Time blocking: planning the day so it does not plan you

A day with no plan is not a neutral day. It is a day whose contents will be decided by whoever asks first and most insistently, because in the absence of a decision about what this hour is for, the loudest available request wins by default.

This is worth separating from a to-do list, which most people already have. A list tells you what needs doing and says nothing about when, so every time you finish something you re-choose from the whole list — and that choice is made with whatever information is most recent and most emotionally available, which is usually the last message you received. A list of twenty items is not a plan for a day; it is twenty opportunities to pick the easiest one.

The effect over a week is that the work which had someone waiting for it gets done and the work with no external pressure does not, regardless of relative importance. Nobody chases you about the thing that would have prevented next quarter's problem. Time blocking is mainly a mechanism for giving that work an advocate, by deciding in advance rather than in the moment.

The core of the method is that every working hour has a stated purpose before the day begins, written somewhere you will see it. Not every hour filled with a task — hours assigned to a kind of work, which is a looser and more survivable commitment.

Doing this the evening before rather than in the morning matters more than it sounds. Planned in the morning, the plan competes with the day's first messages and loses; planned the night before, it is waiting when you arrive, and you begin executing rather than deciding. It also stops the planning itself from becoming the first hour's work.

The purpose of assigning an hour is not to predict the future accurately. It is to make the trade-off visible when something arrives. Without a plan, agreeing to a request costs nothing observable. With one, you can see that saying yes means the pricing review does not happen, which is the information you need to decide properly — and quite often you will still say yes, having actually chosen rather than drifted.

Grouping similar work together — all the calls, all the invoices, all the email — is worth more than it appears, because the cost of switching between kinds of work is largely invisible.

The cost is not the seconds spent moving between tasks. It is that different work needs different mental state loaded, and rebuilding that state after an interruption takes time that does not feel like work at all. Twenty minutes of writing, then a call, then twenty more minutes of writing produces noticeably less than forty continuous minutes, and the shortfall is difficult to notice from the inside because both halves felt busy.

Email and messages are the case worth handling deliberately, because they arrive continuously and each one is individually trivial. Checked whenever they appear, they fragment everything else while feeling like responsiveness. Handled in two or three defined windows, the same volume takes less total time and stops interfering with the rest.

The honest limit: batching depends on being able to defer, and some work genuinely cannot be deferred — an operational role or a customer-facing shift has real-time obligations. The method then applies to whatever is left, which is usually more than it first appears.

Most days contain one or two hours of work that requires uninterrupted thought, and those hours are the ones most easily lost, because they have no external deadline and nobody waiting.

Protecting them means treating the block as a commitment of the same kind as a meeting, which is mostly a matter of what you do when something arrives during it. A block that yields to any request is not protected, and after a couple of weeks it stops being planned at all.

Practically: the block goes on the calendar so the time is visibly unavailable, notifications are off rather than dismissed, and the people who might interrupt know when you are back. That last part does more than it seems — 'available from eleven' is easy to accept, while an unexplained unavailability invites a second attempt.

Put it where your attention is actually good rather than where the calendar has space. A protected block in a window when you cannot think well is protected time producing poor work, which is a common way to conclude that the method does not work. If the only free space is a bad window, the answer is to move something else, not to accept the bad window.

A plan with every minute allocated fails on the first thing that overruns, and something overruns most days. The failure is not partial either: once the day is behind, every subsequent block is wrong, and the usual response is to abandon the plan entirely by mid-morning.

So unallocated time has to be in the plan deliberately, and it has to be genuinely unallocated rather than optimistically empty. A gap that exists because you have not yet decided what goes in it will be filled by the first request that arrives. A gap that exists because overruns and interruptions are certain is doing a job.

The amount is a matter of calibration rather than a rule, and it is easy to check: if you routinely end the day with several blocks untouched, there is not enough buffer, and the plan has been fiction for a while.

Underestimating duration is the related and universal problem. Almost everyone plans as though tasks will take the time they would take with no interruptions and no false starts. The fix is not better estimation, which is hard, but leaving room for the estimate to be wrong, which is easy.

It cannot create hours. If the committed work genuinely exceeds the available time, blocking will show you that clearly and repeatedly — which is useful information and not a solution. The response has to be fewer commitments, delegation, or a change in what gets promised, and no arrangement of blocks substitutes for one of those.

It cannot fix a wrong priority. A day executed exactly as planned, on work that did not matter, is a well-organised waste of a day. Blocking is about protecting whatever you decided to protect; deciding is a separate activity, and the fact that the plan was followed says nothing about whether the plan was right.

It cannot survive being treated as a contract. Plans that are wrong by ten o'clock are normal, and the useful response is to adjust the remaining blocks rather than to conclude the day is lost. People who abandon the method usually do so because a plan failed as a prediction, which was never its job.

And it will not make attention available where there is none. Blocking assumes you can direct your focus once the interruptions are removed. If the difficulty is that focus is not there even in quiet, protected time, that is a different problem, and better scheduling is not the tool for it.