Import first, then adjust. Start with the default categories, upload a real statement, and see which merchants actually appear before you rewrite the list. Designing a perfect category tree in advance tends to produce categories you never use and misses the ones you need. A short amount of setup beforehand, plus a focused tidy-up after the first review, works better than either extreme.
The two orders, compared
| Point of comparison | Categories first | Import first, then adjust |
|---|---|---|
| Effort before you see any data | High, you must guess your own spending | Low, defaults are ready |
| Risk | Categories you never use; missing ones you need | First review has more rows to correct |
| Fit with your real life | Based on imagination | Based on your merchants |
| Good for | People with an accountant's list to follow | Most personal users |
There is one situation where categories first wins: your CA or business already has a chart you must match. In that case, set up those categories and their codes before importing, so rows land in the right place from the start.
Why starting from defaults usually works
A statement is a list of what you really did. Once you see forty rows of food delivery, you know you need a separate Delivery category. If you see one, you do not. SpendixAI starts you with default categories on every plan, including Free, and categorises in stages: your own rules first, then your history and similar merchants, then, only if AI is switched on, an AI model for what is left. On a first import you have no rules and no history, so more rows will need your attention. That is expected, not a failure.
Choose the account before importing, since each transaction keeps its account. If your base currency matters to you, choose that first as well, because it cannot be changed once transactions exist.
The middle path
- Before import: pick the account, check the base currency, and glance at the default categories. Make only the changes you are already sure of, such as adding a category for rent if there is none.
- Import one month, not the whole year. A single statement is enough to see your patterns.
- Review the rows. Fix categories on the merchants that matter, and add rules for the ones that repeat.
- Afterwards, make your category changes in one sitting, using what you saw.
- Import the older statements. They benefit from the rules and history you just built.
What should I change after the first review?
Keep the list short. These are the changes that usually pay off, none of them required.
- Split a category that swallowed too much, for instance Shopping that now holds clothes, gadgets and gifts.
- Merge categories that mean the same thing so reports stop dividing the same spend.
- Add a category for a large recurring cost that deserves its own line, such as society maintenance or school fees.
- Rename anything you cannot explain to a friend in three words.
- Add GST rates, the tax flags or accounting codes only if you actually need them; ask your CA if unsure.
What if I have already imported everything?
Nothing is lost. Corrections you make become history, and categories can be adjusted afterwards. Go through the biggest merchants first, since they account for most of the spend, and ignore the long tail until later. A list built from real data will fit better than the one you would have written in advance, even if getting there took a second pass.
One more reason to wait: your first month will show you things you did not expect, such as a subscription you forgot or a delivery habit larger than you thought. Those discoveries are the real reason to categorise, and they are easier to act on when the list reflects what you actually spend, not what you planned to spend.
Useful ideas start with a clearer picture of your own spending.
Explore your reports

