See the app

Identity-based habits: how they work, and how to build one

Most habit advice starts with what you'll do. Identity-based habits start with who you've decided to be — and that single reordering changes what you track, what counts as progress, and what a missed day means.

If you've ever kept a habit tracker for three weeks and then quietly stopped opening it, the problem probably wasn't discipline. It was that the tracker was measuring the wrong thing. A checklist measures compliance. What actually keeps someone going is the sense that they are turning into a different person — and compliance is a poor proxy for that.

An identity-based habit is a small, repeatable action chosen because it is evidence for a person you're trying to become. The action matters less than what it proves. Ten minutes of reading is a task. Ten minutes of reading, done because "I'm someone who thinks for a living", is a vote.

The idea was popularised by James Clear in Atomic Habits, which frames every action as a vote for the type of person you wish to become. It's a good frame. It's also usually where the treatment stops — one chapter, then back to habit stacking and streaks. This page is about what it looks like when you take the frame seriously for a few months.

Why checklist habit tracking runs out

Checklists have a specific failure mode, and it's worth naming precisely because the fix follows from it.

A streak tells you that you showed up. It can't tell you who showed up.

How to build an identity-based habit

Four steps. The first is the one people rush, and it's the one that determines whether any of the rest works.

1. Name the identity in your own words

Not a role ("a better employee"), not a metric ("someone who reads 50 books"), and not an adjective pile ("productive, disciplined, focused"). A sentence a friend would recognise as being about you. A calm presence in every room. Someone who finishes what they start. A present, patient parent. Quietly becoming wealthy.

Test it: if the sentence would look identical on someone else's page, it's too generic to guide a decision. The point of naming it precisely is that on an ordinary Tuesday you can ask "what would that person do right now?" and get a real answer.

2. Name the obstacle honestly

The thing that actually gets in the way — not the socially acceptable version. "I don't have time" is usually a stand-in for something more specific: self-doubt, a phone within arm's reach, resentment about who does more at home. Identity work that skips this step produces advice that's technically correct and useless to you.

3. Make the first action embarrassingly small

Small enough that you'd feel silly negotiating with yourself about it. Two minutes is a good ceiling. The reason isn't that two minutes produces results — it doesn't, directly. It's that two minutes is too small to fail, and a version of you that never fails to act is a very different input to your self-image than one that keeps setting thirty-minute goals and missing them.

A calm presence in every room takes one deep breath before checking their phone. That's the whole action. It works because it's true to the identity and because there is no plausible day on which it's impossible.

4. Collect evidence, not streaks

Change what you record. Instead of did I do it, record what did today prove. Over a month this produces something a checklist never does: a body of evidence you can read back. The most convincing thing you can show someone who doubts they've changed is a record of the days they acted while rating their own energy at its lowest.

What to do when you miss a day

Miss one. It's fine — and the design of your system should say so out loud, before it happens, or you'll invent the permission after the fact and feel worse about it.

The useful rule is that a single miss is data, and two consecutive misses are a signal. One missed day usually means life happened. Two in a row usually means the action is too big, the identity isn't actually yours, or the obstacle you named wasn't the real one. Treat the second miss as a prompt to revise the system rather than to try harder at it.

This is why forgiveness has to be built in rather than granted. A streak that resets to zero punishes exactly the person who is closest to giving up.

How do you measure something as vague as identity?

You can't measure identity directly, but you can measure the things that move with it, and you can do it without pretending it's precise.

Four mistakes that quietly kill it


Doing this without an app

You don't need software. A notebook works: identity at the top of the page, the obstacle underneath it, one line a day about what you did and what it proved. Most people who succeed at this do it on paper. If that's you, close the tab and go write the sentence — that's a genuinely good outcome.

What Arc does with all this

Arc is an Android app built entirely on the four steps above. You name the identity and the obstacle once; each morning it writes a single 2-minute action for that specific person and that specific obstacle, and you can swap it if it doesn't fit. One missed day a week is frozen automatically. At day 30 it writes you a mirror — a plain-language account of what your own month showed — and a six-dimension compass that keeps updating from there.

It's $49.99 a year or $7.99 monthly, with no free tier and no ads, because the subscription is the whole business model. The FAQ covers the details, and the privacy policy is specific about what's stored.

Whichever way you do it, the reordering is the part that matters. Decide who you're becoming first. Then let the day's action be the smallest honest piece of evidence for that person, and let a month of evidence be the thing you judge yourself on.

Thirty days from now, you'll want the evidence.

Arc writes one 2-minute action a day for the person you decided to become — and shows you the distance at day 30.

See how Arc works