The Day the App Said 'No' — And How I Learned to Listen to the Users
When building the AI-powered planning assistant, Planner, I made a wrong assumption about what users needed. The app didn't respond as expected, and I had to go back to the drawing board — and the users — to find the real problem.
Building Planner was meant to be a tool that helped people think through their goals and figure out the next steps. I assumed that users would appreciate a highly structured interface with clear dependencies and a logical flow. But when I launched the first version, the app didn't feel right to the people who tried it. It was functional, but it didn't resonate.
I had spent a lot of time designing the AI logic to surface dependencies and suggest sequencing. I had tested it with a few people, and it worked. But when I shared it with a broader group of users — professionals, parents, and small business owners — I got a different kind of feedback. They said it felt too rigid. It didn't help them think, it made them feel like they had to follow a script.
That was a wake-up call. I realized I had missed something important. The AI had done its job, but the experience wasn't right. The mistake wasn't in the code — it was in the assumption. I had assumed that people wanted a tool that told them what to do next, but what they really wanted was something that helped them reason through the choices themselves.
So I went back to the drawing board. I talked to more users. I asked them what they needed. I listened to their frustrations. And slowly, I started to see what was missing. The tool needed to be more flexible. It needed to be a thinking partner, not a script reader.
I adjusted the interface. I made it easier to move around ideas. I gave users more control over how they structured their plans. And I made sure the AI didn't push too hard. The result was a product that felt more natural, more intuitive — and more useful.
That experience taught me something important. Building AI-powered tools isn't just about the technology. It's about understanding the people who will use them. And sometimes, that means going back, admitting you were wrong, and listening again.
If you're building something and it doesn't feel right, don't be afraid to step back. The mistake might not be in the code — it might be in the assumptions you made along the way. And the answer might not be in the AI — it might be in the people you're trying to help.
