A first import of a year of bank history can leave hundreds of transactions without a category. Faced with a long list, the instinct is to start at the top and grind downward. That is the slowest route. A smarter plan treats the backlog as a pattern problem: most of the entries come from a small number of merchants, and fixing the merchants fixes the entries.
Count and filter first
Begin by narrowing the transaction list to uncategorised entries only. The filters include an option for exactly that. Note the count. Then sort your attention rather than the list: which merchants appear most? Which amounts are largest? Those two questions decide your order of work.
Scroll for a minute without acting. The shape of the backlog usually becomes clear. A few merchants repeat dozens of times, a handful of big payments stand out, and a long tail of one-off items sits at the bottom.
Attack the repeaters
Choose the most frequent merchant and give it a default category on the merchants screen, or write a rule for it. One action now handles every past and future payment to that merchant. Repeat with the next most frequent. After ten or fifteen of these, a large share of the backlog is gone, and no transaction needed individual attention.
- Set merchant defaults for the top repeaters.
- Write rules where a merchant needs conditions, such as amount ranges.
- Use bulk selection for groups that rules cannot capture.
Then the heavy hitters
With repeaters handled, sort by amount and review the largest remaining items. Errors here have the biggest effect on reports. Take care over each: open the transaction, read the merchant and note, and choose the category that fits.
Finally, the long tail
What remains is a scattering of one-off purchases. Work through them in short sessions, or decide that small items do not need precise treatment and file them in a sensible general category. A rule of thumb: if the payment is a very small share of your monthly spending, accuracy matters less than finishing.
Keep the backlog from returning
Once cleared, protect the result. Check the review queue weekly, keep merchant defaults updated, and add a rule whenever you notice a repeated correction. A clear backlog is a one-time achievement only if the system keeps up with new transactions.
| Order | What to tackle | Why |
|---|---|---|
| 1 | Frequent merchants | One fix clears many entries |
| 2 | Largest amounts | Biggest effect on reports |
| 3 | Selectable groups | Bulk action saves time |
| 4 | One-off small items | Low impact, finish last |
A realistic first week
Spread the work over several days. Day one, set defaults for the top five merchants. Day two, write rules for the next five and clear the heavy hitters. Day three, use bulk selection for the groups that remain. By the weekend, a backlog that looked like a month of toil has become a few short sittings, and the habit of tidying is already in place.
Measure progress
Watch the uncategorised count fall as you work. It is motivating, and it tells you when to stop. There is a point where the remaining entries are not worth the effort, and recognising it is part of good housekeeping. Aim for a state where the count is small and steady, not zero at all costs.
A tidy, mostly categorised history makes every later feature more useful: budgets reflect real spending, reports show real patterns, and insights have something sensible to say.
Useful ideas start with a clearer picture of your own spending.
Explore your reports

