The problem I kept running into
Every time I needed to do competitive research, the workflow was the same painful loop. Open 10 tabs, take notes in a doc, screenshot things, try to remember what you saw on that one checkout flow two sites ago. Then spend an hour reformatting everything into something you could actually show someone.
The insights were always there. The process was the problem.
What I built
Scout is a Chrome extension that does competitive UX analysis as you browse. You open a competitor's site, Gemini analyzes the UX patterns on the page, and annotations surface directly on top of what you're looking at. When you're done browsing, you get an exportable report.
The core loop is: browse, annotate, export. That's it.
The technical bit
The interesting engineering challenge was making the Gemini integration feel invisible. The goal was for insights to appear at the right moment without interrupting the browsing experience. A lot of the work was in figuring out when to trigger analysis, how to render annotations without breaking page layouts, and keeping latency low enough that it doesn't feel like you're waiting for something.
Early versions tried to do way too much. The turning point was stripping everything back to that core loop and making each step feel fast and obvious.
What I'd do differently
Scope creep is the real enemy of side projects. I spent weeks on features that didn't survive to launch. If I started over I'd define the one thing the tool had to do well and ship that first.
Where it is now
It's an extension on the Chrome Store.
Would love feedback from other builders, especially anyone who has worked with Chrome extension architecture or the Gemini API. Always curious how others would approach the same problem.
Link: https://chromewebstore.google.com/detail/ecmkeokcmiflgkfnnhbcmcklmdobkila?utm_source=item-share-cb
Top comments (0)