to proactively define, document, and align on preferred terms so every team could speak the same language going forward.

Sign up → Enroll

One of the reasons I proposed sign up become enroll in the context of Amazon One was because Amazon ("Big Amazon" as it's known internally) uses sign up. Customers using Amazon One already have an Amazon account. Asking them to “sign up” again during palm enrollment was confusing — it implied they were creating a second Amazon account rather than completing a separate biometric setup.

Switching to “enroll” created a clear distinction between signing up for Amazon and enrolling in Amazon One. It also aligned with other identity- and membership-based services, making it more intuitive for customers.

  • Helps differentiate between (big) Amazon and Amazon One

  • Supports a new onboarding state: “pre-enrolled” for users who completed online steps but hadn’t yet completed their palm scans in person

  • Has more secure and official connotations

The main landing page, technically owned by Marketing
I was even willing to let marketing keep their exclamation point, in the name of compromise
There was significant confusion around what customers are signing up for – Amazon or Amazon One.
Changing AO to exclusively using “enroll” language helps mentally differentiate between the two products for the customer.

When Amazon One introduced online enrollment, it created a new “in-between” stage in the user journey: customers could start the process online but still had to complete it in person with a palm scan. The existing “success screen” (not pictured) gave a false sense of completion, leading many users to believe they were fully enrolled when they weren’t. This misunderstanding not only eroded customer trust but also led to an exceptionally high drop-off rate between starting enrollment online and finishing it in-store.

Another challenge was that the existing “sign up” terminology didn’t allow for a natural way to name this interim state — “pre-signed-up” felt awkward and unclear. Switching to enroll made it easy to introduce “pre-enrolled”, a term that clearly signaled a profile had been created but the process wasn’t yet finished. This helped set expectations, reduce drop-offs, and cut down on support inquiries.

"Pre-signed up" is a weird and unfamiliar term
"Pre-enrolled" is already a familiar concept
Passive voice :(
Active voice :)

A wrinkle in roll-out

The problem was getting alignment outside our org. Our Marketing team, which was very siloed, owned the high-traffic landing pages. Their main KPI was top-of-funnel clicks, and they were worried 'Enroll' sounded too heavy and would kill conversion. And honestly, they never did get on board. If you go to the Amazon One landing page today, you'll still see 'Sign up.'

But, for me, it’s okay, because users don't enroll on the landing page. They enroll in the product. My focus was on getting alignment with Product, Eng, and Legal for the actual user experience. We made the in-product system consistent, and that was the win.

Amazon One ID → Amazon One Profile

Why?

At the time, ID, account, and profile were used interchangeably throughout the Amazon One experience. Standardizing on a single term — profile — would create consistency, reduce ambiguity, and better align with the way users naturally think about storing their personal information.

"Profile" is a much more familiar term within the account-creation mental model than "ID"

Can I help you?

This wasn't just a UI problem; it was a nightmare for our Customer Support team. When it came to account issues, the internal help docs, which were made by another super-siloed team, were using such inconsistent terminology that the support reps could hardly determine the customer’s problem, let alone tell them how to fix it. 

A user would call about their 'ID,' the rep would look up their 'Account,' and tell them to check their 'Profile.' It made us look disorganized.

So, after auditing hundreds of those support articles, I went to that team, showed them their own call-log data, and pointed out how much time their reps were wasting just playing translator. Once they saw how standardizing on 'Profile' would make their lives easier, they became my biggest allies.

Users were invited to link other rewards memberships, such as Starbucks, to AO so they could pay with their palms.
Most other memberships also use the term "profile" rather than "ID"

Changing 'ID' to 'Profile' was also a critical future-proofing move.

Additionally, the team was on the verge of launching ID Verify.

ID Verify would allow users to scan their valid government-issued ID in order to use their palms to indicate age eligibility at participating locations. I was the UX writer for ID Verify, and foresaw that it would be accompanied by the introduction of many references to IDs, so I advised we get ahead of this duplicative terminology.

So I sent out this email, got almost all the approvals needed and.......got laid off, along with 18,000 other Amazon employees.

However, we can still poke around the live website and see if I had any lasting impact.

After being notified of my position's dissolution, I applied and was hired onto another team within Amazon.

I reached back out to one of my teammates who made it through the purge, and asked if he could share some stats. 

The defined content strategy was a clear success. By focusing on simple, consistent, human language, we saw a 64.2% drop in abandonment from pre-enrollment to enrollment completion - users now understood they weren’t done until they went to a store and scanned their palms.

Internally, the new lexicon and guidelines dramatically reduced cross-team friction, and the number of QA tickets for simple copy changes dropped by more than half, and updating terminology in help articles and cut support contacts by 13.6% as more users could self-solve. The content workshops were also a huge win for getting everyone aligned on how language impacts the user experience, so content design was included in the beginning of feature development

If I were to change anything, it would be pushing for more testing. We made a strong, successful pivot from technical language to 'human' language. If I had more time, I would have loved to A/B test this. Does a 'friendly' tone ever undermine a 'secure' feeling? I'd want to test our 'human' copy against a 'formal and secure' variant to find that perfect balance and back it with data

Sign-up → enroll seems to have mixed adoption

As for Amazon One ID → Profile,

I could only find one live use of "Amazon One ID" buried deep within a help article.

Other than that, every instance of Amazon One ID seems to have been updated to Amazon One Profile.

In addition to the launch of ID Verify, AO also partnered with AXS Mobile to allow users to use their palms as tickets. AXS also used "ID" so adopting "Amazon One Profile" was good future-proofing.

Linking an AXS Mobile ID to an Amazon One ID would be a potentially confusing flow.

All in all, even though my time on this team was cut short, I take pride in knowing I made a lasting impact in a few linguistic nuances that will hopefully make the user experience (and the world) a better place.

Okay, maybe the world is a stretch.