Two of us decided to learn the same thing at the same time. The material and the starting point were the same, but the schedules were different. One studied on Saturday mornings, three hours with coffee and the door shut. The other studied for ten minutes after lunch on weekdays, at a desk with people walking past.
This was not a study, and we are not going to pretend it was. We have no research and no chart to show you. There were two people and one autumn, and a product decision came out of it.
What happened to the Saturdays
The first Saturday went very well. Three hours is enough to get properly into a subject, and by lunchtime he had covered more than his colleague managed in a fortnight.
The second Saturday was a friend's birthday. The third was moved to Sunday and then did not happen at all. By the fifth week he felt guilty about the session every Friday night, but the guilt did not make him study on Saturday. Instead he decided to start again properly next month. For an adult learner that is the most expensive decision there is.
He did come back. But most of the first session back was lost. After three weeks away from a subject you have forgotten the last thing you learned, and you do not know which part of it you have forgotten until you try to use it.
What happened to the ten minutes
Nothing special happened to them, and that is the point.
Some days he did two sessions. One day he sat down, read the question, realised he was too tired and closed the laptop. That cost him nothing, because there was no plan to break. Ten minutes is too small to be a promise, so there is no promise to fail.
The interesting part was coming back after a break. He had breaks too: a cold, and a week of urgent work. But the last thing he had done was a single question, not three hours of material, so to pick it up again he only had to remember one thing.
By the end of the autumn he was further along. Ten minutes is obviously not better than three hours in itself. His schedule simply survived an ordinary life, and the Saturday schedule did not.
Where the time actually goes
This is the part that changed the product. We asked both of them what they spent their time on, and the answers were completely different.
A Saturday session had a start-up cost. You open the material, work out where you stopped, decide what to do today and get back into the subject. In a three-hour session that cost is tiny and worth paying, and he never noticed it.
In a ten minute session the same cost takes up all the time. If you have to decide what to study, you spend the whole ten minutes deciding, and you do it at the most tired moment of the day.
So a short step only works if the decision has already been made when you open it. That is why the home screen shows one step and a line saying why this one, and not a list of things you could reasonably do next. Without the list, ten minutes is ten minutes of work. With the list, you spend them choosing.
What is inside a step
A step consists of a setup, a question, a hint if you ask for one, and a walkthrough if you get it wrong.
Hints come one at a time. A hint that gives away the answer does not help you, it just ends the task. The walkthrough starts from the point where your reasoning went wrong, not from the correct answer. That is slower to write, but it is the only kind that teaches anything.
When you get something wrong, nothing turns red and no mark is recorded. We have written about why a percent of you is the wrong number. With a step this small it matters even more: a number attached to ten minutes says how your day went, not how much you have learned.
What we refuse to do with your ten minutes
The day counter carries no weight. We count the days you practised in a row, and you can watch the number grow if you like. But nothing depends on it: there is no reward, no debt, and no message when the run of days breaks. If an app makes a missed day the main event of the week, a person who misses two comes back embarrassed or does not come back at all. Here a missed evening costs nothing and is never held against you.
There is no notification saying you have not studied. Such a message is a reproach from an app that knows nothing about your week. A reminder in the messenger you already use is fine. It says what the next step is, not how long you have been away.
There is no catching up. If you miss a week, nothing has piled up, because there was no schedule to fall behind. The route stays where you left it. When you return, it starts with a short review of the part that was shaky and then moves on.
Ten minutes is the minimum, not the limit
When we talk about this, we often make it sound like a limit. It is not one. If you have an hour, the route will give you step after step for the whole hour and will never suggest that you have done enough for today.
The number exists so that a bad day still counts. Ten minutes is the smallest amount of time in which you can really learn something. It is also short enough to use on a train, in a kitchen, or in the quarter of an hour between the children falling asleep and you falling asleep yourself. That last case is described in full in Andrei's story.
Where this argument breaks
Some things do not fit in ten minutes. You cannot write an essay in that time, or sit with a long piece of code until it makes sense. A mock exam, by definition, does not fit either.
For those, the route gives you a task that says plainly it needs a longer session. We prefer that to chopping one long task into a series of short ones that only look like steps. Cutting deep work into fragments is a different mistake from the one we are fixing, and it is just as real.
There is one more limit, and we should say it openly: a short step works because somebody else decided what the step is. If that decision is wrong, you will spend ten minutes a day moving efficiently towards something you did not want. The line under every step says why it was chosen. If the reason makes no sense to you, argue with it instead of carrying on.
