How to Tell if Your Product Is Actually Useful or Just Visually Polished

There is a dangerous gap between a product that looks good and a product that is genuinely useful. A polished UI with smooth animations can feel impressive in a demo but collect dust after the first week. The metrics most teams track - DAU/MAU ratios, session length, engagement scores - do not reliably distinguish between the two.

Session length is especially misleading. A long session could mean users are deeply engaged, or it could mean your interface is confusing and people cannot find what they need. High DAU can mean real utility, or it can mean you have built a habit loop that keeps people opening the app without getting real value from it.

The One Signal That Matters

Unprompted return visits. Someone comes back to your product on their own, without a push notification, without an email reminder, without a streak mechanic guilt-tripping them. They come back because they hit a problem and thought of your tool as the solution.

This is hard to measure directly, but you can approximate it. Look at return visits that happen outside of any notification or marketing touch. Filter out sessions that start from a push notification click or email link. What is left is organic pull - people who actively chose to open your product.

How to Test Early

Before you have enough data for statistical significance, just watch people. Do user interviews a week after onboarding and ask one question: "Have you used it since we last talked?" If they have not, no amount of polish will fix the underlying problem.

The best products we have seen are often ugly early on. They work, they solve a real problem, and people keep coming back despite the rough edges. Visual polish is the last 20% - it makes a useful product delightful, but it cannot make a useless product useful.

Build something people reach for when they have a problem. Everything else is decoration.

Fazm is an open source macOS AI agent. Open source on GitHub.