Body doubling is the only ADHD trick that's never failed me

I can lose an entire afternoon to a task that takes eleven minutes. Not because it's difficult — because starting it means being alone with it. The task sits there, I know exactly what it is, and some part of my brain refuses to make first contact. I'll reorganise a folder. I'll read the documentation for a library I'm not using. I'll get a third coffee.

Then someone sits down at the other end of the table with their own laptop, and I just… do it. Eleven minutes. No willpower involved. Nothing about the task changed. The only new variable in the room was a person who wasn't even looking at me.

That's body doubling. It's the most reliable executive-function hack I've found in a decade of having this brain, and it has one enormous flaw: it needs a second human.

What body doubling actually is

You work on your task. Somebody else is present, doing their own unrelated thing. That's the whole technique.

They don't help. They don't check your progress. They don't know what your deadline is. There's no pact, no report at the end, no "did you do it?" text the next morning. They are just a warm body in the vicinity while you do the thing you couldn't start.

Worth separating from its noisier cousin: an accountability partner is about consequences after the fact — you tell someone whether you did it. Body doubling is about the twenty seconds before you begin, and involves no reporting whatsoever. That's exactly why it doesn't add its own layer of dread. Accountability gives me a second thing to be avoidant about. A body double gives me nothing to avoid.

Why it works, and why nobody can fully explain it

I want to be straight about the evidence here, because ADHD content has a bad habit of dressing up folk wisdom as neuroscience. Body doubling is widely used and constantly recommended in ADHD communities and by ADHD coaches. It is not a technique with a deep pile of controlled trials behind it. Anyone telling you the mechanism is settled is guessing with more confidence than I am.

The most directly relevant study I've found is a 2025 preprint by Zinat Ara and six co-authors, "You Are Not Alone: Designing Body Doubling for ADHD in Virtual Reality". They put participants through a bricklaying task in VR under three conditions — alone, with a human body double, and with an AI one — and found people finished faster and reported better sustained attention with either kind of double than with none.

Encouraging, and I'd rather you knew its limits than took my word for it: twelve participants, in a headset, laying virtual bricks. That is not the same as answering your email on a Tuesday, and the paper itself notes that earlier work found no conclusive effect and that some neurodivergent people are actively uncomfortable with someone else being present. Twelve people is a hint, not a finding. I mention it because it's the only real data in the neighbourhood and because most pages selling you an AI body double quietly round it up to proof.

What I can offer is the three explanations that match my own experience most closely:

It moves the intention out of your head. Holding "I am going to write this email" in working memory is, for me, genuinely effortful, and the intention evaporates the moment anything else appears. Another person in the room seems to hold the frame of "we are working now" on my behalf, so I don't have to keep regenerating it.

It raises the stimulation floor. A lot of ADHD avoidance isn't fear, it's boredom — the task is too low-stimulation to engage with, so my brain goes shopping for something better. Another person's presence adds a small, constant background hum of input. Enough to make the room tolerable. Not enough to be a distraction.

It makes drifting slightly embarrassing. Not shameful — just visible. When I'm alone, opening a new tab costs nothing. When someone is three feet away with their own work open, the cost of visibly doing nothing ticks up just enough to tip the balance toward starting.

Notice none of those require the other person to do anything. That detail matters, because it's the reason substitutes can work at all.

The problem with needing a human

Here's where the best trick I know falls apart in practice.

The moments I most need to start something are the moments a body double is least available. It's 11pm. Or it's a Tuesday and everyone I know is at their own job. Or the thing I'm avoiding is embarrassing and I don't want a witness. Or — most often, and least flatteringly — arranging the body double is itself a task, and I can't start tasks. That's the whole problem. I need a solution that doesn't have a prerequisite.

There are real services for this. Focusmate pairs you with a stranger on a video call for a session you book ahead (25, 50 or 75 minutes, as of October 2026), and for a lot of people it genuinely works — I'd recommend trying it before you take my word for anything. But it's built on scheduling and on being seen, and both of those are load-bearing costs. If I have to book a slot with a stranger, I've added a social appointment to the thing I was already avoiding. Some days that's fine. On my worst days, which are precisely the days I need it, it's another wall.

So I spent about a year collecting substitutes.

How to body double alone: six ways when nobody can sit with you

Ranked roughly by how much of the effect they preserve for me. Your ordering will differ — the useful move is to try them badly a few times rather than pick the theoretically best one.

1. Go somewhere with strangers in it. A café, a library, a coworking desk, the quiet carriage. This is the closest substitute because it's not a substitute — it's real presence, minus any obligation to interact. The cost is transit, which on a bad day is the whole battle. But if you can get out the door, this is the strongest version.
2. A silent video call, left running. Ring a friend who also has things to do, both say "right, working now," and then don't talk. Cameras on, microphones on, no conversation. It feels absurd the first time and then it works. The unspoken rule that makes it survive past week one: nobody has to entertain anybody. Either of you can leave without it being rude.
3. Study-with-me recordings. Hours of someone silently working at a desk, of which there are thousands on YouTube. Weaker than a live call because the person can't perceive you, so the visibility effect disappears entirely — but the "we are working now" frame survives, and it's available at 3am with zero coordination.
4. Make the clock visible. A timer you can actually see, counting down, is a surprisingly good stand-in for a presence in the room. It's the one external thing that keeps insisting time is passing while you sit there. This is most of why Pomodoro works for ADHD people at all, and it's why I care so much about the timer being visible on the Lock Screen rather than buried in an app you have to open. If you want to try it right now, there's a free focus timer you can keep on screen on this site. It runs in any browser tab, with no account and nothing to install.
5. Narrate the first physical action out loud. Not the task — the first physical movement. "I am opening the document." Said aloud, in an empty room. It sounds ridiculous. It also externalises the intention in roughly the way another person does, and it's free. This one has the highest ratio of stupid-feeling to effective of anything on the list.
6. Something that checks in on you. The last gap I couldn't fill with any of the above: nothing on this list ever notices you're stuck. A café doesn't say anything when you've been staring at the same paragraph for nine minutes. This is the piece I ended up building.

What I ended up building

I make FocusDrop, an iPhone and Apple Watch app (a Mac app is coming), for ADHD adults, largely because I needed it. For a long time it had a swipe-based task deck and a focus timer and no answer at all to the starting problem — which was, embarrassingly, the problem I had built the app to solve.

So since version 1.4, in July 2026, FocusDrop has had a feature called Co-Focus (part of Pro; more on that below). It's an AI body double that lives on your phone: you start a session, and something stays with you through it. With Apple Intelligence, it breaks the task into steps small enough to be insulting and writes its check-ins for that task; without it, it uses the steps you wrote and a set of calm lines I wrote. Either way it checks in while the timer runs, and — the part I'm most attached to — behaves honestly when you haven't picked a task at all. No fake checklist, no pretending to know what you're doing. It just keeps you company while you work on whatever it is. Since version 1.6.1 it also remembers your recent sessions on a task.

Two honest caveats, because this is exactly where a blog post starts lying to you.

It is not a person. It cannot perceive you, and it doesn't create the social-visibility effect that makes a café or a live call work. If you can get to a café, go to the café. Co-Focus is for 11pm, for the days leaving the house is not happening, and for the tasks you don't want a witness to.

The AI runs on your phone. Not because that's a nice bonus, but because I couldn't stomach the alternative. The whole point is company during your least composed moments, and I'm not routing that through a server so it can be logged. It uses Apple Intelligence on your iPhone, so nothing you type is sent to an AI service or to me. If you leave iCloud sync on, your tasks and session history sync through your own private iCloud, which I can't read. There's no account, and nothing for me to look at.

An AI body double on your iPhone (no call, no camera, no stranger)

If you landed here looking for an AI body double, here's the plain version, checked against the app on 23 October 2026.

What it is

Co-Focus, a companion inside FocusDrop that stays with you through a focus session and checks in as it runs. Software, not a live person.

Where

iPhone only, on iOS 18 or later. FocusDrop's Apple Watch app doesn't have it.

Price

Pro only (prices above, US).

Apple Intelligence

Not needed for Co-Focus itself, which falls back to built-in prompts. Its personalised coaching lines and step breakdowns need Apple Intelligence: iPhone 15 Pro or later on iOS 26 or later.

Camera and call

None. FocusDrop doesn't use your camera or microphone, and there's nobody to book or meet.

Privacy

The AI features run on your device through Apple Intelligence; nothing you type is sent to an AI service or to me. With iCloud sync on, your tasks and session history sync through your own private iCloud, which I can't read. There's no account.

In the app since

Version 1.4, July 2026.

If you want a human body double, try Focusmate first: it pairs you with a real person on a video call. Co-Focus is for the hours when that isn't going to happen.

One naming note, because the names get mixed up: Co-Focus is FocusDrop's on-device AI companion. It is unrelated to Cofocus, the video co-working site, and to OVR IT's Co-Focus.

For the other camera-free options side by side, including the ones I didn't make, see Body doubling apps without a video call, compared.

If you try one thing

Pick the ugliest version of any of these and do it badly today. Ring someone and don't talk to them. Say "I am opening the file" out loud like a lunatic. Take the laptop to the café even though it means buying a coffee you don't want.

The reason I trust body doubling over every other technique I've tried isn't that it's clever. It's that it doesn't ask me to become a more disciplined person first. It changes the room instead of the person in it, and that turns out to be the only intervention my brain has never learned to route around.

What's the substitute that works on you? I'm always collecting more — the six above are just the ones that survived contact with my own worst days.

---

Common questions

What is body doubling in ADHD?

Working on your own task while another person is present, doing their own thing. They don't help you, supervise you, or hold you accountable in any formal way — they're just there. For a lot of ADHD adults that presence is the difference between starting a task and circling it for three hours.

Why does body doubling work?

Nobody has proven a single mechanism and the formal research is thin. The three explanations that best match my experience: it moves the intention out of your working memory, it raises the stimulation floor enough for an understimulated brain to engage, and it makes visibly drifting off slightly more awkward than starting.

Can you body double alone, without another person?

Partly. Full presence seems to matter, but several substitutes recover a useful share of the effect: a café or library, a silent video call left running, a study-with-me recording, a visible countdown, narrating your first physical action out loud, or an app that checks in during a session. None fully replaces a person — but each one removes the scheduling problem, which is what makes body doubling fail in practice.

Is there an AI body doubling app?

Yes, a few, including Co-Focus in FocusDrop for iPhone. It's a software companion that stays with you through a focus session and checks in as it runs, with no call, no camera and no stranger. It's part of FocusDrop Pro and iPhone only, and its personalised coaching lines need Apple Intelligence (iPhone 15 Pro or later on iOS 26); other iPhones get its built-in prompts. It is not a person: if you want a human body double, Focusmate pairs you with one on a video call. Co-Focus is FocusDrop's on-device AI companion. It is unrelated to Cofocus, the video co-working site, and to OVR IT's Co-Focus.

Is an AI body double as good as a real person?

I don't think anyone can honestly claim that yet, and I'd be suspicious of anyone who does — including me. The one VR study that compared them found both beat working alone, with participants split on which they preferred: some wanted the realistic social pressure of a human, others valued being free of judgement. Twelve people in a headset, though. What I'd say from using mine daily is that it's clearly worse than a friend at the same table and clearly better than nothing at 11pm, and that the gap between "nothing" and "something" is the one that actually decides whether I start.

Does FocusDrop's AI body double work offline?

Yes. FocusDrop's Co-Focus (a Pro feature, iPhone only) makes no network call: its built-in prompts ship inside the app, and on an iPhone with Apple Intelligence the personalised lines come from Apple's on-device model. There's no FocusDrop server for it to reach.

How is body doubling different from an accountability partner?

An accountability partner is about consequences after the fact — you report whether you did the thing. Body doubling is about the moment of starting and involves no reporting at all. Your body double often doesn't know what you're working on, which is exactly why it doesn't add its own layer of dread.