Budgets tend to gain categories over time and rarely lose them. Past a certain point the effort of maintaining the structure exceeds anything it tells the person maintaining it.
Categories are added to explain surprises
When spending in a broad category comes in higher than expected, the natural response is to split it and find out where the money went. Each surprise adds a subdivision.
The split does answer the immediate question, so the change feels productive. Nothing prompts anyone to remove it once the question has been settled.
Repeated over a year or two, this produces a structure with many categories, most of which are never consulted again.
Every category imposes a classification cost
Each transaction must be assigned somewhere, and more categories mean more decisions and more ambiguous cases. A supermarket trip containing household goods and food fits two places at once.
Ambiguity is what makes maintenance tiring. The work is no longer recording spending but adjudicating boundaries that carry no consequence.
Automated categorisation helps but introduces its own errors, and correcting misclassifications becomes an ongoing task of its own.
Detail without a decision attached is noise
A category earns its place only if a different number in it would change something. If no plausible figure would prompt an action, tracking it separately adds nothing.
Much fine-grained detail fails this test. Knowing the split between two kinds of household consumable rarely alters behaviour in either.
The useful question for each category is what would be done differently at twice the amount. Categories with no answer can be merged.
Abandonment follows friction, not failure
People stop budgeting when the process becomes tedious rather than when it stops working. A system that takes an hour a week is abandoned regardless of accuracy.
Once a few weeks are missed, the backlog itself becomes a barrier, and restarting requires reconstructing transactions nobody remembers.
Simplicity is therefore a durability property. A coarse budget that is still being maintained outperforms a precise one that stopped.
What a small structure still captures
A handful of categories separating committed costs, essentials, discretionary spending and saving is usually enough to answer the questions a household actually acts on.
Temporary detail can be added deliberately when a specific question arises, then removed once answered, which keeps investigation separate from routine tracking.
The aim is a structure that survives a busy month, since a budget only informs decisions during the periods when it is still being kept.