What We Learned From the First 100 Beta Users

Sixty-one of our first 100 beta users opened the app fewer than five times in the first month, and most of them are still using it.

That number confused us for weeks. We built Echos of Mind around the idea that noticing a pattern takes repetition — you log a moment, then another, then a third one that looks like the first two, and something clicks. Low open counts looked like failure. Then we read the actual entries people were leaving, and the picture changed. People weren't opening the app to check in. They were opening it when something already felt off, writing down what happened, and closing it. The app wasn't a habit. It was a specific tool for a specific moment, used less often than we expected and, apparently, more usefully.

What we thought mattered vs. what people used

We spent most of our pre-launch time on the pattern timeline — the view that strings related entries together and shows you the shape of a recurring behavior over weeks. We assumed that was the core of the product, the thing people would open daily and study.

In the beta, the timeline got used by about a third of people, and mostly after they’d already noticed something on their own. What people actually leaned on was the single-entry view: write down what happened, what you were doing right before it, how it felt in the body. No trend line, no aggregation, just the raw moment. Several people told us, unprompted, that seeing the pattern laid out for them felt like being told the answer before they’d worked it out themselves. One user put it more bluntly: "I don’t want you to tell me the pattern. I want to see it myself, later, and be annoyed that I didn’t notice sooner."

We hadn’t designed for that. We’d designed for insight delivery. What people wanted was a place to be honest in the moment, with the pattern-recognition part left for them to do on their own time.

The feature nobody asked for that people kept returning to

Buried in the entry screen was a small field we almost cut before launch: a spot to note where you were and who you were with, separate from the emotional description. We added it late, mostly because a friend testing an early build said something like "I keep forgetting the context matters as much as the feeling."

It became the most consistently filled-in field in the entire app — more than the mood-adjacent language, more than the trigger notes. People used it to catch situational patterns they hadn’t clocked: the same coworker, the same time of day, the same kind of Sunday. We’d built it as an afterthought. It ended up doing more of the actual behavioral-pattern work than the features we spent months designing.

What "self-awareness" meant to them, versus what we assumed

Going in, we imagined self-awareness as something like insight — a moment of "oh, that’s why I do that." What we saw in the beta was closer to permission. People used the word "relief" more than we expected, and almost never in relation to a discovery. It showed up when they described just writing something down without a score attached, without a streak to keep, without anyone — including the app — telling them whether the entry was good or concerning.

A few people said flatly that other apps they’d tried made them feel judged by their own data. That wasn’t a compliment about our design. It was a low bar other products had set, and we happened to clear it by not doing the things that made people feel judged in the first place.

What we changed, and what we're still not sure about

We killed the daily reminder notification two weeks into the beta. It was driving opens, technically, but the entries that came from notification-triggered sessions were shorter and vaguer than entries people wrote unprompted. We were optimizing for engagement and getting worse data in return, which is its own kind of warning sign. We also moved the context field — the one nobody asked for — higher up in the entry flow, because burying a feature people actually want is just bad design dressed up as minimalism.

What we haven’t figured out is what to do with the timeline. It’s still there. Some people do want it, later, after they’ve already noticed something themselves — a way to confirm what they suspected rather than be told it. We don’t know yet whether to keep building it out or quietly let it become a secondary feature nobody misses if it’s gone. We’re leaving it as-is for now, mostly because we don’t have enough beta users who’ve used the product long enough for the timeline to mean anything.

We built a tool to show people their patterns and learned that the tool people wanted was a place to describe a moment honestly and leave the pattern-spotting to themselves. The gap between those two things — what you build and what people actually reach for — is more useful than any launch metric we’ve tracked so far.