The Budget Category No App Offers: "I Don't Know Why I Bought This"

Every budgeting app ships with the same buckets: food, transport, housing, entertainment, shopping. Every one of them assumes that each transaction was a decision, and that the only open question is which drawer it belongs in. This piece argues for adding one bucket none of them offer, labelled something like “unexplained”, for the transactions where you look at the line a week later and cannot reconstruct a decision at all. It covers what belongs in it, why ordinary categories hide exactly these purchases, and how to read the pile at the end. There are no results of mine in here. It is a method, with the published research that explains why it works.

What goes in it, and what does not

It is not the same thing as impulse spending, and calling it that lets you off. An impulse purchase is one you decided to make quickly. You remember wanting it. The unexplained pile is for purchases where the wanting itself has gone missing: a cable at a price you would not normally pay, a charge from a shop you remember visiting but not why, the same kind of pen bought three times at three-week intervals.

Take the pens as a worked example, because the hypothetical is a common one. Say you buy the second because you cannot find the first, and the third because you cannot find either. The cost is not three pens. It is three pens plus the fact that your desk has no fixed place for the things you use daily, and no budget category will ever show you that. If the pens are a legitimate work expense they sit quietly under “supplies”, each one defensible, the pattern invisible.

The working rule for what goes in: if placing a transaction took real effort, it goes in the unexplained pile. Do not be strict about the definition. The effort is the signal.

Why ordinary categories hide these purchases

Two findings from the mental-accounting literature explain the blind spot.

The first is that budgets track what is typical. Heath and Soll’s 1996 studies in the Journal of Consumer Research found that people do set category budgets and track against them, but that the effect is strongest for purchases highly typical of the category. A cinema ticket registers against “entertainment” and blocks the next cinema ticket. An atypical purchase registers weakly or not at all. A replacement charger, a duplicate pen and a gadget bought in an airport are atypical of everything, so they pass through a category system without touching the sides.

The second is that ambiguous expenses get filed wherever they cause least discomfort. Cheema and Soman (2006, Journal of Consumer Psychology) showed that when an expense could plausibly belong to more than one mental account, people assign it flexibly, in the direction that justifies the spending. A dinner becomes “work” or “food” or “a celebration” depending on which account has room. That is the mechanism behind defending purchases you regret: the category is part of the defence.

Put those together and a conventional category system has a predictable failure. The purchases most worth examining are atypical and ambiguous, which is precisely the kind it records least faithfully. An explicit “unexplained” bucket removes the option of filing the awkward ones somewhere comfortable. Survey evidence suggests the amount hiding is not small: in C+R Research’s 2022 survey of 1,000 US consumers, people estimated their monthly subscription spending at $86 and then itemised it at $219. Subscriptions are only one line, but they show how far a felt total can sit from a counted one.

Sort the pile by what went wrong, not what was bought

This is the step that makes the method worth doing. At the end of a quarter, take the unexplained pile and ignore what the items were. Sort by the failure that produced them. In practice most entries fall into one of four groups.

Failure type	What it looks like on the statement	What it points at	The fix is usually
Replacement because lost	Duplicates of cheap durable things: chargers, pens, tape measures, umbrellas	No fixed place for frequently used items	Organisational, not financial: one drawer, one hook
Out of routine	Clusters around travel, moves, illness, deadlines	Decisions made tired, rushed or without your usual options	Often none worth building; budget it as a known cost of being away
Maintenance avoidance	A new item bought while a working-but-neglected one sits at home	Fixing or cleaning feels harder than buying	A short repair list with dates; the purchase shows up, the unfixed thing never does
Genuinely mysterious	No memory, no pattern	Usually nothing	Leave it; a small residue is normal

The table is this site’s characterisation, not a published taxonomy. Your pile may produce a group that is not on it, and that is the point: the groups come from your statement, not from a template.

Three things are worth noticing about the fixes column. Most of them cost nothing. Almost none of them are budgeting actions. And the third row is the one to flag to anyone else, because it is the most invisible: buying a new thing is an event and appears on a statement, while failing to fix the old thing is not an event and appears nowhere. You end up owning both.

How to run it

1. Keep whatever categories you already use. If you are starting from nothing, a four-category worksheet is one option, and this method needs only one addition to them.
2. Add “unexplained”. Apply the effort rule above. Expect it to feel slightly embarrassing; that feeling is the ambiguity finding at work.
3. Review weekly, not monthly. Memory for small purchases decays quickly, and a transaction you could have explained on Sunday is a false positive by the end of the month. The same applies if you are logging every purchase for a month; the two methods share a log.
4. After about three months, sort by failure type and write one structural change per group. Not a spending target. A change to the environment that produced the purchases.
5. Judge it by the pile, not the total. If the replacement-because-lost group stops growing after you fix the drawer, the method worked, whatever total spending did.

What not to expect

This is a diagnostic, not a savings plan. Nothing here predicts that total spending falls during the months you run it, and anyone promising a percentage is guessing. What it produces is a short list of specific, mostly free, mostly non-financial changes, which is a different output from the “spend less on coffee” conclusion that category totals tend to push people toward. Small purchases are where this pile lives, which is also why a list of everything bought under fifteen dollars is a useful companion view of the same statement.

Sources checked September 18, 2026: the Heath & Soll and Cheema & Soman abstracts (publisher pages behind the DOIs linked above), and C+R Research’s published survey summary. This article is general information about tracking your own spending, not financial advice.