Focus in a cafe: the distraction nobody names
Working in a coffee shop fixes the noise and creates a different problem: you are surrounded by strangers who have no way to know whether you are available. This is about the second problem, and about what actually helps, cheapest option first.
A café is a good place to focus, and for a specific measurable reason: the noise level is roughly right. A series of five experiments, summarized by David Burkus in Harvard Business Review, found the sweet spot for creative work sits near 70 decibels, close to the murmur of a busy coffee shop. So people go there to escape the open office and they are right to. What they usually find instead is that the ambient noise was never the main cost of the office. The main cost was being approached, and a café does nothing about that — it just replaces colleagues with strangers who are even less able to tell whether you are busy.
Key takeaways
- Creative work peaks near 70 dB, roughly the murmur of a café. Too quiet (~50 dB) and too loud (~85 dB) both lose the benefit.
- Speech is the damaging part, not volume. Noise with speech broke prose memory as well as arithmetic; noise without speech broke only arithmetic.
- In a 2018 survey of 2,014 office workers, conversations were named as a noise source by 71% — ahead of phone calls and ringing phones.
- Cafés solve sound and create approach. Headphones block sound, not approach, so they do not address it either.
- The cheapest fix is a different seat or a different hour. The most reliable is a visible signal that a meeting is in progress.
Why people go to a café in the first place
The usual explanation is that the coffee shop is quiet. It is not. It is measurably noisy, and by several measures more acoustically complex than the average office: the espresso machine, the door, three conversations you cannot follow. The claim that people go for a "calm atmosphere" is a post-hoc story attached to a place that mostly functions as an acoustic mask.
The five-experiment series points somewhere more useful. Moderate, continuous, unpredictable background noise appears to help creative work relative to total silence, and the mechanism usually offered is that it masks the small internal interruptions that compete for attention. Around 70 dB you get enough masking without the masking becoming a load in itself. Both ends fail: around 50 dB the benefit largely disappears, and around 85 dB it is gone entirely.
This is one research programme, reported secondhand through a business publication rather than read here in the original papers, and "70 dB" is a rounded summary of a range measured in laboratory conditions. It is good enough to explain why people move to cafés. It is not a target to optimise for, and it says nothing at all about interruptions.
What the noise research actually says
The second, better-established finding is that it is not volume that matters. Banbury and Berry (1998), in the British Journal of Psychology, played office noise in two conditions. Noise containing speech degraded performance on both prose memory and mental arithmetic. Noise without speech degraded only mental arithmetic. Their conclusion was that noise distracts through its unpredictability, and that intelligible speech is the worst case of it.
Morrison and Macky have argued the same line from the open-plan side: the consistent finding across studies is that overhearing other people's conversations is the defining problem of open offices, and that it tracks negatively with performance, stress and how people feel about the place they work. A café is full of conversations you cannot follow. So is an office. The difference is that in the office the ones you overhear are your colleagues, who will come and find you afterwards.
Read that ranking again with the Banbury and Berry result in mind. The four most-cited office noise sources are all people, and the top one is people talking. A soundproofed office with a door is a different proposition from an open floor, which is why a review of the open-plan literature by James, Delfabbro and King (2021) keeps finding the same consistent losses in satisfaction, acoustic privacy and focus. The European Parliament briefing on open-plan offices reaches the same conclusion from the policy side.
And on background music, the evidence is genuinely mixed, so you should be sceptical of anyone selling a playlist as a productivity intervention. Johansson et al. (2012) found non-preferred music hurt performance while café noise had no measurable impact at all. Furnham and Strbac (2002) found performance dropped against silence on cognitive tasks, with introverts affected more than extroverts. A systematic review in Liverpool covering 95 articles and 154 experiments found an overall negative effect on memory and language tasks, worse with lyrics, and worse on difficult tasks. Music with a beat can be a useful mask. It is not a substitute for removing the actual problem.
The signal that strangers do read correctly
One person talking, two listening, one active speaker. Everyone who has sat in a meeting knows what this is, which is the entire value of the signal.
Start a meetingThe problem cafés do not solve
In an office, an interruption is expensive for the person causing it. They have to walk over, interrupt a conversation, and possibly remember to follow up. The office is a place where asking a stranger a favour has a social price, which is why colleagues mostly hesitate. That hesitation is the only real thing protecting you, and it is invisible in every productivity framework because it is not a mechanism, it is a courtesy.
A café removes the price and keeps the request. Someone needs a charger, someone wants to show you a photo, someone wants to know if this seat is free, someone wants you to look at a screen because the words "have you got a minute" are the only tool they have for getting your attention. None of them are being malicious. All of them are cheap for them and expensive for you, and none of them can tell from twenty feet away that you are mid-paragraph.
Headphones are the standard answer and the standard answer is aimed at the wrong target. They reduce sound. The café problem is approach, which is visual and social. Someone reads your headphones as "someone who is listening to something", which is a weaker signal than a call, and a stranger who wants to ask you something will still ask you — more likely, in fact, if you are visibly absorbed and they assume you have not noticed them. You also cannot hear the person who genuinely needs you, which is a real cost in a place full of people you may end up needing.
A ladder of options, cheapest first
Several of these have nothing to do with any tool, and several are better than the tool. The order below is roughly order of cost, and roughly order of reliability.
- Change seat or change hour. Free, immediate, and more effective than most people expect. Distance from the door, the counter and the main walkway removes most approaches. Mid-morning on a Tuesday is a different room from 4pm on a Friday, and you can know which is which after two visits.
- Use the noise as a cue for yourself. The 70 dB finding suggests moderate sound can help creative work. Lean into that: headphones off, ambient noise on, and treat the café as an instrument for creative tasks rather than for exacting analysis.
- Make a visible signal. A notebook open beside the laptop, a second monitor showing your actual work, a "busy" calendar status, a sign. Cheap, and it works much better than a verbal request because it does not require the other person to cooperate with anything.
- Book the room. In a coworking space or a meeting room, this is decisive. A closed door removes every ambiguity at once, and you can say "I'm in a meeting all afternoon" because it is true.
- Put a simulated meeting on screen. The strongest legible signal that exists in a public place, because a live call is a signal every reader already has a template for. MeetingSimulator is free and needs no account. Use it for the block where the first four options have failed.
Where the tool is genuinely weak is exactly where you should expect weakness: a stranger has no context in which your meeting is real. A colleague in your own office will assume it is real. Someone at the next table has no reason to. If you care about that, the honest answer is that a visible sign or a booked room is more honest and equally effective, and you should reach for those first.
Coworking spaces specifically
A coworking space is a café with better acoustics and worse escape routes. You have paid a monthly fee for a desk, and a room is usually the thing you actually want — so you book it, and then the booking is the bottleneck. The European Parliament briefing notes that around 32% of workers are in open environments while only 31% of employers provide rooms suitable for calls. The scarce resource is not attention. It is a door.
Two practical consequences. First, check the booking situation on day one rather than discovering it on the day you needed the room — the ones that exist are heavily contested at peak hours. Second, if you cannot get a room, you are back to the approach problem, and the ladder above is the answer. The full analysis of the productivity tool and how interruption actually works is worth reading if you want the mechanism behind the signal.
How to look busy without anyone noticing
Look busy is the wrong frame, and it produces the wrong behaviour. Trying to look occupied is a performance that people learn to read, and it also means you never actually settle into the work. The version of this that holds up is the one where you stop managing appearances and change a fact instead: something visible is genuinely happening on your screen, so there is no appearance to maintain. We wrote about the mechanics in how to look busy at work without anyone noticing — including why the common version fails and what the alternatives are.
Why your colleagues keep interrupting you anyway
Because a café solves the wrong half of the problem, and the same is true of every personal boundary you have tried. Interruptions in the office are not a volume problem, and they are not a willpower problem either: Gloria Mark's observational work at UC Irvine found the average time on a single work event before switching or being interrupted is 3 minutes and 5 seconds, with 44% of interruptions self-inflicted. Whatever is happening in a coworking space, it is happening for the same structural reasons.
The longer version of that is in why colleagues keep interrupting you, which goes through the research properly, including a widely repeated statistic that turns out to be misreported. If you are in a coworking space because the office was worse, that article is the more useful one.
Questions people ask
Why is it easier to focus in a cafe than in a quiet office?
A series of five experiments, summarized by David Burkus in Harvard Business Review, put the sweet spot for creative work near 70 decibels, close to the murmur of a busy cafe. Both ends fail: around 50 dB the benefit largely disappears, and around 85 dB it is gone. Mild, unpredictable background sound seems to mask the small internal interruptions that compete for attention, and total silence does not.
Is loud cafe noise better than the office I am trying to escape?
Sometimes, but not because of the volume. Banbury and Berry (1998) found that office noise containing speech disrupted both prose memory and mental arithmetic, while noise without speech disrupted only the arithmetic. The unpredictable, intelligible human voice is the damaging part, which is why an open office can be harder to work in than a busy cafe full of strangers.
Do headphones solve the cafe distraction problem?
No. Headphones reduce sound, and the cafe problem is mostly social. Someone still walks up, reads your posture, decides you look available, and asks to borrow a charger. You also lose the ability to hear anyone who genuinely needs you, and you signal unavailability in a way a stranger cannot interpret as a boundary.
What is the cheapest way to protect focus in a coworking space?
Change your seat or your time slot before anything else. Distance from the counter, the door and the main walkway removes most approaches for free. A booked meeting room is better still, but demand usually outstrips supply, which is the underlying problem in coworking: around 32% of workers are in open environments while only 31% of employers provide rooms for calls.
Should I use a simulated meeting screen in a cafe?
It works, and it is free, but be honest about the limitation: a stranger has no context in which your meeting is real, so they may read it as deception rather than as a boundary. Seat choice, headphones and a visible sign solve the same problem at no cost. Reach for the screen when those have already failed, and spend the time you win on the work.
Does background music help me concentrate in a coffee shop?
The evidence is genuinely mixed, so treat confident answers with suspicion. Johansson et al. (2012) found non-preferred music hurt performance while cafe noise had no impact, and Furnham and Strbac (2002) found performance fell against silence, with introverts affected more. A systematic review of 95 articles found an overall negative effect, worse with lyrics. Useful as a mask, unreliable as a fix.
Try it on your next focus block
Open the tool, pick the number of guests, press start, and get 45 uninterrupted minutes wherever you happen to be sitting. No account, no download, no cost.
Free forever No signup Real people, not AI Nothing leaves your device
Sources cited on this page: Burkus, D., Why You Can Focus in a Coffee Shop but Not in Your Open Office, Harvard Business Review (summary of a five-experiment series on noise levels and creative work); Banbury, T. & Berry, C. M. (1998), Disruption of office-related tasks by speech and office noise, British Journal of Psychology 89(3), 499–517; Morrison, M. & Macky, A. (open-plan overhearing research, as cited in the interface acoustics summary); Radius Global Market Research for Interface, What's That Sound? (n = 2,014, US/UK/AU, 2018); James, P., Delfabbro, P. H. & King, S. L. (2021), open-plan versus cellular office review, International Journal of Environmental Research and Public Health; European Parliament, briefing on open-plan offices (2020); Johansson, O. et al. (2012) on music and cafe noise; Furnham, A. & Strbac, E. (2002) on music, cognition and personality; University of Liverpool systematic review of 95 articles on background music; Mark, G., "Too Many Interruptions at Work?", Gallup Business Journal (2006).