ADHD accountability partner, at the scale of a day
Body doubling gets you through the first ten minutes. This is for the fortieth day. Different problem, different fix.
CommitMe is a habit app, not a treatment and not a clinical tool, and it makes no claims about ADHD beyond the obvious one: a lot of people who find habit trackers useless find them useless for the same reason. A tracker that only you can see is asking the part of you that is already struggling to also be the enforcement.
Two different problems that get the same name
"I cannot get started" and "I cannot keep it up" both get called an accountability problem and they need different tools.
Starting is a problem of the next ten minutes. Somebody present, right now, is what moves it. Body doubling works for exactly that reason, and Focusmate built a whole product on it.
Keeping it up is a problem of the fortieth day. Nobody can sit with you for that, and booking a call every time is its own barrier when the task takes fifteen minutes and you need to do it four times a week.
Where body doubling stops
Focusmate and the rest solve task initiation, and they solve it well. You book a session, somebody is there, you start. What they do not do is carry across days. Nothing persists between sessions, so week four looks exactly like week one.
CommitMe works a day at a time. People you invited can see whether a shared habit is logged. A nudge is one tap for them. Nobody has to be available at the same time as you.
The case for body doubling rests on reports rather than trials. Medical News Today says no controlled experiment has tested it. ADDA published the article that named the practice in 1996 and says the same in it. Enough people describe the same effect that arguing with them would be strange, so the honest form of the claim is that people report it works.
The two are not competing. Book a session when starting is the problem, and log the habit here when the session ends. One habit, two tools, and neither needs to know about the other. If you are choosing between them rather than stacking them, the question is whether your bad days look like a task you cannot open or a week you quietly stopped doing.
Why the standard advice fails here specifically
Most habit advice quietly assumes that a future consequence feels real today. Build the habit now, benefit in six months. For a lot of people that assumption holds well enough to carry them through a dull Tuesday. For a lot of people it does not, and no amount of restating the benefit makes a distant reward feel present.
Two things follow, and both point the same way. A consequence has to arrive close to the behaviour rather than at the end of the quarter. And it works better coming from outside your own head than from a promise you made to yourself, because the promise is subject to renegotiation by the same person who made it.
A crew supplies both cheaply. The gap shows up today rather than in March, and it shows up to somebody else.
There is a second reason to be wary of advice built on trying harder. Most of it assumes willpower behaves like a fuel tank that empties as you spend it. That model failed a pre-registered replication run across 23 labs with 2,141 people in 2016, and researchers have been split about what survives ever since. A plan that depends on rationing willpower is built on a foundation the field is still arguing about. Structure outside your head does not need that argument settled.
Interesting is not a plan
The first fortnight of anything new tends to go well, and it is easy to read that as proof the system works. It is proof the system is new. When the novelty goes, whatever is left has to be structural: the slot, the size of the action, the reminder, the person who notices.
This is also why switching apps rarely helps. A new tracker restores the novelty for about ten days and then leaves you exactly where you were, having spent the intervening time setting it up.
Why trackers get abandoned
Measuring something has a cost, and it is easy to miss because the measurement is also working. Etkin ran six experiments on this and published them in the Journal of Consumer Research in 2016. In one, 95 students spent a day walking. The ones who could see their step count walked more than the ones whose pedometer was sealed shut, and enjoyed the walking less. Getting to choose whether to be measured did not protect them. Attention on the output made the activity feel like work.
That is a live problem for anything built around a grid of squares. The answer is not to stop tracking. It is to track less of your life and be deliberate about which parts.
- Track the ones you drop, not the ones you like. If you already run three times a week and enjoy it, putting it on the board buys you nothing and risks turning it into a chore. The reading you keep meaning to do is the candidate.
- Share fewer than you track. Sharing is per habit and opt in, so a habit is visible to your crew only if you turn it on. The default is private, and most people should leave most habits there.
- Three or four, not twelve. A long list is the fastest route to a day where you cannot face opening the app at all, and one skipped day of logging tends to take the rest of the week with it.
- Let a habit leave. Once something happens without you thinking about it, delete it.
Missing a week does not reset you to zero
This usually decides whether an app survives contact with a real month. You earn one pause day for every four completions, banked up to thirty. Opening an absence spends them and the streak carries through. The record survives the weeks that were not your fault, and that is the difference between a tool you come back to and one you delete in shame.
The fortieth day is not an arbitrary number to design around. Lally and colleagues followed 96 people picking up a new daily behaviour and reported in 2010 that it took a median of 66 days for the behaviour to reach its plateau, with individual times ranging from 18 to 254. Whatever your own number turns out to be, you will still be somewhere in the middle of it when the first bad week arrives.
The same study found that missing a single day did not measurably change the curve. An app that treats one miss as a reset is punishing you for something the evidence says you recover from, and that is what gets the app deleted rather than the habit.
Smaller than you think it should be
The daily action wants to be small enough that it does not need a decision. Not thirty minutes of admin, five minutes. Not the gym, the walk. The size that feels too small to count is usually the size that survives a bad week, and a habit that survives bad weeks beats an ambitious one that runs for nine days.
You can also anchor a new habit to one you already do, so the trigger is something that is already happening rather than a moment you have to remember.
Deciding in advance when and where beats deciding to try harder. Gollwitzer and Sheeran reviewed 94 studies covering more than 8,000 people in 2006 and found that plans of the form "when X happens, I will do Y" had a medium to large effect on follow-through, d=0.65. A later and much larger review across 642 tests put the range lower, from d=.27 to d=.66. Both are worth having for something that costs one sentence to write.
A habit here has a time and a reminder at that time, or an anchor to a habit you already keep. Either one is the same sentence, already written down and no longer your job to remember. What it will not do is choose the time for you, and a time you picked because it sounded responsible rather than because you are reliably free then is the usual reason a habit dies in week three.
Who to put in the crew
Five people by default, private to whoever you invite. Who you pick matters more than how many, and the wrong pick is worse than an empty crew.
- Someone who taps and moves on. A nudge is one tap. You want a person who sends it and then gets back to their day, not one who treats it as an opening for a conversation about whether you are serious this time.
- Someone with something of their own on the board. If you are the only person being watched, it stops being a crew and becomes supervision, and supervision is a thing people escape rather than keep.
- Not the person who already brings it up. If somebody in your life already comments on this particular habit, inviting them turns a recurring argument into a live feed. Pick the friend who has no opinion about your sleep.
- Three who look beats five who forget. Five is the default, not a target. An invite link sent to somebody being polite produces a name in a list and nothing else.
Then tell them how to read a blank square, because they will get it wrong on their own. A kind person sees a gap, assumes you are having a bad day, and says nothing, so the one day you needed a nudge is the day nobody sends one. Say it out loud: a blank square is not a crisis and not a confession, and tapping is pre-authorised on every one of them. Given that, they never have to judge whether today is the day to say something, and you never have to explain the gap afterwards. Shame usually enters through that explanation.
What it does not do
There is no automatic pairing. You invite your own people, five by default, or you join a community of people working on the same kind of thing. Nothing puts a stranger in your crew for you. There is also no live video and no session booking, so if what you need is somebody on a call right now, this is the wrong tool and Focusmate is the right one.
There is no money at stake and nothing to lose but the square. Some people need a deposit on the line to move, and if that is you, the apps that hold your money are a different product and you should use one. There is no native app either. It runs in a browser and installs to your home screen.
It is free, there is no paid tier, and the free tier is one crew. That is the whole product today.