How to Do a Time Audit in One Working Week

A time audit is a small, boring exercise people avoid because it looks like extra homework: write down what you actually did with your working hours. Peter Drucker argued in The Effective Executive that knowledge workers cannot manage their time until they have measured where it goes, and that the first measurement always contradicts what they believed. Here is how to run one in a single working week without it becoming a second job.

Measure before you optimise

The case for auditing is that people are poor estimators of their own day. Observational studies of office work that watched people directly found task switches happening every few minutes, while the same people, asked afterwards, described long uninterrupted stretches. Memory keeps the shape of a day and drops the texture, and the texture is where the fixable problem usually sits.

Five working days is the minimum, and ten is better if your weeks vary. One day tells you about that day. Five days with different meeting loads tell you something about the job itself.

Log 30-minute blocks in the moment

Use a spreadsheet with one row per half hour, or a notes file if you will actually keep it open. A phone alarm every 30 minutes is enough of a nudge. Fill each row in as the block ends, not at 6pm, because the fragments are the finding. Reconstructing the day later smooths out exactly the interruptions you are trying to count.

Capture six fields per block and no more. More fields means you stop after two days.

• What you did - A five-word label, for example 'rewrote onboarding email' or 'weekly planning call'.
• Category - One of five, and only one. Splitting a block across categories lets you hide the time you are trying to see.
• Who was with you - Alone, one other person, or a group. Group time has a different cost structure from solo time.
• Energy 1 to 5 - Your level at the end of the block. Patterns here show which work drains and which restores.
• Did it move a committed outcome - Yes or no against something you had already promised. This is the field people find hardest to fill honestly.
• Interruptions - A tally of times you were pulled off the block by someone else, and a separate tally of times you pulled yourself off.

Sort the week into five categories

Deep work is effort that needed your full attention and produced something. Communication is email, chat and messages. Meetings are synchronous calls and rooms, including internal ones. Admin is scheduling, expenses, forms and travel. Unplanned work is everything that arrived without being on the plan. Sort the whole week, then total the hours in each column and convert to percentages of your contracted hours.

Expect the sorting itself to take 45 minutes. It is the cheapest part of the exercise and it is usually where the surprise lands, because unplanned work rarely shows up as a category people predict in advance.

Read three numbers

Three figures are worth extracting from the log. The first is your average daily hours of deep work. The second is your average daily hours in meetings. The third is your self-interruption count, which is the number of times you left a block by choice, not because someone walked over.

A day with under two hours of depth and over three hours of meetings is usually a structural problem, not a discipline problem, and no willpower technique will fix it. A day with a healthy deep block but a self-interruption count near 30 points at habit, and that one you can change alone. The audit tells you which problem you have, and those two problems need different responses.

Change one thing, then re-check in six weeks

Pick the single largest block of low-value time in the log. That might be a recurring meeting that takes 90 minutes and produces nothing you can name, or an hour of evening email that only exists because you did not batch it earlier. Change that one thing and leave the rest of the week alone, so that in six weeks you can see whether the change did anything.

Re-audit for three days at the six-week mark, not five. You are checking whether the change held, not rebuilding the full picture. If the log shows the same pattern, the change was wrong for that constraint rather than the general idea being wrong.

• Move the standup - If your best focus window is 9am to 11am and the daily call sits at 9.15, move the call before you try anything harder.
• Batch the inbox - Three fixed checkpoints a day instead of continuous checking. The audit shows how many hours the continuous version costs.
• Cap the meetings - Set a weekly ceiling in hours and decline anything that pushes past it. A ceiling without a decline is only a wish.
• Protect the first block - If morning depth already exists, defend the start time and stop letting it drift by ten minutes a week.