The Feature We Loved. Users Didn't.
We spent days building a recommendation feature, but users cared more about basic things like bookmarks and navigation. After fixing the friction they actually experienced, the product improved and I learned to separate exciting features from useful ones.
Where it started
We were building a small student platform and got very excited about a recommendation feature. It felt like one of those things a “proper” platform should have. Personalisation. Categories. Recommendations based on interests. Very startup-y.
Where I got stuck
We spent days talking about it. Where should recommendations appear? Should users choose interests? Should newer content rank higher? Then we finally showed the product to students. Almost nobody cared. One person asked: “Can I save this for later?” No. Another person asked the same thing. Still no. We had spent days building recommendations and somehow hadn't built bookmarks. During the same session, someone couldn't find the back button. Another person kept clicking something that wasn't clickable. One person opened the homepage and just sort of… stared at it. That hurt a little more than criticism would've.
What changed
We stopped building new stuff for a bit. Instead, we watched people use the site. No explaining unless they were completely stuck. Every hesitation got written down. Every wrong click. Every “where do I go now?” The list was incredibly boring. Add bookmarks. Make this button clearer. Fix navigation. Move this section. So that's what we worked on.
After that
The product became less exciting to describe. It also became much easier to use. And people stopped needing us standing beside them explaining what everything did. That felt more useful than the recommendation system ever had.
What stayed with me
Builders get excited about features. Users get frustrated by friction. Those are not always the same conversation.
