Weekly planning that survives Tuesday

Sunday evening, and the week looks reasonable on paper. Three big things, two calls, gym on Tuesday and Thursday. By Tuesday lunchtime the 9 a.m. block has gone to a standup that ran long and the gym has quietly become Wednesday, and the document you made is now a description of a week that is not happening.

That is the normal outcome of weekly planning, not a failure of it. Microsoft's telemetry from Microsoft 365, covering the year to 15 February 2025, puts Tuesday at 23 percent of all meetings, the heaviest day of the week, tapering to 16 percent by Friday. Tuesday is the test.

Sunday is also the most flattering moment of the week in which to make a promise, and that shows up in the data. Hengchen Dai, Katherine Milkman and Jason Riis went through Google search records, university gym turnstile data and goal commitments made on stickK, and reported in Management Science in 2014 that commitments to a goal rise 62.9 percent at the start of a new week, with gym attendance falling steadily across each day that follows it. You write the plan at the top of that spike. It then has to work on the days the same data shows sloping away from it. The rest of that Sunday ritual, the tidying and the shopping list, is worth about 40 minutes once you sort it by evidence; the plan is the part that has to survive contact with Tuesday.

That changes the question. Not how to write a better Sunday plan, but what that plan is still good for on Wednesday, once a third of it has been overtaken. What follows is a five-line structure that assumes the week will go wrong, and the evidence behind each line.

What weekly planning is actually for

A weekly plan has one job that survives contact with the week: it makes Wednesday's decisions cheap. When something gets displaced, you already know what it was competing with, which day it can move to, and what you would drop to make room. An accurate forecast of your own hours is what it looks like it is offering. It cannot deliver that.

There is a second payoff, and it lands the moment you finish writing rather than at the end of the week. E. J. Masicampo and Roy Baumeister put 73 undergraduates through a study at Florida State, published in the Journal of Personality and Social Psychology in 2011. Two groups wrote about a pair of important errands they had not done, and one of those groups then wrote down how, when and where they would do them. Everyone afterwards read an unrelated passage and was interrupted four times to report what they had been thinking about. Among those who had made no plan, 65.2 percent had let their minds wander at least once, against 33.3 percent of the planners, who also answered more of the comprehension questions correctly. Nothing had been completed in between.

So the honest case for fifteen minutes on Sunday is not that you will predict the week, which you will not, but that your head gets quieter for having decided.

The week you plan is not the week you get

Roger Buehler, Dale Griffin and Michael Ross followed 33 psychology students through their honours theses and published what happened in 1994. Asked for their best estimate, the students said 33.9 days. They took 55.5. Only 29.7 percent finished inside their own best guess, and fewer than half of them, 48.7 percent, finished by the date they named when asked to imagine everything going as badly as it possibly could.

Plan around that. Their best estimates came to 61 percent of the time the work actually took, and even the deliberately gloomy figures they produced when told to assume the worst fell short more often than they held. If your own estimates run anywhere near that short, a week booked to the brim is oversold by roughly 40 percent before a single thing has gone wrong. Book about three-fifths of your available hours and leave the rest genuinely unbooked, rather than filled with optional work you would like to reach.

The empty part is not spare capacity. Microsoft reports that people working in Microsoft 365 are interrupted every two minutes across the day by a meeting, an email or a notification, that half of all meetings fall between 9 and 11 in the morning and 1 and 3 in the afternoon, and that 57 percent of meetings now arrive as ad hoc calls with no invitation attached. Those hours are already spoken for. You are choosing whether to find that out on Sunday or on Wednesday.

A weekly planning template in five lines

The whole structure fits on an index card and takes about fifteen minutes, and none of it needs software.

Line two looks like a formality. It has more direct experimental support behind it than anything else in the table.

Peter Gollwitzer and Veronika Brandstätter ran the cleanest version of the test in 1997. Eighty-six German students were asked, a week before Christmas, to write a report on how they spent Christmas Eve and post it back within two days. Half were additionally asked to name when and where they would sit down to write it. Among those who named a time and a place, 71 percent hit the deadline; among those who only held the goal, 32 percent did. Same task, same holiday, one extra sentence.

That result has since been pooled with a great many others. Gollwitzer and Paschal Sheeran gathered 94 independent tests of if-then plans and put the average effect on goal attainment at d = .65, which is medium to large by the conventions of that literature.

The field version is far less dramatic, and more useful for it. Katherine Milkman and four colleagues mailed all 3,272 employees of a Midwestern utility firm a reminder about free flu clinics on site, and some of those mailings carried a blank line asking for the date the employee planned to go, while others asked for the date and the time. The control group vaccinated at 33.1 percent. Writing a date raised that by 1.5 points, a difference that was not statistically significant. Writing a date and a time raised it by 4.2 points, which was.

A day alone barely registered. The day plus the hour is what moved it, and that is the difference between a weekly plan and a weekly list.

Plan the thing that will break it

Most weeks have one predictable point of failure, and by Sunday night you usually already know exactly which one it is going to be. The 9 a.m. block a standup eats. The gym session that dies whenever you leave the office at seven. If you keep the week in software rather than on paper, it is worth knowing what your app does once a block overruns, because the answers range from rebuilding your afternoon uninvited to nothing at all.

Naming it in advance, with a specific answer attached, is its own intervention, and it has been tested against the alternative. Falko Sniehotta, Urte Scholz and Ralf Schwarzer randomly assigned 211 people leaving a cardiac rehabilitation programme to three groups: standard care, a group that wrote action plans covering when, where and how they would exercise, and a group that wrote those plus coping plans naming the barriers they expected and what they would do about each one. Two months after discharge, the combined group was doing significantly more exercise than either of the others. Writing the action down helped. Writing down what would get in its way, and the answer to it, helped more.

One line, in practice. If Tuesday's block goes, I draft 500 words at 8 p.m. instead. The fallback has to be visibly smaller than the original, small enough that refusing it would feel silly, which is the same move that gets a stalled task started at all.

The Wednesday check, not the Friday catch-up

Wednesday afternoon, ten minutes, three questions: what is done, what moves, and what gets dropped this week. Dropped on purpose, in writing. A plan you keep pretending is still live is worse than no plan at all, because it turns every glance at it into a small report of failure, and the week is not even over.

The evidence here is unusually strong. Benjamin Harkin and seven co-authors pooled 138 randomised studies covering 19,951 people for Psychological Bulletin in 2016 and found that prompting people to monitor progress toward a goal raised attainment at d+ = 0.40, with larger effects in two specific conditions: when the information was physically recorded, and when the outcome was reported to someone or made public.

David Allen calls the weekly review the "critical success factor" in making his method stick, and his version runs three stages, get clear, get current and get creative. Most weeks need less ceremony. The part worth taking is that the review is scheduled rather than summoned by guilt.

A dropped item is not a failure. It is the same principle as restarting a run of days after you miss one, where the recovery has to be designed in or it does not happen.

What to leave loose

Not everything improves for being given a start time. Gabriela Tonietto and Selin Malkoc ran 13 studies on activities people genuinely put in calendars, coffee with a friend, a film, a break, and reported in the Journal of Marketing Research in 2016 that scheduling a leisure activity made it feel more like work, and measurably reduced both the anticipation of it and the enjoyment when it arrived. The fix in their own data is precise. Rough scheduling, meaning the activity gets a day but no start time, removed the effect completely.

So the week has two zones. Work you are likely to avoid gets a day and an hour. Things you actually want get a day and nothing else, which is the difference between "Saturday afternoon, see Sam" and "2:15, coffee with Sam, 90 minutes".

The same restraint applies inside each day, where three commitments beats a list of twenty-seven for reasons that are about the reporting rather than the work.