Atomic Habits in Practice: Identity Over Outcomes

A great idea, an underused one.

One of the more useful ideas to come out of recent habit-formation writing is simple: every action you take is a small vote for the type of person you want to become. It reframes habits from tasks you complete into evidence you collect. Most people read it, nod, and then go back to a to-do list that has nothing to do with identity.

Understanding the idea is the easy part. The hard part is translating it into something you actually do at 7am when you're deciding whether to get up or hit snooze one more time.

Outcomes are noisy. Identity is steady.

Outcome-based goals (lose ten pounds, save a certain amount, hit a revenue number) are noisy in the short term. You can do everything right for a week and see no movement, which is discouraging enough to make people quit before the outcome shows up.

Identity doesn't have that problem. If your identity is "someone who trains," then training today is a direct win regardless of what the scale says. The outcome eventually follows from a large enough stack of these direct wins, but the wins themselves don't require waiting for it.

Turning votes into a system.

Naming your identities explicitly instead of leaving them implied is where this starts to work. "I want to be healthier" is a wish. "I am someone who trains four times a week" is an identity you can vote for daily. The more specific the identity, the easier it is to know what today's vote should look like.

This is the mechanic behind Ascend: you define identities, attach small actions to each one, and complete them daily. Every checkmark is a vote, tracked and grouped by identity instead of floating in a generic list, so the pattern of your votes is visible at a glance.

Let the weekly tally do the convincing.

Willpower is a bad long-term strategy, but a running tally of votes is a good one. When you can see, at the end of a week, exactly how many times you voted for a given identity, the case for continuing tends to make itself. You're not arguing with motivation. You're looking at a record.

That's also where course correction happens. If one identity has a lot fewer votes than the others, that's worth noticing before another month goes by the same way.