Think about how food gets managed in a household. There are two decisions, and anyone who has run a kitchen knows they don't mix.
One is the weekly shop. Once a week you decide what comes in: rice, pasta, the chicken for Wednesday, some fruit, and roughly how much of each. The other is dinner. Every day, around six, you open the fridge and decide what goes on the stove, with what's there and whatever energy you have left.
Skip the shop and you open an empty fridge at eight and eat crackers for dinner. Plan every meal of the week on Saturday, down to when you'll chop the onion, and on Thursday you get home late, order a takeaway, and the fish goes off in the fridge.
Time works the same way. Weekly planning is the shop. Daily planning is dinner. And most plans that fall apart do so because one tried to do the other's job.
The difference in one line
The study where the detailed plan lost
In the early 1980s, Daniel Kirschenbaum and colleagues followed first-year university students enrolled in a study-improvement programme. Everyone got the same training in study and self-control skills. What differed was the plan: one group planned each day, with specific goals and activities; another planned by the month, in broad strokes; a third made no explicit plan at all.
The intuitive bet is that the daily plan wins. It is the most concrete, the most controlled, the one that tells you exactly what to do at every moment. In the best-known account of the 1981 study, that is what the researchers expected, and the monthly group was the one that improved most in study habits and attitudes.
The one-year follow-up, published in 1982 with 48 freshmen, adds nuance. Programme participants in general raised their grades and were less likely to drop out than students who had not taken part. And monthly planning led to the biggest gains over the year, but only among students who already had better grades (a B average). For those who started with low grades, the abstract reports no such advantage. The authors discuss a possible inhibiting effect of highly specific plans and a helpful effect of moderately specific ones.
The most sensible reading I can take from it: the overly detailed plan doesn't fail on paper, it fails on Tuesday. It describes a day that won't happen, and every deviation gets logged as a small defeat. The looser plan gives direction and leaves room for life to get in.
What the week should decide
The week is where you make the decisions that need some distance.
What has a date. Appointments, deadlines, birthdays, the bill due on the 10th. None of this is negotiable; it goes on the calendar straight away.
A few outcomes. Two or three things that, if they happen, make the week worth it. Outcomes, not time slots.
How much fits. This is the part almost nobody does. When Roger Buehler and colleagues asked students to think aloud while estimating how long a task would take, about three quarters of their thoughts were about how the future would go. Only 7% were about how things had gone in the past. When the researchers made people connect past experiences to the new prediction, the optimistic bias disappeared.
It is the shop done after checking the cupboard. Before deciding what fits into next week, look at what fitted into the last one. If you planned eight things and did four, next week is not the exception; it is last week under a different name.
What the day should decide
The day is where you make the decisions that can only be made up close.
The order. Given the energy you have, what landed in the morning, the meeting that moved.
The first step. Not "work on the report" but "open the file and write the heading for section two". The day turns the week's outcome into a concrete start.
What gets dropped. Every day something doesn't fit. Deciding that in the morning is a choice; discovering it at six in the evening is guilt.
How to split the work
| Weekly planning | Daily planning | |
|---|---|---|
| Question it answers | What goes in, and what matters? | In what order, starting where? |
| Level of detail | Outcomes and dates | Sequence and first step |
| When to do it | Once a week, looking back at the last | The evening before or first thing |
| Sign it's going wrong | Time slots booked four days out | More items than today can hold |
How far this goes
What I'm more confident about is the principle that shows up on both sides: a plan that details what can't be known yet tends to get abandoned, and a forecast that ignores the past tends to be too optimistic.
Why I split the two
For a long time I planned my week with the precision of a day. On Sunday night the whole week was drawn up, hour by hour. By Tuesday it was a museum piece, and I was back to deciding everything on the fly.
With ASB I wanted the two conversations to be different. Once a week, the assistant asks on WhatsApp what has a date and what matters, and reminds me what actually fitted into the week before. At the end of each day, it only asks what comes first tomorrow.