Why do time tracking apps make you choose a category before you start?

Before a timer will run, it needs an answer to a question the next hour has not gotten around to answering yet: what is this? Work, or a specific piece of work? An errand, or the start of something else entirely? You have to pick a box before you have done the thing that would tell you which box it belongs in — a guess, made in the first few seconds of an activity you have not actually lived through yet.

The label comes due before the evidence does

Normally you would decide what something was after it happened, once you had seen the shape it actually took. A timer reverses the order. It wants the category at the moment of starting, which is the one moment you know the least about what is about to unfold. You are not describing an hour; you are predicting one, and predictions about your own day are wrong more often than the tidy category list would suggest.

Most stretches of a day do not stay in the box you put them in

Real time does not sort itself neatly. A block you started as "work" turns into a phone call with a friend partway through. An "errand" turns into forty minutes of research once you are standing in the shop trying to decide what to buy. None of this is unusual — it is just what an hour is actually like, several small things arriving inside what looked at the start like one thing. A category chosen in advance has no way to account for a shift that has not happened yet, so the label and the hour start drifting apart from the moment the clock starts running.

Guessing wrong has a cost that writing it down afterward does not

Once a timer is running under a category, correcting it mid-stream feels like admitting the first guess was wrong, so a lot of people just let it ride — the mislabeled block stays mislabeled, or gets stopped and never properly restarted under anything else. Do this often enough and the record you end up with is not a picture of the day, it is a picture of your first guess about the day, repeated and quietly compounding across weeks until the totals describe an hour that never quite happened the way the label says it did.

The upfront decision becomes its own kind of friction

There is a second cost, and it shows up before the timer even starts. If you are not sure yet what to call something — is this "work" or "admin," is this "health" or just an errand that happens to involve a pharmacy — you can end up standing at the category screen for longer than the task itself would have taken, or putting off starting the timer at all until you have sorted it out in your head. The category picker was supposed to be the fast part. Instead it becomes one more decision standing between you and just doing the thing.

Naming something after it happened is a different, easier job

Once an hour is over, you already know what it was. Describing it is no longer a prediction, it is a report. "Helped with paperwork, then answered emails, then a short walk" is one sentence to write in hindsight, even though guessing that exact mix in advance, at the moment you would have needed to start three separate timers for it, would have been close to impossible. Hindsight is not a trick — it is just what you already have once the thing is done, and a log is built to use it. Where a timer needs you to know the future, a line written afterward only needs you to remember the last hour, which is a much smaller ask.

TimeMap works from that side of the problem. A line goes in the calendar with a time and a category colour attached to it, but the category gets picked when you are writing about what already happened, not when you are about to start something you have not done yet. There is nothing to predict correctly and nothing to correct mid-stream if the shape of the hour turns out different from what you expected walking in — you just describe it once, when you actually know what it was.

The category screen is supposed to help you see a pattern later, not force you to forecast one before it exists. Moving that decision to after the fact is a small change, but it removes a specific kind of friction that has nothing to do with the activity itself and everything to do with being asked to know something too early.