Building an identity-based habit and goal tracker: why votes beat streaks
The Problem
Every habit tracker I tried had the same core model: do the thing, get a streak, protect the streak. And every time the streak broke, the motivation evaporated.
I dug into the behavioral science and found the issue. Streaks run on loss aversion. The fear of losing your streak becomes the primary motivator. But loss aversion is psychologically exhausting. It turns a positive behavior (building habits) into a defensive one (protecting a number). And when the number resets, people quit.
The research on identity-based behavior change pointed to a better model: treat every habit completion as a vote for the person you're becoming. Votes accumulate permanently. They don't reset. Missing a day means one fewer vote, not starting over.
What I Built
Atomic Wins is a habit and goal tracker built around identity voting instead of streaks.
Core mechanics:
- Identity Voting: Link habits to identity dimensions (Runner, Reader, Writer, etc.). Each completion casts a vote. Votes accumulate through tiers: Bronze (25), Silver (50), Gold (100), Platinum (200).
- 2-Minute Fallbacks: Bad day? Just show up. Every habit has a minimal version. Can't do a full workout? Three push-ups still cast a vote. This kills all-or-nothing thinking.
- Compassionate Streaks: Streaks exist for consistency tracking, but missing a day doesn't reset progress or trigger punishment.
- AI Atomization: The AI breaks big goals into specific next actions. "Write a book" becomes "Open a blank document and write a one-sentence summary of chapter 1."
- Dual-List System: Habits (recurring, identity-building) and Goals (one-time achievements) live side by side as Wins. Different mechanics for different purposes.
44 identity dimensions across 11 life domains. Two dashboard modes: a warm card-based layout for most users and a data-dense spreadsheet mode for power users.
The Approach
Solo build in ghost mode. Brand-forward, no personal brand. The product speaks for itself or it doesn't.
The tech stack is straightforward. The interesting part is the behavioral design: making the identity voting feel meaningful instead of gimmicky, making the tier progression satisfying, designing the fallback system so it's effortless to use on bad days.
Where I Am Now
Launching an alpha with a Founding Members program. 250 spots total, released in 3 waves. Wave 1 is 50 spots with invite codes.
Founding Members get lifetime access to all current and future premium features. No expiration, no strings. In exchange: feedback, bug reports, and hopefully some word-of-mouth if the product earns it.
The bet is that 250 genuinely engaged early users are worth more than 2,500 passive signups. Founders who use the product daily and see real identity progression will tell others. That's the growth model.
What I'm Looking For
Honest feedback on the concept, the positioning, or the launch approach. Not looking for validation, looking for the blind spots I can't see because I've been too close to this for too long.
If the identity voting model resonates with you, Founding Member spots are open at atomicwins.ai.
It's really tough to see your own blind spots after working on something for so long. I happen to know a few founders and indie hackers who'd probably be happy to answer your questions about the concept and strategy.