How to do an honest time audit without a spreadsheet
· 6 min read
A time audit is a week of recording where your hours go, followed by one honest look at the totals. You do not need a spreadsheet. An index card, the Screen Time report built into your Mac, or a printed end-of-day receipt all produce the same thing: a list of categories with hours next to them. The honesty comes from not editing the record while you make it, and the spreadsheet is usually where the editing creeps in.
What a time audit is for
The purpose of an audit is to replace a belief with a measurement. You believe the day is mostly spent on your main work; the audit tells you what it is actually spent on. You believe meetings take a couple of hours a week; the audit tells you the number. The gap between belief and measurement is the whole product. Everything you do afterward comes from that gap.
This means an audit has to be done on an ordinary week, without changing behavior, and without tidying the numbers to look better. An audit of a good week is not an audit; it is a performance.
It also means the audit only needs to run once, or once in a while. It is a diagnostic, not a habit. People who try to make it permanent usually abandon it and feel they have failed at time tracking, when in fact they finished the audit and did not need the rest.
Why spreadsheets kill audits
Spreadsheets seem like the natural tool, and they are the reason most audits die by Wednesday.
The first problem is friction. Opening a sheet, finding the row, typing a start time and an end time and a category: it is enough work that you defer it, and deferred entries get reconstructed from memory, and memory is what you were trying to get away from.
The second problem is categories. A spreadsheet invites you to design a taxonomy before you have any data: Deep Work, Admin, Meetings, Communication. Then the real day arrives and nothing fits. Was that call admin or communication? The time spent deciding goes unrecorded, and the categories quietly become a story about the week you wish you had.
The third problem is the edit. Once the numbers are in cells, they can be adjusted. A fifteen-minute browse becomes ten. A forgotten hour is filled in as “research”. Nobody does this dishonestly; it just happens, because a spreadsheet looks like a document you are responsible for rather than a record of what occurred.
A good audit instrument is one you cannot easily edit and barely have to touch.
The index card method
This is the manual version. It works if you can keep it up for five days, and it captures things no app can see, like phone calls and whiteboards.
- Take one index card per day. Write the date at the top and your planned closing time.
- Set an alarm to go off every hour during the working day.
- When it goes off, write the time and one short line describing the hour just gone. Use the words you would use to a colleague, not a category: “drafting the proposal”, “mail and chat”, “call with the printer”.
- If the hour was split, write both, with a rough fraction: “mail, then mostly the proposal”.
- Do not go back and revise earlier lines. If you realize an entry was wrong, add a new line saying so.
- At the closing time, put the card in a pile face down and do not look at it again until the weekend.
- On the weekend, lay the five cards out and, with a pen, circle recurring words. Count the hours next to each circled word.
The face-down rule is important. Reading the card each evening tempts you to do better tomorrow, which ruins the sample. The audit week is for observation only.
The automatic method
If the weakness of the card is that you have to remember it, the fix is a record that makes itself. Punchcard is a Mac menu bar app that notices which application is in front during the day and, at the closing time you set, prints a receipt: one line per app with its time, a day total, a stamp. You tear it off and put it on the pile. Nothing to fill in, nothing to remember, and because it is a printed record rather than a cell you can type in, it is not something you edit.
It makes a good audit instrument for a particular reason: it does not know your categories and does not ask for any. The lines are app names, which are the one taxonomy you cannot argue with. If mail was in front for two hours, the receipt says two hours, and there is no cell to change it in. It records app names only, never window titles or document names, and asks for no macOS permissions. The first seven receipts are free, which is precisely an audit week, and tracking itself never expires. It cannot see phone calls, whiteboards, or which project the hours belonged to, and if all your work happens inside one app the receipt will show one line and tell you less. In that case, combine it with the card: the receipt for the honest totals, the card for what the totals were made of.
The Screen Time report in System Settings is a third option, free and already installed, with the limitation that its totals run around the clock and include weekends, so you will need to read it at the same time each day and subtract.
Reading the audit honestly
Lay out the five days and resist the urge to explain anything for the first few minutes. Just look.
Find the biggest category. Not the one you expected to be biggest; the one that is. In most first audits it is communication of some kind, and it is bigger than the main work.
Find the total. Add up the recorded hours per day and compare with the hours you would have claimed. The difference is the unrecorded day: lunch, walking, thinking, the phone. It is normal for this to be an hour or two, and knowing its size is useful on its own.
Find the variation. Which day had the longest main-work line? Look at the calendar for that day. Usually it was the day with no morning meeting, which is a finding you can act on.
Then, and only then, explain. Write one sentence per surprise. “Mail is double what I thought.” “Tuesday was the only real work day.” Three or four sentences is a complete audit report. There is more on why the feeling and the number differ so much in why-your-estimate-of-your-own-workday-is-wrong-by-hours.html.
Turning it into one change
An audit that produces a list of ten resolutions produces nothing. Pick the single biggest gap between what you wanted and what the cards say, and make one change aimed at it. Two fixed mail times instead of all day. A protected first ninety minutes. A real closing time.
Then leave it alone for a month and run another audit week. If the category moved, the change worked. If it did not, the change was aimed at the wrong thing, and the second audit will show where the time went instead. Either way you have two measurements to compare, which beats a feeling about whether you have been productive lately. If you want a standing record between audits, the-end-of-day-receipt-a-two-minute-review-that-sticks.html describes a review that takes about two minutes.
Questions
How long should the audit run?
Five working days. Shorter and one unusual day skews everything; longer and the record-keeping itself starts to change the behavior it is measuring.
What if I forget to write on the card for a few hours?
Write a line that says “forgot, roughly 11 to 2, probably mail and the report”. Do not reconstruct it in detail. A visible gap is honest; a filled-in guess is not.
Should I audit evenings and weekends too?
Only if the question you are trying to answer is about them. For a working-hours audit, stop at the closing time every day and leave the rest unrecorded. Mixing them in makes the weekday totals harder to read.