The short answer
StickK is the original. Founded in 2007 by Yale economists Dean Karlan and Ian Ayres with Jordan Goldberg, it turned a piece of behavioural economics — that people will pay to constrain their future selves — into a website. You write a Commitment Contract, put money behind it, name a referee, and invite supporters. It is free to use.
Commit takes the same premise and moves the verification into the phone. Instead of asking a friend to confirm a report, it asks for evidence the moment the thing happens: a photo taken in-app, a GPS-verified run, a screen-time reading, health data. Commit AI reviews it, and a friend can still sign off if you want a human in the loop.
StickK's weak point isn't the theory — the theory is excellent. It's that the whole contract rests on someone else remembering to click a button.
Side by side
| Commit | StickK | |
|---|---|---|
| Founded | 2026 (public beta) | 2007, by Dean Karlan, Ian Ayres and Jordan Goldberg |
| Verification | Photo, video, GPS, screenshot, file, time-lapse, health data or screen time — reviewed by Commit AI, with optional friend verifier and an appeal | A human referee you nominate confirms your self-report. No automated or photo verification |
| If nobody verifies | Proof is reviewed automatically; it doesn't wait on a person | Progress depends on the referee acting |
| Where money goes | Charity, anti-charity, or Commit — as a service fee | A friend, a charity, or an anti-charity |
| Social layer | Squads of up to 10, shared commitments, a group streak that only advances when everyone shows up | Supporters who can follow and cheer |
| Stake amount | $1–$500 per missed check-in (USD, CAD, GBP or EUR) | You set the amount per contract |
| Cost to use | Free during beta | No subscription fee |
| Platform | iOS (public beta) | Web, with mobile apps |
| Research pedigree | Built on the same behavioural-economics literature | Wrote a good deal of it — see Ayres' Carrots and Sticks |
StickK facts verified 12 August 2026 from stickk.com/aboutus and public records of its founding.
The referee problem
StickK's design is elegant on paper. You report progress; a referee you chose confirms or denies it; money moves accordingly. A human can exercise judgement that no algorithm can — they know that you were genuinely ill on Tuesday.
In practice, the referee is the single point of failure. They are doing you an unpaid favour, indefinitely, on a schedule set by your goals rather than theirs. Enthusiasm fades in week three. And there's an awkwardness the design never resolves: you've asked a friend to be the person who takes your money away. Some people quietly stop enforcing rather than have that conversation.
Commit's answer is to make verification the default path rather than a favour. Proof is captured in-app at the moment of the act — not uploaded later from your camera roll — and Commit AI checks it against what you committed to. If it makes the wrong call you can appeal and a human reviews it. A friend can still be your verifier, but nothing waits on them.
This trades away something real. An AI reviewer doesn't know your life. That's why grace tokens exist, and why appeals do.
Anti-charities: StickK's best idea
The anti-charity is StickK's sharpest contribution. Instead of your forfeit going somewhere pleasant, it goes to an organisation whose views you actively oppose — which converts a mild financial sting into something you'll genuinely reorganise your morning to avoid. StickK's own materials report that users who choose an anti-charity are meaningfully more likely to hit their goal than those who pick a charity or a friend.
We think they're right, which is why Commit offers anti-charity as a stake recipient too. Credit where it's due — this was their insight, and it's a good one.
One honest difference: on Commit, a stake is a service fee paid to Cognitive Health Research LLC, Commit's parent company — not a donation. You won't receive a tax receipt, and costs are deducted before disbursement through our Charitable Giving Program. We'd rather say that plainly than let you assume otherwise.
When StickK is the better tool
- You want it today, on any device. StickK runs in a browser and has for eighteen years. Commit is iOS-only and still in public beta.
- Your goal needs human judgement. "Be kinder to my sister", "practise piano properly rather than mechanically" — no camera adjudicates that. A referee can.
- You already have a reliable referee. A partner, a coach, a sibling who enjoys this. If that person exists, StickK's model works as designed.
- You want the original. There's a real argument for the tool built by the people who ran the studies.
- You want to stake toward a specific person. StickK lets a friend receive the forfeit. Commit removed that option; ours go to a charity, an anti-charity, or Commit.
When Commit is the better tool
- You don't have a referee, or you've burned one out. This is the most common reason people leave StickK.
- Your goal is provable with a camera or a sensor. Gym sessions, runs, screen time, cooking, tidying, reading.
- You want a daily habit, not a single deadline. Commit is built around recurring commitments with windows and schedules.
- You want a group, not spectators. A squad of up to ten shares commitments and a streak that only advances when every member shows up — rather than following along from the sidelines.
- You want it to feel like a phone app from this decade. Bluntly, this is the most common criticism of StickK, and it's fair.
Commit is in public beta
iOS, free while in beta. There's no App Store listing yet — we'll email you when your spot opens.
Join the betaRead the Commitment Agreement first if you want to see exactly what you'd be signing.
A note on fairness
StickK made the case that commitment contracts work, with actual research, years before anyone was building habit apps. If anything here is inaccurate, write to [email protected] and we'll fix it.