How to categorise a shopping trip that was half food and half everything else
One till receipt with vegetables, dish soap and a birthday card on it. Splitting it three ways is accurate and unsustainable; booking it all as groceries is fast and quietly wrong. Here is how to decide which error you can live with.
You are standing outside the supermarket with one receipt covering forty-one pounds of vegetables, chicken, dish soap, a pack of socks and a birthday card for someone at work. Your ledger wants a category. The receipt does not have one. This is the single most common reason a recording habit stalls in week two, and it is not because the person is disorganised — it is because the question genuinely has no obviously correct answer, and having to solve it fresh at every till is exhausting.
There are only two real options, and both of them are wrong in a different direction. It is worth knowing exactly how each one is wrong before you pick, because then you are choosing an error you understand rather than discovering one six months later.
What each option actually costs you
Booking the whole receipt as groceries is fast, it never stalls, and it inflates your food number by whatever the non-food share was. If that share is small and steady — a bottle of cleaner here, a roll of foil there — the inflation is small and steady too, which means your month-to-month comparisons still work. A food figure that is consistently eight percent high is still perfectly usable for spotting the month that went wrong, because the bias applies to every month equally.
Splitting the receipt is accurate and it is where habits die. Not because splitting is hard, but because it converts one decision into four, at the exact moment you are holding bags and someone is waiting. Do it once and it is fine. Do it at every shop for a month and you will stop recording altogether, and a complete-but-approximate ledger beats a precise one you abandoned in February.
The useful question, then, is not "which is more accurate" but "would the split ever change a decision I make?" If you have never once looked at a household-goods figure and done anything about it, splitting produces a number nobody reads at a cost you pay several times a week. If you are actually trying to find out what the non-food half of the supermarket is costing you, the split earns its keep — but usually only for as long as it takes to find out.
The rule that survives contact with a real shop
A workable policy is a size threshold, decided once, out loud, and then never re-litigated at the till. Something like: anything under a set amount goes in with the groceries; anything above it gets its own line. The number matters far less than the fact that it exists, because its whole job is to remove the decision, not to be optimal.
Almost everyone sets that threshold too low at first. If a five-pound item triggers a split, you are splitting most receipts, and you have effectively chosen the accurate-and-abandoned option without noticing. Set it high enough that the split is rare — the coat, the kettle, the case of wine for a party — and it will still catch the purchases that actually move a monthly total, because those are the only ones large enough to move it.
The other half of the policy is that the whole household uses the same one. Two people applying different thresholds to the same shared book produce a category that means one thing on Tuesday and another on Thursday, and that is worse than either policy applied consistently. This is worth five minutes of conversation once, and it is a conversation about a threshold rather than about anybody’s spending, which is a much easier one to have.
Why one big category is not the failure it looks like
There is a persistent belief that a good ledger is a finely divided one, and that lumping things together is a sign of not taking it seriously. The opposite is closer to the truth. Detail is only worth what it changes, and most category detail changes nothing — it exists because it was easy to create on setup day, not because anyone ever acted on it.
Fambook’s categories are arranged with this in mind: two levels, laid out flat so one tap reaches any of them, ordered by how often a household typically records that kind of thing rather than by how much it costs, and after a few weeks reordered by your own entries rather than the default. That does not make the split decision for you. It does mean that when you do decide to split, the second line costs a tap rather than a search.
If you want the detail without paying for it at the till, there is a third path worth knowing about: record the single line now, and export a month to CSV later if you ever genuinely need to take the supermarket apart. The receipt is gone by then, but the pattern is not — and the pattern is what you were after.
The point of a category is to let you notice something and act on it. If a split would not change what you do next, the split is decoration, and decoration is exactly the kind of small ongoing cost that quietly ends a recording habit.
Frequently asked questions
Should I split every supermarket receipt into food and non-food?
Only if you would act on the difference. A split costs a few seconds every shop, several times a week, and that recurring cost is what ends recording habits. If your non-food share is small and fairly steady, a single grocery line is biased in a consistent direction, which leaves your month-to-month comparisons intact.
Does putting household goods under groceries ruin my food budget?
It inflates it, but predictably. A figure that is consistently high by a similar margin every month still tells you which month went wrong, because the bias applies equally to all of them. What actually ruins a budget is a category whose definition changes from week to week.
What size purchase is worth splitting out?
High enough that splitting is rare. If a small item triggers a split you will be splitting most receipts, which is the version people abandon. Pick a threshold once, apply it without re-deciding at the till, and let the everyday items ride along with the food.
How do we keep two people in a household consistent about this?
Agree the threshold out loud once and write it down somewhere both of you can see. The conversation is about a number, not about anyone’s spending, which makes it a much easier one to have — and inconsistency between two people is worse for the totals than either policy applied uniformly.
Can I go back and split a receipt later?
You can edit an entry afterwards, but the practical answer is usually no, because the itemised detail is on a paper receipt you no longer have. That is the real argument for deciding a policy in advance rather than intending to tidy it up at the weekend.
Try it for one month
Fambook gives a household one shared ledger: anyone can add an entry in seconds, every entry says who spent it, and the month adds up in one place instead of two. Records with no signal and syncs afterwards. Recording, categories, budgets, statistics, CSV import and export, sync and sharing for two people are free — the subscription only buys you less typing.