Ask someone what their rent or mortgage costs and they will tell you to the last unit. Ask what they spend on food, transport or the small stuff and the answer is a guess, and the guess is almost always low. Fixed costs announce themselves once a month. Variable costs arrive in forty small pieces, and forty small pieces do not add up in anybody's head.
That is what a spending log is for. Not virtue, not restraint. You write down what leaves your hands for thirty days so that you are arguing with a record instead of a memory. The single most important rule is the one people skip: for the month of the exercise, you change nothing. No cutting back, no proving a point. This is a diagnostic, and a diagnostic you interfere with tells you about the interference.
Pick a method you will still be using on day nine
Any method works for four days. The one you want is the one that survives a Tuesday when you are tired, in a hurry, and standing in the rain. There is a real trade-off here, and it is not the one people assume: the effortless methods capture more and teach less, because the whole benefit of writing it down by hand is the half-second of attention at the counter.
| Pocket notebook | Most attention per entry. Needs to be on you, and it can be lost. |
|---|---|
| Notes app on your phone | Always with you, fast, and nobody wonders what you are doing. |
| Card and account exports | Complete and effortless, but blind to cash and silent about why. |
| A tracking app | Sorts things for you, and you will spend the month correcting how it sorted them. |
Whichever you choose, fix the moment of recording. Either you write it at the counter, or you sit down at the same time every evening and do the day. Floating between the two is how logs die. And decide now what happens with cash, because cash is what breaks almost every log: either you stop using it for the month, or you note the withdrawal and then account for it as it goes, never both halfheartedly.
The field everyone leaves out
Date, amount, what it was. Those three are obvious and they are not enough. Add a fourth: why, in three words. A month later, a line reading lunch out is indistinguishable from every other line reading lunch out. A line reading lunch out, no food at home tells you something different from lunch out, colleague leaving. One of those is a shopping problem. The other is a life, and you are not going to budget your way out of a colleague leaving.
- Date, and mark the ones that were not ordinary days
- Amount as it actually left, including any delivery or service charge
- What it was, in plain words, not a category name
- Why, in three words, written before you have had time to be clever about it
- How it was paid, if you use more than one card or account
Week two, when it falls over
Somewhere around day nine you will miss three days. Then you have a backlog, and a backlog feels like a debt, and the tidy solution is to start the whole thing over on the first of next month. Do not. Draw a line, write missed, and carry on from today.
Reconstructing four days from memory produces a log that looks complete and is quietly false, which is worse than one with visible holes. Holes you can see are data too. It is usually the busy, expensive, disorganized days that go unrecorded, and knowing that those days exist is half of what the exercise was going to tell you anyway.
The other thing to expect is that you will spend less this month than you normally do, purely because you are watching. That effect is real and it is not cheating. Just do not treat the month as a typical one when you set future numbers from it. If anything, nudge the results upward a little.
Reading it at the end
The wrong move on day thirty is to add everything up, feel bad, and close the notebook. A total is the least informative number in the whole document. Read it in three passes instead, and give each pass a few minutes on its own.
- Split committed from chosen. Anything you could not have avoided this month without changing where you live or how you get to work goes in one pile; everything else in the other. Only the second pile is negotiable, and it is usually smaller than you feared and different from what you expected.
- Count frequency, not just size. Sort the chosen pile by how many times each thing appears rather than by amount. The item you bought eleven times is a habit with a price attached; the one you bought once for a lot was a decision. Habits and decisions need completely different treatment.
- Mark the lines you cannot remember at all. A purchase you have no memory of two weeks later bought you nothing lasting. That list, not the total, is the actual output of the month.
Three things usually fall out of that. A number you were badly wrong about, and it is rarely the one you were worried about going in. A recurring charge you had forgotten you were paying. And at least one thing where the amount is perfectly reasonable but the frequency is not, which is a much easier problem to fix than it feels, because you are not giving anything up, only doing it less often.
After the thirty days
Very few people keep a full log running for years, and there is no need to. Once you know which two or three numbers you are systematically wrong about, watch only those, weekly rather than daily. Then run the full exercise again in six months or after anything that changes the shape of the household, because a new job, a new home or a new person in it makes every previous number obsolete.
One more thing worth saying out loud, because it is the reason people abandon the exercise on day twelve. The log is a mirror, not a verdict. It records what happened without any opinion about whether you should have done it, and reading it as an accusation is the surest way to stop keeping one.