How to use a daily total to set a weekly limit

23 August 2026 · 6 min read

A weekly hours limit only works if it is built from real daily totals, not from the number you wish were true. Measure one ordinary week, pick a ceiling slightly below what you found, split it across the days unevenly to match how your week actually runs, and check the running total each evening against the day’s receipt. On Sunday, compare the week roll with the ceiling and adjust by no more than an hour.

Why limits set from a guess do not hold

“I am going to work forty hours a week” is a decision made without data. If your real weeks are forty-eight, a forty hour limit is not a limit; it is a wish, and it will be broken by Wednesday with no idea why. The limit has to start from the number you are actually at, because the gap between the two is the thing you are trying to close, and you cannot close a gap you have not measured.

The second problem with guessed limits is that they are usually flat: eight hours a day, five days. Real weeks are not flat. Monday has the planning meeting, Thursday has the report, Friday is shorter. A flat limit fails on the heavy day and then gets abandoned for the week.

Step one: measure a baseline week

Pick a week with nothing unusual in it. No travel, no launch, no holiday.
Let the days record themselves. Punchcard prints a receipt at the closing time you set, with each app on its own line, the time it had, and a total at the bottom. You do not start or stop anything; it notices which app is in front during the day (app names only) and itemizes it at closing time.
Each evening, write the day total on one line of a note. Five lines by Friday, seven if you worked the weekend.
On Sunday, Punchcard prints a week roll without being asked. Its total is your baseline.

If you prefer, how many hours do you really work: a one-week measurement walks through this week in more detail. Either way, do not set the limit until you have this number.

Step two: pick the ceiling, then budget the days

Take the baseline and subtract a modest amount. If the baseline was forty-seven hours, a reasonable first ceiling is forty-four. Not forty; that is a twenty percent cut in one step and it will fail. The goal is a number you will actually hit in week one, because a limit that holds once is easier to lower than a limit that has never held.

Write the ceiling down next to the baseline. Both numbers, in the same note. You are going to look at them every Sunday.

Now split the ceiling across the week to match your real shape, not an ideal one. Look at the five daily totals from the baseline week:

The heaviest day gets the largest budget, slightly below what it measured.
The lightest day gets the smallest, and it is fine if that is five hours.
The rest land in between.
The five (or six, or seven) budgets add up to the ceiling.

An example, with a forty-four hour ceiling: Monday 9:30, Tuesday 9:30, Wednesday 9:00, Thursday 9:00, Friday 7:00. This is not a recommendation; it is a shape. Yours should come from your receipts. Using your receipts to plan a lighter week has more on reading the shape.

Write the daily budgets in the note, one per line, with a blank beside each for the actual.

Step three: check the total each evening

At closing time, read the receipt’s total.
Write it in the blank next to that day’s budget.
Add up the actuals so far and subtract from the ceiling. That is what is left for the rest of the week.
If you are ahead (used less than budgeted), do nothing. Do not “spend” it tomorrow; it is margin.
If you are behind (used more), take it from the remaining days’ budgets, starting with the lightest one.

This takes under a minute and it is the part that makes the limit real. The daily total is printed in the same place every evening, so the comparison is mechanical rather than a feeling.

Punchcard does not do this arithmetic for you. It has no limits feature, no budgets, no warning when you cross a line, and no alerts of any kind. It prints the receipt; the note and the subtraction are yours. If you want an interruption at a certain hour, a calendar alarm is the tool for that. The receipt’s job is to make the number honest.

Step four: the Sunday check

On Sunday the week roll prints. Read it next to the note:

Did the week land under the ceiling? If yes, hold the ceiling for one more week, then consider lowering it by an hour. Not more.
If no, by how much? An hour over is noise; keep the ceiling. Three hours over means one of the daily budgets is wrong. Find the day that blew up and give it more next week, taking from another day.
Which line grew? If the roll shows the overage sitting in mail and chat, the limit problem is an interruption problem, and the fix is on those apps rather than on the clock.

The Sunday roll: a weekly review you will not skip is the longer version of this five minutes.

What the receipt cannot tell you about the limit

Punchcard measures time with an app in front on your Mac. That is the right measure for most desk work and the wrong one for some jobs. If a third of your week is on the phone, in a meeting room, or on paper, the receipts understate the week and the ceiling has to account for that separately. Add a line to the note for off-Mac hours and estimate it honestly.

It does not know whether the time was work. A browser line is a browser line. If you use the same apps for work and for everything else, the total includes both, and a limit built on it is a limit on screen time rather than on work. The usual fix is one browser for work and another for the rest.

And it has no per-project or per-client view. A weekly limit on total hours is something it supports well; a limit on hours for a particular client is not, and a timer-based tracker with tags is the better tool for that.

Questions

Should the limit include the weekend? Count whatever you actually worked. If Saturday has a receipt, it is in the week. The limit is a ceiling on the week, and a week has seven days. Many people find that simply seeing a Saturday receipt in the roll is enough to stop producing one.

How fast can I lower the ceiling? An hour a week, at most, and only after a week that held. A ceiling that drops fast gets abandoned. A ceiling that drops by an hour every other week is ten hours lower in five months, and it will still be standing.

What about weeks that are genuinely heavier? Decide in advance. On Monday morning, if you know the week is a launch week, write a different ceiling for that week in the note and return to the normal one afterward. A ceiling changed on Monday is a plan; one changed on Thursday evening is a collapse.

Does Punchcard keep the history so I can see the ceilings over months? It keeps every day in one SQLite file on your Mac and prints month rolls as well as week rolls, and you can export CSV at any time. The ceilings themselves live in your note, not in Punchcard; it has no field for them.