We spent three months building the dashboard.
It had charts, filters, export options. You could slice the data by time period, by user segment, by geography. It was comprehensive. It was polished. The design team had iterated on it for weeks.
Two months after launch, the analytics showed that almost no one used it.
A handful of users had clicked on it once. Fewer had clicked on it twice. The export feature had been used exactly four times, and two of those were by people on our team testing it.
This was not a failure of execution. The feature worked exactly as designed. The problem was that nobody had wanted it in the first place.
The request had come from a sales call. A potential customer had asked if we had "advanced analytics." We didn't. The sales team flagged it as a blocker. Product added it to the roadmap. Engineering built it.
Nobody went back to ask what "advanced analytics" actually meant to that customer. Nobody checked if other customers wanted the same thing. Nobody built a small version first to see if anyone cared.
We just built it, because it was on the roadmap, and things on the roadmap get built.
The feature still exists. It's still in the product, still maintained, still occasionally causing bugs that someone has to fix. It has a carrying cost that will continue indefinitely, for something that provides almost no value.
This happens more than anyone admits. Most products are full of features that nobody uses. They accumulate over time, each one added for a reason that seemed good at the moment, each one now quietly consuming resources.
The problem isn't that we build things that fail. Failure is fine. The problem is that we build things without ever checking if they're needed, and then we keep them forever because removing them feels like admitting a mistake.
The dashboard taught me something I should have learned earlier: the most expensive features aren't the ones that are hard to build. They're the ones that are easy to build but shouldn't exist.
Building is cheap now. Judgment is expensive. And the hardest judgment is deciding not to build something that someone, somewhere, once asked for.