Identify Amazon brand recognition signals
I hypothesized that shopper trust and Amazon brand recognition were tied to specific signals. To identify them, I ran unmoderated studies testing the styling changes merchants requested most: fonts, background color, and star rating colors. We learned that much of what merchants wanted was safe within the right constraints, like limiting fonts to four research-backed presets or allowing transparent backgrounds with accessibility guardrails. And we found the key recognition signals, like the classic Amazon orange stars, had to remain consistent across different merchant sites.
Offer flexibility within guardrails
We turned the findings into a simple model: protect the elements that carry trust or Amazon brand recognition, and offer flexible controls for the rest, like buttons, links, and expanders, to match the merchant's site. Once I defined what shoppers see on merchant sites, the merchant UX designer shaped how merchants configure those controls in their editor. As a result, reviews looked integrated into the merchant's site without losing Amazon's trust signals.
Build a system, not one-off decisions
As we took an iterative approach to find how much customization was too much, I designed a framework to help the team assess each new request against brand recognition and shopper trust at scale. Partnering with a technical writer, PM, and legal, I published the first guidelines for how Amazon reviews appear outside Amazon, now referenced by merchants and partner teams.
Default vs. Customized
