The spending spreadsheet, and why it stops

Nearly everyone tries a spreadsheet first, and nearly everyone stops updating it around week five. The reason is not discipline, and it is worth understanding before you try again.

By António Avelar · Updated 18 August 2026 · 3 min read

Almost everyone tries a spreadsheet first, and they are right to. It is free, it is yours, it does exactly what you tell it, and it will still open in thirty years. As a way of thinking about money it is close to perfect.

It also, in the overwhelming majority of cases, stops getting updated somewhere around week five. That is worth explaining properly, because the usual explanation — that you lacked discipline — is both unkind and wrong.

Spreadsheets do not fail at maths

The spreadsheet is not the part that breaks. A sum is a sum. What breaks is the input, and the input is a daily chore with no immediate reward.

Consider what a month actually asks of you. Perhaps 120 transactions across two or three accounts. For each one: find it, read a description like a terminal code and a city, work out what it was, decide which category it belongs to, type it in. Call it twenty seconds each if you are quick and you know all the merchants, which you do not.

That is forty minutes a month of clerical work, and — this is the important part — it has to happen before you get anything out of it. The insight is at the end of the chore. Weeks where you are busy are exactly the weeks with the most transactions, so the backlog grows fastest precisely when you have least appetite for it. Once you are three weeks behind, catching up is a two-hour job, and it never happens.

What the failure is actually made of

Broken down, it is four distinct problems, only one of which is about you:

1. Transcription. Getting the transactions from the bank into the sheet at all.
2. Identification. Turning CT 4471 LISBOA 0821 into a merchant you recognise. This is genuinely hard and no amount of discipline makes it faster.
3. Classification. Deciding which category it belongs to — and, crucially, deciding the same way you decided last month.
4. Interpretation. Looking at the totals and drawing a conclusion.

Only the fourth is thinking. The first three are data entry. Almost all the time goes into them, and none of the value does.

The honest case for keeping the spreadsheet

There are situations where the spreadsheet genuinely wins, and pretending otherwise would be dishonest:

• Anything modelling the future. What happens to the mortgage if rates move two points, whether you can afford a year off. Spreadsheets are unbeatable at questions with assumptions in them.
• Accounts nothing can read. Property, a private loan, cash, an asset held somewhere obscure.
• Not connecting anything. If you would rather no software read your accounts, a spreadsheet fed from CSV exports is the correct answer, and it is a perfectly respectable one.
• Total control of the categories. A sheet does precisely what you say, forever.

Moving without losing the method

If the spreadsheet worked when you kept it up, the thing to keep is the method, not the file. What you built was a set of categories that meant something to you and a habit of looking at them. Both transfer.

What you are handing off is steps one to three: the transactions arrive on their own, the terminal codes get resolved into merchant names, and each one lands in a category. Step four — looking at the totals and deciding what to do — is still yours, and it is the only step that was ever the point.

Two things worth carrying across deliberately. Keep your own category names rather than adopting whatever the tool suggests: categories that survive covers why the set you already trust beats a generic one. And check the first month’s numbers against a month you did by hand — particularly that transfers between your own accounts are not being counted as spending, which is the most common way an automated total quietly disagrees with a manual one. And if the open question is which tool to hand the work to, choosing a personal finance app in Portugal separates the four kinds before you commit to one.