What It Feels Like to Build Something People Use at Their Worst

Most apps get opened casually. Someone's bored, curious, killing time between meetings. Unheavy gets opened when something broke — a fight that didn't resolve, a thought that's been running for three days, something that happened that you haven't been able to put down.

That asymmetry changes what you build. It changes what questions you ask. It changes what you choose not to build. And it sits with you differently than a normal product, because you know what people are carrying when they open the thing you made.

The Weight of Knowing

I don't know what people write in Unheavy — that's the point, and I'll get to the architecture in a moment. But I know the category. People aren't journaling about their day. They're writing the thought they've never said aloud. The accusation they'll never send. The grief they've been carrying without naming it. The thing that surfaces at 3am when there's nothing left to distract them.

When you build something and you know that's how it gets used, the usual product questions feel beside the point. Session length, DAU, retention curves — those metrics assume a kind of casual, replaceable engagement. This isn't that.

I'm not saying Unheavy is important because I built it. I'm saying the category is serious, and building in a serious category changes your priorities whether you want it to or not.

The Question That Shapes Every Decision

The question I kept coming back to during development: what would I want to be true if I was the user right now? Not "what maximizes session length" or "what drives return visits." Just: what does this person actually need, and is this product getting out of their way?

That question sounds simple. In practice, it cuts against most of what product development culture rewards. Engagement mechanics, habit loops, notification strategies — these exist because they work. They're good at extracting attention. They're bad at serving someone who opened an app in the middle of a hard night.

Why Nothing Is Stored

The obvious reason to not store data is privacy. But there's a more specific reason that only becomes clear when you think about what people actually write.

If you write something you've never admitted — the real version, not the version you'd tell a therapist — having it saved somewhere changes the act of writing it. There's a small part of you that knows it exists. That it could be found. That it lives on a server, subject to data breaches, policy changes, legal requests.

No storage isn't just a privacy position. It's what makes the writing fully honest. I've heard from people who say they wrote things in Unheavy they couldn't write in their private journal. I think that's the architecture doing its job — removing the witness, even the imagined future one.

What I Chose Not to Build

No accounts. No social layer. No mood tracking. No trend graphs. No badges. No streaks. Each of these is a reasonable feature somewhere. None of them belonged here.

Accounts mean someone can log in from another device and find what you wrote. Mood tracking means numbers, and numbers mean you're performing something measurable. Emotional processing isn't a performance. Streaks reward consistency in something that should be irregular — you don't want to be using an emotional release app every single day. If you are, something else is going on.

Social features are the most obviously wrong choice. "Share this to your story" after writing something you've been carrying for months. That's not a feature. That's a betrayal of why someone opened the app.

Why One-Time Purchase

A subscription creates an ongoing commercial relationship with a product people should be able to forget about between sessions. The last thing someone needs when they're in the middle of processing something real is friction about payment, upgrade prompts, or the feeling that they're paying monthly for something they haven't used.

Mental health tools shouldn't have retention mechanics. They should work, then disappear until you need them again. One-time purchase means the product has one job: be good enough that people remember it and recommend it. There's no extraction mechanic. The business model aligns with the user's interest.

I'm not going to claim this was a pure ethical position — it was also a practical one. But the alignment matters. When you're building something for vulnerable moments, misaligned incentives cause real harm.

The Harder Part

Some percentage of people who download Unheavy are in serious pain — not "had a hard day" pain, but the kind that makes you consider things you shouldn't. I can't know who or when. I can't read what's written. The architecture prevents it.

What I can do is make sure the product doesn't make things worse. Simple, calm, non-alarming. Clear that it's not a substitute for professional help. Not overclaiming what the app does — "a place to put it down" rather than "heal what's hurting." There's a difference between a tool that helps someone manage a hard moment and a tool that implies it's enough on its own.

If you're in crisis, please reach out to a therapist, counselor, or crisis line. In the US, you can call or text 988.

If You're Building in This Space

The responsibility isn't abstract. Ask before you ship: could this data be subpoenaed? Could a vulnerable person misuse this mechanic? Am I introducing friction that makes someone feel worse for not using it?

The bigger question: does this make the user stronger, or does it make them dependent on the product? Most apps I've seen don't ask that. The ones that do tend to look different — simpler, quieter, less sticky. That's not a failure of product thinking. That's what the category requires.

Build like you know someone's going to open it at 2am with something they can't say to anyone. Because they will.