I once stopped work mid-sentence because a timer told me to, stood in the kitchen for five minutes feeling extremely disciplined, came back, and could not remember what the sentence was going to say.
I had been stuck on that paragraph for forty minutes. It unstuck itself about ninety seconds before the bell. I obeyed the bell anyway, because that was the rule, and the rule had quietly become more important than the work.
The Pomodoro Technique works because of three things (a defined start, a commitment small enough to accept, and a break you cannot skip) and none of those three require the interval to be 25 minutes. Francesco Cirillo, then a university student, picked 25 because that is roughly what he could get out of a tomato-shaped kitchen timer. It is a perfectly reasonable starting number. It is not a finding about human attention, and treating it as one is how people end up interrupting their best work on a schedule.
Cirillo's own account is refreshingly unmysterious. He was struggling to study, challenged himself to focus for ten minutes, grabbed the kitchen timer to hand, and the technique grew from there into the 25 on, 5 off, long break after four cycles structure people know today. There is no experiment behind the 25, and no attention research that produced it. There was a student, a kitchen, and a plastic tomato.
Which is fine. Defaults have to come from somewhere, and an arbitrary default that gets people started beats a perfect number nobody has. The problem is what happens when the default gets promoted to doctrine, and people start calling a good working session a failed pomodoro because they went twelve minutes over.
If you strip the technique down, three things are load-bearing. Everything else is decoration.
A defined start. Most people do not have a distraction problem so much as a starting problem, and a timer converts a vague "I should work on this" into a specific physical action: press the button. The session begins because you pressed something, not because you finally felt ready. That is the most underrated part of the method.
A bounded commitment. "Work on the report" is an open-ended obligation with no visible end, and open-ended obligations are expensive to accept. "Work on the report for 25 minutes" is a small, finite thing you can agree to while still dreading it. The bound lowers the activation energy, which is exactly why the technique is so good at getting people into work they have been avoiding for days.
An enforced break. Left alone, most people either take no breaks and degrade for hours, or take breaks that expand into the afternoon. A scheduled break removes that decision, and the value is less the rest itself than the fact that stopping and restarting are both decided in advance.
Keep those three. Everything else is negotiable, including the number on the timer.
It interrupts flow on a schedule. Deep concentration has a ramp. You spend the first ten or fifteen minutes loading the problem back into your head, and the good work happens after that. A 25 minute block often means ten productive minutes, a stop, and then the loading cost again. Stopping at the exact moment the hard part clicks is not disciplined. It is expensive, and you do not always get that state back.
It suits shallow work far better than deep work. This is the pattern I would bet on: the technique performs best on work you are avoiding and worst on work you are absorbed in. Email, admin, revision, marking, tedious refactors, the tax return. All of it responds beautifully to 25 minute bursts, because the working set is small and the resistance is high. Original writing, architecture, composition, hard problem-solving? The same 25 minutes is a knife through the middle of the thing you were building.
Timer-watching replaces working. Once the interval becomes the unit of achievement, people start managing the interval: glancing at the countdown, refusing to start a task at 3:52 because it would not be a clean pomodoro, counting completed tomatoes in a spreadsheet and feeling productive about the spreadsheet. The scaffolding turns into the building.
Five minutes is often the wrong break. After genuinely hard concentration it is not a rest, it is a pause. After light work it is enough to lose the thread for no benefit. A fixed break length has the same problem as a fixed work length: it assumes every hour of your day is made of the same material.
Here is the decision I actually make, which takes about four seconds before a session.
| Type of work | Interval that works | Why | |---|---|---| | Dreaded admin, email, forms | 15 to 25 minutes | The bound is the point. You need the commitment to be small enough to accept. | | Studying, revision, practice problems | 25 to 40 minutes | Short enough to keep retrieval sharp, long enough to finish a topic. | | Editing, review, structured planning | 40 to 50 minutes | Needs judgment but has a small working set, so it survives a stop. | | Writing, design, hard debugging | 60 to 90 minutes | Long ramp, large working set. Below an hour you mostly pay loading costs. | | Open-ended creative or research work | No countdown at all | You cannot predict the natural seam. A countdown will land in the wrong place. |
Two things to notice. First, 25 is genuinely good for a real category of work, it is just not good for all of them. Second, the bottom row breaks the technique entirely, and it is the row most likely to contain your best work.
Interval length is personal as well as task-shaped, and the honest way to find yours is to measure rather than guess.
For one week, use a stopwatch instead of a countdown on your main work. Start it when you actually begin, note the first moment you feel real pull toward something else, and keep going until you genuinely cannot hold the thread. Two numbers come out: where concentration first wobbles, and where it stops. Do that four or five times and the range is obvious.
Then set your interval a little below the second number, not at it. Ending while you still have something left is what makes the next session easy to start. Ending in collapse makes tomorrow feel like a punishment.
Sleep, caffeine, stress, and how interesting the work is all move that ceiling around. It is a range rather than a constant, and anyone insisting their number is 52 minutes on principle has made the same mistake as the people insisting on 25.
The risk of abandoning fixed intervals is that you abandon the three mechanisms with them and end up with no structure at all. Keep them explicitly.
Building the sequence in advance is worth the two minutes it costs. In Deep Focus the Session Planner is a drag-and-drop list of focus and break blocks with a duration on each, so a morning can run 20 warm-up, 10 break, 75 deep, 20 break, 45 cleanup, which is nothing like 25 and 5 and much more like how a real morning goes. Blocks auto-advance, so there is no decision at each transition, and each focus block carries its own profile, so the warm-up block can leave your inbox reachable while the 75 minute block does not.
Most break advice is "step away from the screen," which people ignore because it is vague. The useful version: a break should restore the resource the work just spent, and different work spends different things.
After hard concentration the resource is attention itself, and the worst refill is a feed, because that is attention-intensive in exactly the same way the work was. You come back with less than you left. What helps is boring on purpose: walking, staring out of a window, stretching, making a drink, standing outside. After sitting still, the resource is physical, so move. After talking to people all morning, be alone.
One rule saved me a lot of afternoons: no messages during short breaks. A five minute break containing a Slack thread is a task switch with a rest-shaped label on it, and you return carrying the thread instead of the work.
I want to be fair to the technique, because it deserves it.
If you cannot currently work for twenty minutes without reaching for your phone, 25 and 5 is an excellent intervention and better than anything more sophisticated. If the task is boring and mandatory, the small bound is the only thing that will get you moving. If your attention genuinely falls apart after half an hour, your ceiling is thirty minutes and the standard interval was right all along. In a shared space with unavoidable interruptions, short blocks lose less when one gets destroyed.
It is a good default and an excellent on-ramp. The failure is only ever in the last word of "Pomodoro Technique, forever, for everything."
For starting work and for getting through tedious tasks, yes, and the reason is the low activation energy of a small bounded commitment rather than anything specific about 25 minutes. For sustained creative or complex work it often costs more than it returns, because it interrupts on a schedule that has no relationship to the work. Judge it by task type instead of adopting or rejecting it wholesale.
Task-matched intervals: a countdown for work you resist, a stopwatch with a minimum duration for work you sink into, and a break length chosen at the same time as the work length. You keep the defined start, the bounded commitment, and the enforced break, and you drop only the fixed number. For most people that means something in the 45 to 90 minute range for deep work and 15 to 25 for admin.
Because the timer has no idea what you are doing, and the interval is measured from when you started rather than from when your concentration arrived. Deep work needs a ramp, so a 25 minute block can easily contain ten good minutes wrapped in loading and unloading. If this keeps happening, your interval is shorter than your ramp plus a useful stretch of work.
Long enough to cover the ramp-up plus real progress, short enough that you stop before you are wrecked. For most people that lands between 45 and 90 minutes for demanding work. Test it with a stopwatch for a week rather than adopting someone else's number, and set the interval slightly below the point where your attention collapses.
No. Finish the thought, reach a natural seam, then break, and take a slightly longer one to compensate. The real risk of overriding a break is doing it every time and burning through the afternoon, so treat it as an allowed exception rather than a new rule.
Every productivity system starts as a tool and drifts toward being an identity, and that drift is where the value leaks out. The point was never to complete pomodoros. It was to make the thing.
Keep the timer. Move the number. Stop when the work is done, not when a tomato says so.
Get the latest productivity tips and Deep Focus updates delivered to your inbox
