How to Batch Small Tasks Without Losing the Morning
Small admin jobs do most harm by breaking into the hours that need continuity. A single fixed window late in the day keeps them contained.
By eleven you have submitted an expense claim, confirmed a meeting, filled in a form someone needed by Friday, and sent the two-line answer a colleague was waiting on. Each one took a few minutes and each one was genuinely necessary. Together they come to a couple of hours a week, and nobody is going to take them off your hands. Yet the morning has gone, and the piece of work you meant to start is exactly where it was when you sat down.
The small jobs themselves are rarely what did the damage. The damage came from letting them arrive one at a time, all through the day, into the same hours as work that needed an unbroken run, and that is how two hours of admin ends up producing a day with nothing substantial in it.
Why a dozen small jobs cost more than their minutes
Small jobs are expensive mainly because each one breaks into whatever else you were doing, and the minutes they take are the smaller part of the bill. Twelve five-minute tasks scattered through a morning add up to an hour on paper, but in practice they are twelve separate interruptions.
Sophie Leroy’s research on attention residue found that after switching tasks, part of your attention stays with the previous one and performance on the next suffers, especially when the earlier task was left unfinished. For a morning of scattered admin, that cuts in both directions. The report you were drafting leaves residue on the expense claim, and the half-answered email leaves residue on the report when you go back to it.
Mark, Gudith and Klocke found that people who were interrupted finished their tasks in less time and with no loss of quality, but reported more stress, frustration, time pressure and effort along the way. Applied to admin, the result explains why a fragmented day rarely looks costly on the calendar. Everything got done, and on time, and you are still wrung out by four with nothing much to point at.
So the aim here is to contain the small work rather than get rid of it. The jobs themselves are fine; where they land in the day is what needs changing.
Where the admin window belongs in the day
The admin belongs in a single fixed slot in the afternoon, after the demanding work, and that position matters more than how long the slot runs.
Most people have a stretch of the day when sustained concentration is unlikely but routine work is perfectly manageable. That stretch is the natural home for admin, because admin asks for attention in short bursts and very little continuity, and you were never going to use those hours for anything deep.
The instinct usually runs the other way. Clearing the small things first feels like getting ready, and it gives an immediate sense of order. What it actually does is spend the day’s best uninterrupted hours on work that never needed them, so the substantial task inherits whatever is left. That is the morning being eaten before nine, and it may be the most common way a well-meant day slips away.
Forty-five minutes at the same time each day is enough for most people. If it regularly runs short, that tells you something about the volume of admin, and it is worth dealing with that directly instead of letting the overflow spread back into the morning.
The fixed time matters as much as the fixed length. A window that drifts to wherever there happens to be space will be overridden on precisely the days it was meant to protect. Keeping the same slot daily, even when there is little to put in it, gives the rest of the day a boundary you can plan around with some confidence.
Why small requests feel more pressing than they are
Small requests feel urgent because they arrive with a sender, a timestamp and someone waiting, and none of that says anything about how much they matter. This is why batching takes deliberate structure and cannot be left to judgement in the moment.
Zhu, Yang and Hsee, describing what they called the mere urgency effect, found that people chose tasks with short deadlines over tasks offering a larger reward, even when the better payoff was plainly in front of them. For admin, the important detail is that the pull towards the time-pressured option did not wait for the urgent task to be worth more.
A quiet, important piece of work makes no such noise. Nobody is chasing it and nothing about it pings. If the two are left to compete as each one turns up, the message will win nearly every time, so the protection has to be built into the shape of the day before the competition starts.
Write it down now and do it in the window
Anything that arrives outside the window goes onto a list to be dealt with later, even if it looks like a ninety-second job.
That can feel wasteful, but the arithmetic favours the list. Ninety seconds of work plus the residue on either side of it costs considerably more than ninety seconds, whereas writing it down takes a few seconds now and the task simply joins the afternoon batch.
The exceptions should be few and stated in advance: something genuinely time-critical, or a named person you have agreed to stay responsive to. Naming them is what makes the rule hold for everything else, because an exception you never put into words tends to grow until it covers whatever is in front of you.
Grouping admin by the kind of work inside the window
Inside the window, it helps to group jobs by the kind of attention they need, which keeps the batch itself from becoming fragmented. Do all the written replies together, then everything that involves a form or a system, then the calls. Hopping between kinds within the batch brings back a small version of the residue the batch was meant to avoid.
The same reasoning applies to the hours outside the window, which is why the inputs need to be closed while the substantial work is under way. Reading a message rarely takes long, but every check produces a small decision about whether to act, and that decision is itself a switch, even when the answer is no. It is the reason checking the same inbox four times an hour costs so much when each check lasts only seconds.
What the protected morning still needs from you
A protected morning still needs a decision about what it is for, because containing the admin only solves half the problem.
If you batch small tasks successfully but have not chosen what the freed hours are for, they tend to fill with the most manageable work within reach, which is often more admin dressed up as something weightier. The window keeps the small things out, and it does nothing on its own to put the right thing in.
So the batch works best alongside a decision made in advance, in writing, the evening before: one line, specific enough that you can begin without deliberating. Then the morning holds a known piece of work and the afternoon holds everything else, and neither has to negotiate with the other while the day is happening.
The Do One Thing diary is built for that half of the job, with a printed page that holds one line for tomorrow, written tonight. It has no room for the batch and makes no attempt to hold it. The small things belong on a list, and the list was never the part that failed.
Admin will always be part of the day. It only starts to cost you the morning when it is allowed to trickle into the hours that needed to stay whole.
