StoweBody doubling is doing your own task while another person is present — in the room, on a video call, or in a shared online session. The other person is not helping, supervising, or even necessarily talking to you. They are just there, and for many people, especially with ADHD, that is the difference between starting and not starting.
On paper it should not work. The task is the same, your abilities are the same, the double contributes nothing. And yet it is one of the most consistently reported focus strategies in ADHD communities — common enough that a small industry of matching services and co-working apps now exists to supply strangers on demand. What follows is why presence plausibly works, what the research honestly shows, and how to try it for free before paying anyone.
Why does another person's presence help?
The most useful frame is external scaffolding: presence supplies, from the outside, some of what executive function is supposed to supply from the inside. Starting a task alone requires you to generate the start signal, hold the intention, and resist every competing pull — precisely the machinery ADHD taxes. Another person changes the physics of the moment. The session has a start line, drifting off to your phone now has a tiny social cost, and the room contains one running example of "person working" to mirror.
None of this requires the double to do anything, which is what separates body doubling from accountability. An accountability partner checks your output. A body double just exists nearby. The pressure is ambient and mild — for many people exactly mild enough to help without tipping into the anxiety that a supervisor would create.
What does the research actually say?
Honestly: it is community knowledge that research has begun to document, not a validated intervention. The first substantial academic work is Eagle, Baltaxe-Admony and Ringland (2024), who surveyed 220 mostly neurodivergent people about the practice. Participants consistently described it helping them start, stay on, and finish tasks, across a spectrum of forms — same room, video call, matched strangers, recorded videos — and the authors propose the external-scaffolding account above. What does not yet exist is a controlled trial measuring the effect on task completion.
There is older, adjacent evidence that presence as such changes performance: social facilitation, studied since the 1960s (Zajonc, 1965), found that being observed tends to improve simple or well-practiced tasks and impair complex, unpracticed ones. That is not a body doubling study, but it makes two useful predictions that match what people report — presence has real effects, and those effects have conditions.
So the accurate sentence is: widely and credibly reported, mechanistically plausible, evidence emerging. If it works for you, the absence of an RCT is no reason to stop. It is a reason to be suspicious of anyone selling it as proven.
How do you try body doubling without buying anything?
Start with people you already know and tasks you already avoid. The cheapest experiments, roughly in order:
- Same room, separate work. Sit where someone else is working — a partner answering email, a library, a café. Announce nothing. Do your task.
- A silent video call. One friend, cameras on, microphones off, each doing your own work for a set block. Say what you are doing at the start, say whether it happened at the end.
- "Study with me" videos. A recorded person working, with a timer. The weakest form of presence — and still enough for some people.
- Matched-stranger services. Focusmate and similar pair you with a stranger for a timed block. This is the paid tier of the same mechanism; try the free forms first so you know whether the mechanism works for you at all.
Two rules make the experiment fair. Pick the task before the session starts — a double helps you do a chosen task, not choose one. And use it for work where presence helps rather than hurts: clearing, sorting, and admin respond best; delicate creative work sometimes wants an empty room.
Where does planning fit in?
Body doubling solves the staying problem, not the deciding problem. A double beside you while you stare at forty undifferentiated tasks is two people watching you freeze. The session works when it has one pre-decided task — which makes the pairing obvious: decide first, then borrow presence for the doing.
The deciding half is the problem I built Stowe around, so I will say plainly where it sits: it is not a body doubling app — no rooms, no strangers, no camera. It handles the step before: you talk out the pile, on-device AI sorts it, and today holds 3–5 chosen things — which is what a focus session wants as its input. Toki, the small creature that listens while you talk and sits by while you work, is the gentlest possible cousin of the same idea: something is present, and the moment is witnessed. I would not claim a drawn creature does what a human does. The method that transfers is the pairing: one pre-decided task, plus some form of presence, is a repeatable way to start.
References
- Eagle, T., Baltaxe-Admony, L. B., & Ringland, K. E. (2024). "It was something I naturally found worked and heard about later": An investigation of body doubling with neurodivergent participants. ACM Transactions on Accessible Computing, 17(3), Article 16.
- Zajonc, R. B. (1965). Social facilitation. Science, 149(3681), 269–274.
Related: Planning methods: what the evidence actually supports · Task paralysis: why a long list freezes you · What is a brain dump, and what should happen next
Frequently asked questions
Does the body double have to do anything?
No — that is the strange, load-bearing part. The double does not supervise, help, or check your work. They exist nearby, usually doing their own thing. The presence itself is the mechanism: it gives the moment a start line, a mild social stake, and a reason not to drift, without anyone applying pressure.
Does body doubling work over video, or only in person?
People report both working. The survey research that exists (Eagle et al., 2024) deliberately covered a spectrum — in the same room, on a video call, in a stranger-matching service, even a recorded "study with me" video. Which works for you is individual; the common thread is that some other human is plausibly aware you are supposed to be working.
Is body doubling scientifically proven?
Not in the controlled-trial sense, and it is worth being plain about that. What exists is credible documentation: a survey of 220 mostly neurodivergent people consistently describing that it helps them start, stay on, and finish tasks (Eagle et al., 2024), plus older adjacent findings that mere presence changes performance. Widely reported, mechanistically plausible, no RCT yet.
What if being watched makes me perform worse, not better?
Then trust that reading — it has research behind it. The oldest finding in this area (Zajonc, 1965) is that presence helps practiced, simple work and can hurt complex, unpracticed work. If a double helps you clear email but wrecks your writing, use it for the email and write alone. Body doubling is a tool with conditions, not a universal setting.