Getting started

Importing a US bank CSV: dates, refunds and column mapping

Import a Wells Fargo checking CSV into an expense tracker without errors: month-first dates, signed amounts, refunds, checks and column mapping.

Editorial illustration for Importing a US bank CSV: dates, refunds and column mapping

A US checking-account CSV usually goes wrong in three ways when it is imported into an expense tracker: dates read the wrong way round, signs reversed so spending looks like income, and columns mapped to the wrong fields. Look at ten rows before importing, test a purchase and a refund, and compare the end balance with your statement. Amounts below are in dollars and made up.

Which dates trip you up?

US files usually write the month first, so 03/04/2026 means 4 March. A tool expecting day-first reads it as 3 April, and a date such as 25/03 simply fails. Find a row with a day above 12 to see which order your file uses. If the tool lets you choose a date format, set it; if not, convert the column in a spreadsheet before importing. we are not describing Wells Fargo's own file layout, so read your headings.

What does column mapping mean?

A tracker needs to know which column is the date, which is the description and which is the amount. If the file has a single amount column, negative numbers usually mean money out; if it has separate debit and credit columns, make sure they are mapped separately. Unneeded columns, such as a running balance, can be left out. Check the mapping against two rows you recognise.

A worked example: signs and refunds

DateDescriptionAmount
03/04/2026Grocery store-$72.30
03/06/2026Employer payroll$1,850.00
03/09/2026Online store-$39.99
03/12/2026Online store refund$39.99
03/15/2026Check 1042-$420.00

The balance change for these rows is −72.30 + 1,850.00 − 39.99 + 39.99 − 420.00 = $1,357.70. Spending is $72.30 + $420.00 = $492.30, because the refund cancels the online purchase. A tracker that reads the refund as income would overstate income by $39.99 and spending by the same amount. The check payment of $420.00 has only a number, so label it from your own records, for example rent or a repair.

  1. Import a handful of rows first if the tool lets you preview.
  2. Make sure purchases show as expenses and the refund reduces spending.
  3. Label checks from your stub or memory the same week.
  4. Compare the end balance with your statement.

Does an India-built tracker suit this?

SpendixAI is built around India and the rupee. It accepts PDF, CSV or Excel, shows rows for review before saving and asks you to pick the account first, but its base currency, once chosen, cannot be changed after transactions exist. we would not promise it fits a US dollar account; a spreadsheet or a tool made for your country is a safer first choice.

How do you separate transfers from spending?

A checking account also records moves to your own savings and payments to a card. Neither is spending, so tag them as transfers. If you import your credit card too, the payment will appear in both places, and one side must be marked as a transfer. A quick test is to ask whether the money has left your whole household. If it has not, it is a transfer, and a tracker that counts it as spending will overstate your costs every month.

What should you keep?

  • The original CSV, by month.
  • A note of the date format and column mapping you used.
  • A list of check numbers and what each was for.
  • A monthly total tied back to the statement.

Once the first month reconciles, the same settings will usually work for the next, so the setup cost is paid once.

SPENDIXAI JOURNAL

Useful ideas start with a clearer picture of your own spending.

Explore your reports