How to audit a week
Start in Budget mode and write down where you think the week goes. Sleep first, because it is the largest single block and the one people quietly borrow from. Then paid work, the commute, the chores. Lock those with the padlock: they are committed time, the part of the week you do not really choose. Whatever the counter shows after that is your actual discretionary budget, and it is usually both smaller and larger than expected. Smaller because the committed blocks are bigger than they feel, larger because almost nobody guesses they have thirty odd hours a week of their own.
The 168 hour frame is Laura Vanderkam's, from her book of that name. Her argument is that a week, not a day, is the right unit for this. A day is too short to hold a life: any single Tuesday can be a write off, and comparing Tuesdays makes you feel like a failure. A week is long enough that the exercise, the reading and the two evenings with friends all fit somewhere, and short enough that you can still remember it.
Then log a real day
Switch to Log a day and fill in the half hours of the day you just finished. This is the part that changes minds. Researchers who collect time diaries have found for decades that estimates and diaries disagree, and they disagree in a flattering direction: people who claim the longest working weeks overstate them by the widest margin, and the hours that go missing from the estimate usually reappear as television, phone or unaccounted time. The tool cannot tell you which of your numbers is wrong, but it can put them side by side.
Log yesterday rather than today, in one sitting, and guess when you have to. A diary that is roughly right beats an estimate that is confidently wrong. Two weeks of honest logging will teach you more about your schedule than a year of feelings about it, and the whole exercise costs about three minutes a day.
What to do with the gaps
Read the comparison as information rather than a verdict. A gap between budget and log is not a moral failure, it is a mislabelled hour. If scrolling comes in at three times its budget, the useful question is which category it is stealing from, because that is the thing you said you wanted. Move the budget to match reality where reality is fine, and move reality to match the budget where it is not.
Then change one thing. Auditing a week and changing nothing is a hobby. Pick the single largest gap, adjust one recurring block, and log another week to see whether it held. The time blocking planner is where that change becomes a plan.