How many expense categories you actually need

When someone starts tracking expenses, the temptation is to build the full catalogue: groceries, produce, butcher, corner shop, delivery, restaurants, coffee, bars… It looks tidier. In practice, thirty categories don't tell you more than eight: they just mean more decisions per expense and more expenses left uncategorised.

A category costs you twice

The first cost is at logging time. Every expense means choosing between thirty options, and when two of them are close — "groceries" and "produce", "going out" and "restaurants" — the choice turns arbitrary. The same purchase lands in a different category depending on the day, and at month end both columns are wrong.

The second cost is at reading time. A thirty-slice chart can't be read: you see the first four and the rest is a band of colours. The thing you wanted to know — where is my money going — gets buried under the detail.

The rule: a category exists if you'll act on it

This is the only test worth applying. Before creating a category, answer what decision you'll make with that number.

• "Groceries" passes: if it runs away from you, you change how you shop.
• "Produce" split out from groceries almost never does. What would you do differently knowing exactly what you spent on vegetables? If the answer is "nothing", it isn't a category — it's a data point.
• "Subscriptions" passes easily: it's the category where people find things they pay for and don't use.

If a number won't change any decision, it doesn't need its own category.

A starting set of eight

These eight cover one person's spending in a city reasonably well, and each has a decision behind it:

1. Housing — rent or mortgage, service charges, utilities.
2. Groceries and home — everyday food and everything consumed at home.
3. Transport — fuel, transit, taxis, maintenance.
4. Going out — restaurants, bars, delivery, tickets.
5. Health — insurance, pharmacy, appointments.
6. Subscriptions — everything that charges itself.
7. Shopping — clothes, tech, gifts.
8. Other — the catch-all, which has to exist.

If you own a car, "Transport" probably deserves splitting. If you travel often, "Travel" earns its own. A starting set is a starting set.

"Other" has to exist, and it has to stay small

Without a catch-all, every odd expense stalls the logging: you hesitate, you put it off, and you end up not logging it at all. "Other" is what keeps logging from ever getting stuck.

That said, it's also the best signal that your scheme is too thin. If "Other" is 5% of the month, it's doing its job. If it's 25%, there's a category you're missing — and finding it is as simple as opening "Other" and seeing what repeats.

When splitting a category is worth it

When both conditions hold at once:

• It's big enough. If it's 3% of the month, splitting it changes nothing.
• There are two different behaviours inside. "Going out" with the daily coffee and the weekly dinner mixed together tells you nothing; split apart it does, because one is a daily habit and the other is an occasional decision.

If either condition is missing, leave it alone.

Tags, for the things that cut across categories

Some things aren't a category, they're a context: a trip, a renovation, one specific month. That doesn't get solved by creating "Trip to Lisbon" next to "Groceries" — because the trip includes groceries, transport and going out.

That's what tags are for: the expense stays in its category and also gets marked as part of the trip. So you can look at the month by category, and at the whole trip, without one view breaking the other.

In one line

Eight well-chosen categories tell you more than thirty, because you'll actually keep them up. Create a new one only once you know what you'll do with that number, and use tags for anything that cuts across several.

Kesef is free, on iPhone and Android

Shared expenses with no cap, 60+ currencies, and the numbers explained.