# Setting up spending categories that match your life > Most category systems collapse because they have too many categories. Here's how to build a set you'll still be using in a year. Source: https://trybica.com/guides/spending-categories-that-work/ Updated: 2026-08-13 --- The standard failure looks like this: someone sets up thirty-two categories in a burst of enthusiasm, spends two weekends sorting transactions into them, discovers that *Household — cleaning products* has €14 in it, and never opens the app again. Categories aren't free. Each one costs attention every time you read a breakdown. A system with too many of them is harder to read than no system at all. ## The test a category has to pass **Would you do something differently based on this number?** That's it. If the answer is no, the category is decoration. *Groceries* passes: €640 instead of €400 changes what you do next week. *Rent* passes, barely — you can't change it this month, but you need it to know what's fixed. *Bakeries* separate from *Supermarkets* fails: there is no decision that treats them differently. Apply the test honestly and most people land on eight to twelve categories. That's the right order of magnitude. ## A starting set If you're building from scratch, this works for most people: - **Housing** — rent or mortgage, condominium, utilities - **Groceries** — supermarket, market, corner shop - **Eating out** — restaurants, coffee, delivery, lunch at work - **Transport** — fuel, public transport, ride-hailing, parking, insurance - **Subscriptions** — everything recurring and digital - **Health** — pharmacy, appointments, insurance - **Shopping** — clothes, household goods, electronics - **Fun** — cinema, concerts, trips, gifts - **Other** — genuinely everything else Nine. Note that *Eating out* is deliberately separate from *Groceries*: they're both food, but one is a fixed requirement and the other is the single most elastic cost most households have. Merging them hides the only actionable half. ## Let "Other" exist The instinct is to eliminate the catch-all. Don't. Every real month contains a vet visit, a parking fine, a wedding present and a plumber, and forcing each into a named category creates categories that see one transaction a year. Watch its size instead. Under about 10% of spending, it's healthy. Above 20%, it's hiding something — look at what's in it and you'll usually find one recurring thing that deserves a name of its own. ## Split by *behaviour*, not by *thing* The categorisation people reach for is by object: food, clothes, transport. The categorisation that actually drives decisions is by how much control you have: - **Fixed** — same every month, can't change it this month - **Variable** — you influence it week to week - **Savings and debt** — money leaving to your future self You can layer this on top of the nine above, and it's the split that answers the real question, which is not "what did I buy" but "what could I change". [Fixed vs variable spending](/guides/fixed-vs-variable-spending/) covers how to use it. One more reason to keep the health, education and housing categories clean: they are the ones you compare against e-Fatura each February. The [IRS deductions guide](/guides/irs-deductions-from-your-statement/) covers that check. ## Let corrections do the work Building the category set is the manual part. Maintaining it shouldn't be. When transactions are [named properly](/docs/merchant-enrichment/), most of them categorise themselves — a supermarket is groceries wherever you shop. Your job is only the ambiguous ones, and each correction should stick: recategorise a merchant once and Bica applies it from then on. See [categorization](/docs/categorization/). For the cases a merchant name can't settle, use a rule. The two that earn their keep for almost everyone: - **The supermarket that isn't groceries.** Household goods and food from the same store, split on description text. - **The transfer that's actually rent.** A payment to a person's name that only you know is your landlord. ## Rename things to sound like you If you think of it as *Fun money* rather than *Discretionary*, call it *Fun money*. This sounds trivial and isn't: a breakdown written in your own vocabulary gets read, and one written in accountancy language gets skimmed. Some people go further — *Fixed*, *Guilt-free*, *Future me* is a complete and perfectly workable set of three. The categories are for you, not for a filing standard. ## Revisit once, then leave it alone After two months, look at your breakdown and ask two questions: which category never had anything interesting in it, and which number surprised me. Merge the first. Consider splitting the second. Then stop. A category system that changes every month can't show you a trend, and trends are the entire reason for having one. The [weekly review](/guides/weekly-money-review/) is where you use it; it isn't where you keep rebuilding it.