The most reassuring thing about importing in Spendix is that nothing lands in your records immediately. Rows from a statement are staged first. You get a chance to read them, fix them and decide what to include. The quality of that review determines how clean your books stay, so it deserves a method.
Start with the flags
The review screen marks rows that need attention. Two kinds are common: rows that look like a transaction you already have, and rows that look like another row in the same import. These are the likeliest sources of double counting, so look at them first.
For each flagged row, compare it with the existing entry. If it is a true duplicate, exclude it. If it is a legitimate repeat, such as two identical fares on one day, keep it. Filter chips help you jump between flagged and unflagged rows.
Use bulk actions
Reviewing every row individually is rarely necessary. The screen supports bulk changes: setting one category for several rows, marking selected rows as business or personal, or editing all rows at once. Group similar rows, apply the change, and move on.
Some imports are large, and the screen notes that reviewing many rows is faster on a computer. The same review is available on the web, so for a statement with hundreds of lines you may prefer to switch. You can still do the whole job on your phone if you want to.
What to check in the unflagged rows
Flags catch likely problems, not all problems. A quick pass through the rest is still worthwhile.
- **Dates** fall within the statement period.
- **Amounts** look plausible; compare the largest few with the statement.
- **Directions** are right: refunds shown as money in, payments as money out.
- **Categories** make sense for the biggest merchants.
Include only what you intend
Rows can be left out of the commit. If the screen tells you there is nothing to commit, you have excluded everything. This is useful for rows that duplicate existing records or that you simply do not want. Remember that excluded rows do not become transactions, so double-check the list before you confirm.
Compare totals
After editing, compare totals with the statement. A banner may tell you when the imported rows match the statement total. If they do not, look for a missing row, a double-counted row, or a sign error. Resolving the difference at this stage is far easier than after the import is committed.
| Check | Done when |
|---|---|
| Flagged rows | Each one kept or excluded deliberately |
| Largest amounts | Match the statement |
| Dates | All within the period |
| Totals | Agree with the statement or are explained |
Pace yourself on big statements
A statement with several hundred rows does not need a single sitting. Review in passes: first the flagged rows, then the largest amounts, then categories for the busiest merchants. Anything left can be tidied through the review queue after the import. Trying to perfect every row before committing is the main reason people abandon imports halfway.
Leave yourself a trail
After committing, add a short note to anything you resolved by judgement, such as a repeat you kept on purpose. Months later, when a reconciliation or a question about a charge arrives, that one line saves you from redoing the thinking.
Finish with a clean commit
When satisfied, continue to confirm the import. Afterwards, spot-check the transactions list. If you notice a problem, do not panic: a whole batch can be discarded, and individual transactions can be edited or deleted. Careful staging simply keeps such repairs rare.
A disciplined five-minute review per import is the best insurance against the slow accumulation of small errors.
Useful ideas start with a clearer picture of your own spending.
Explore your reports


