App idea · not built

Underneath, a log for the times you overreact.

Thirty seconds to record what happened: the word for the feeling, what it threatened, how hard it hit. Then one notification a day later asking where it's at now.

Where it came from

An article about emotional inquiry, which is a three-step way of unpacking a strong reaction. Name the feeling as precisely as you can. Work out what it threatened. Then decide what the situation actually needs from you. It's a good method. The problem is that nobody runs it in their head in the moment they'd need it.

My first instinct was a journal with three prompts, and I think that's wrong. Any single entry is close to useless. You already know you were annoyed on Tuesday. It gets interesting around thirty entries, when it turns out most of your worst reactions this quarter were about competence and most of them happened in the same meeting.

So it's a log, and that means the fields have to be structured. You can't count free text. Make the whole thing a text box and you end up with a diary nobody can summarize, including whoever wrote it.

Thirty seconds or it doesn't get used

Four taps and one optional line. Around forty emotion words to pick from, with a note to pick the closest one rather than the strongest one. Six things that might feel threatened. A slider for how bad it was. A line of context if you feel like it.

Then it makes you close it out: what does this need? Nothing, wait, find out more, or say something. Nothing goes first on the list, because most of the time that's the honest answer and an app that never offers it is teaching you the wrong thing.

The four screens

OneName it

Closest word, not the biggest one

Exposed, dismissed, resentful, deflated, embarrassed, anxious, and about thirty-five more. "Upset" isn't a feeling, it's a category, and it doesn't tell you anything later.

TwoUnderneath

What got threatened

Competence, fairness, belonging, certainty, autonomy, respect. Six, fixed. If people can add their own the monthly summary stops meaning anything, and a short list is what makes the question answerable in four seconds.

ThreeClose it

What does this need?

Nothing, wait, find out more, say something. One tap and you're done. Without this it turns into a list of grievances with nice typography.

The part I actually like

24 hours

One notification, a day after you log something, asking where the intensity sits now.

Everything else in this app is a tidier version of something that already exists. This bit isn't. Being told that most things fade doesn't help. Watching your own fours turn into twos, thirty times over, might change how the next four feels while you're still in it.

It's also the only thing keeping anyone coming back, since I cut everything else that would. So it has to be good, and it has to be one notification and not two.

Things it won't do

Left out

Streaks, badges, levels
An AI companion to talk it through with
A feed, or anything shareable
Open text as the main field
More than one notification per entry
A list of needs you can add to

A streak here would punish you for having a good week and get you making things up to keep it alive, which wrecks the only data the app has. The chat companion is worse. The method works because it's three questions and then you're done. Open it up into a conversation and you're back to going round in circles about why you're annoyed, which is the thing it was supposed to replace.

What it would take

Platform

iOS. Four screens, a local database, and a notification schedule. No account.

Data

Stays on the phone. This is a list of someone's worst moments of the month and it shouldn't be sitting on a server I'm running.

RiskTiming

The moment you'd most want this is the moment you're least likely to open it, and the phone is part of the problem right then. Realistically most entries get made an hour or a day later. That has to be the normal case in the design, not the exception.

RiskRumination

Going back over bad moments in a structured way helps some people and makes it worse for others. The forced close helps, so does putting "nothing" first, so does a soft limit on entries per day. I'd want to talk to someone who actually works in this area before shipping it, rather than sticking a disclaimer in the App Store listing and calling it handled.

RiskRetention

Mood trackers lose most of their users within a couple of weeks, and the ones that don't are using exactly the tricks I've taken out.

Money

Not great as a consumer app. No network effect, nothing to defend, and three screens somebody could copy in a weekend. Fine if I'm building it to have built it. If I wanted it to make money the better version is selling into leadership development programs, where the buyer is a training budget and the thing they're buying is the pattern report.

Not sure yet

I called it Signal to start with, which was a bad idea given there's already a messaging app by that name on most people's phones. Underneath was the label on the second screen and it works better. It says what the app asks you to do without promising you any insight.