How to categorise transactions so the categories actually help
Categorisation is the part of budgeting people spend the most time on and get the least value from. The reason is simple: most category sets are built to describe spending accurately rather than to change it.
Accurate is not the same as useful
A forty-category breakdown is a beautiful description of your past and a terrible input to a decision. When every category holds one or two percent of your spending, no single category is ever worth acting on.
Useful categories are large enough that a ten percent change matters and few enough that you can hold them all in your head.
Design categories around decisions
The test for a good category is whether you can imagine a specific action that reduces it. 'Dining and bars' passes — you can cook more or go out less. 'Miscellaneous' fails. 'General merchandise' usually fails too, because it merges impulse purchases with genuine needs.
Where a category fails the test, split it once along the line that separates the decision, then stop.
- Fixed obligations you cannot change this month
- Fixed obligations you could renegotiate or cancel
- Variable essentials like groceries and transport
- Discretionary spending you would defend
- Discretionary spending you would not
Handle the recurring charges separately
Subscriptions deserve their own view rather than being scattered across categories. A streaming service in entertainment, a cloud drive in software and a gym in health are the same kind of commitment even though they land in different buckets.
Reviewing them as one list once a month is far more effective than seeing each one buried inside a category total.
Let the merchant do the work
Manual recategorisation is where budgeting habits go to die. Set rules at the merchant level so a decision made once applies forever, and accept that a small percentage of transactions will always land in a catch-all bucket.
Chasing perfect categorisation costs more attention than the accuracy is worth. Ninety percent correct and reviewed monthly beats ninety-nine percent correct and abandoned.
Key takeaways
- Categories should map to decisions, not to accounting precision
- Six to nine categories is the practical ceiling for most people
- Review subscriptions as one list, separate from category totals
- Merchant-level rules beat manual cleanup every time
Let Ledger run this for you
Link an account and Ledger sorts six months of history, surfaces recurring charges and flags the ones that have gone dormant.
See pricing