Routines
Doing the same kind of work at the same time of day
Grouping similar tasks removes the cost of switching between different modes of thinking. The gain is largest for the tasks people find most tedious.
Comparisons of batching similar tasks usually pick a winner. This one picks the circumstances, which is more useful.
The difference in one place
- Switching between task types carries a setup cost that repeats every time.
- Batching works best for small, similar, low-stakes tasks rather than deep work.
- Attaching a batch to a fixed time removes the daily decision about when.
The cost that repeats every switch
Moving between different kinds of task requires reloading context: the relevant files, the tone, the people involved and what you had already decided. That reload takes time and attention, and it happens again in full every time you return, which is why a fragmented afternoon produces so little. The cost is largest when the two tasks use similar mental machinery, because the previous one keeps intruding on the new one.
People notice the interruption itself and rarely notice the several minutes of reassembly afterwards, so the true cost is systematically underestimated. Grouping tasks of one kind together pays that setup cost once instead of five times, which is the whole mechanism in a sentence.
What batches well and what does not
Short, repetitive, similar tasks batch extremely well: replies, forms, calls, invoices, bookings and anything with a template behind it. Demanding creative or analytical work batches poorly beyond a point, because concentration has a duration and stacking four hours of it does not produce four hours of output. Tasks with genuine time dependencies cannot be batched at all, and forcing them into a slot creates delays elsewhere that cost more than the switching saved.
The useful distinction is between work that benefits from momentum and work that benefits from freshness, and batching serves the first far better. Trying to batch everything is the common failure, and it usually collapses within a fortnight and takes the useful parts down with it.
Attaching batches to fixed times
A batch with no fixed slot competes for attention every day, and that daily negotiation costs more energy than the tasks themselves. Anchoring it to an existing marker in the day — after lunch, before the first meeting, the last half hour — removes the decision entirely.
For most people, matching the batch to your own energy pattern matters: admin suits the flat parts of a day, and defending the sharp parts for harder work. A weekly slot works better than a daily one for anything that does not accumulate quickly, and it is easier to protect from encroachment. The slot needs a defined end as well as a start, or the batch expands to fill whatever time follows it.
Batching outside work
Errands cluster geographically as well as by type, and one route covering four stops costs far less than four separate departures from the house. Cooking more than one meal at a time uses the same heat, the same chopping and the same washing up, which is the domestic version of the same principle. Household admin — renewals, appointments, forms — benefits most, because these are the tasks people postpone until they become urgent and expensive.
In practice, phone calls in particular batch well, since the reluctance is mostly about starting and the second call is much easier than the first.
The pattern to avoid is batching things that are actually pleasant, because reducing their frequency reduces the benefit rather than the cost.
Where it goes wrong
Batching delays individual items by design, so anything with a genuine deadline inside the batch interval will be late. People sometimes use it to justify not answering anybody for days, which shifts the cost onto colleagues rather than removing it. A batch that has grown too large becomes intimidating, and an intimidating batch gets postponed, which recreates the backlog it was supposed to prevent.
In practice, the fix is a shorter interval rather than more discipline, because the interval is the variable actually controlling the size. It also fails for anyone whose day is genuinely reactive, and pretending otherwise is a common source of guilt in jobs built around interruption.
Some of this will suit you and some will not, and that is the point.
Setting one up this week
Start by listing a week of small tasks and marking which type each belongs to, which usually reveals two or three obvious clusters. Pick the single most annoying cluster and give it one slot, rather than redesigning the entire week at once.
Put simply, keep a running note for items belonging to that batch, so they stop being handled individually as they arrive. Give it three weeks before judging it, since the first week will include a backlog and the second will still feel unfamiliar. If a batch is repeatedly skipped, the slot is in the wrong place, and moving it is far more effective than resolving to keep it.
Side by side
| Consideration | What it means in practice |
|---|---|
| The cost that repeats every switch | Switching between task types carries a setup cost that repeats every time. |
| What batches well and what does not | Batching works best for small, similar, low-stakes tasks rather than deep work. |
| Attaching batches to fixed times | Attaching a batch to a fixed time removes the daily decision about when. |
The takeaway
Pay the setup cost once. That is the entire idea.
Small and repeatable beats ambitious and abandoned, almost every time.
Questions readers ask
Does batching work if my day is full of interruptions?
Partially. Even a protected half hour delivers some of the benefit, though jobs built around responding to others will never batch as cleanly as jobs built around output.
How long should a batch be?
Long enough to justify the setup and short enough to finish in one sitting, which for most admin clusters lands somewhere between twenty and sixty minutes.
