A daily expense tracker with CSV import lets you log cash spends by hand and bring in everything else from a bank or card file. For the import to be useful it must let you pick the account, preview rows before saving and flag rows that look like repeats. Prepare the file first, then import in small batches.
Why import a CSV when I could type everything?
Typing is fine for cash, but a month of card and UPI payments can run to well over a hundred rows. A CSV from your bank carries date, narration and amount already, so importing saves time and avoids typos. The trade-off is that bank narrations are messy, so you need a way to clean and categorise them.
What should a CSV import let me do?
- Choose the account (savings, credit card, wallet) before importing, so each row belongs to the right place.
- Preview rows and fix or untick the wrong ones before anything is saved.
- Show an import history with a status, so you know whether a file went through.
- Warn about possible duplicates when the same file or an overlapping period is uploaded twice.
- Let you correct a category once and remember it next time.
Use these as questions for any app you evaluate.
How do I prepare the file before importing?
- Open the CSV and confirm there is one header row and no summary lines at the top or bottom.
- Check that dates are in one format throughout.
- Look at the amount columns: some banks give separate debit and credit columns, others give one signed column. Know which yours is.
- Remove opening and closing balance rows if the file has them.
- Save a backup copy of the untouched original.
Do not edit narrations by hand unless needed; the importing tool should handle those. Keep the file as the bank produced it as far as you can.
A worked example with made-up numbers
Say you download March activity: 42 rows, with debits totalling ₹38,640 and credits of ₹52,000 from salary. After importing, check three numbers: the row count shows 42, the spending total is ₹38,640 and the credit total is ₹52,000. If the app shows 41 rows, one row was dropped or merged, and you should compare against the file before moving on. If you later import April and the bank's file starts on 28 March, three days overlap. Say those three days held 5 rows; a good tool flags them, and you untick them rather than counting them twice.
How should the import behave in practice?
SpendixAI accepts bank or card statements as PDF, CSV or Excel. You pick the account first, review rows before saving and can upload several files in a batch. It keeps an import history with status, sends an Import failed notification when a file does not go through, and flags possible duplicate imports. It does not connect to your bank or read SMS, and it cannot promise that every unusual layout parses perfectly, so the preview step is where you confirm.
How do I keep the daily part going?
Use the import for bank-side spending once a week, and add cash by hand each evening. Then your review on Sunday is just checking uncategorised rows and totals.
What if the numbers do not match after import?
Compare three things in order: the number of rows, the total of debits and the total of credits. If rows match but totals do not, a sign or a column was misread, for example a debit treated as a credit. If rows do not match, look for lines the tool skipped, such as page headers or balance rows. Fix the file or the mapping, delete that import and run it again instead of patching rows one by one.
Checklist before you click save
- Right account selected.
- Row count and totals match the file.
- Overlapping dates ticked off or removed.
- Unknown merchants categorised so the next import is faster.
Useful ideas start with a clearer picture of your own spending.
Explore your reports

