Export / tutorial

Import a bank statement CSV into Google Sheets

Start with the original CSV and import it into a separate sheet. Check that Google Sheets has interpreted the records correctly before combining them with your existing work.

By Fulla LedgerUpdated 4 min read
Statement entries organized into columns

Step by step.

  1. Keep the original file

    Save an unchanged copy of the downloaded CSV. Record its account, period and transaction count so you can compare the imported result with a known starting point.

  2. Import into a separate sheet

    In desktop Google Sheets, choose File and Import, upload the CSV, then select an import location and the matching separator. Use a new sheet or spreadsheet for the first inspection.

  3. Check the interpreted values

    Compare the headers, first and last dates, signed amounts and identifier fields with the CSV. Investigate unexpected types or split descriptions before adding formulas or sharing the sheet.

Separate importing a CSV from extracting a PDF

A CSV already contains a table of records. Google Sheets can import that file; this does not recover transactions from the visual layout of a PDF. If you only have a statement PDF, use the PDF-to-CSV guide first and review the extracted rows before creating an approved output.

If your bank already supplies a usable CSV, you may not need a PDF conversion service. Retain the original statement as evidence where your task requires it, and compare the downloaded records with the intended account and date range. A native export still deserves a coverage check.

Use the separator that matches the file

Google’s import instructions include a separator choice. A comma-delimited record should produce the expected columns, while a semicolon-delimited record needs its own separator. Check a row whose description contains punctuation as well as a simple row; punctuation within a quoted description must stay with that description.

For a fictional record containing the description “Supplies, office,” the comma inside the description should not create an extra amount column. If the import shifts values sideways, stop and inspect the original text. Manually moving a few cells can conceal the same problem elsewhere in the file.

Inspect dates, numbers and identifiers independently

Google states that a spreadsheet’s locale affects default date, currency and number formatting for everyone using that spreadsheet. Set the locale deliberately when starting the working file. Changing how a value looks does not establish that an ambiguous source date was interpreted correctly.

Check at least one date that distinguishes month from day, a negative withdrawal, a decimal amount and an identifier with leading zeros if present. Compare the stored values with the source. Do not convert an account or reference number into an amount simply because it contains only digits.

Build analysis on a checked copy

Keep the imported source tab identifiable, and place calculations in a separate analysis tab or working copy. Count transaction records without including the header or balance summary. Sum the signed amounts, then compare that movement with the source before trusting a chart, filter or pivot table.

When sharing a real financial sheet, use the intended recipients and access level. A spreadsheet created from a CSV is a separate copy of the data with its own sharing settings. Fulla does not automatically synchronize it, and later corrections in Fulla will not silently repair a sheet you imported earlier.

Questions about this article

Can I rehearse the import without uploading a real statement?

Yes. Use the fictional CSV in the sample pack. Its reviewed fixture rows let you compare fields and totals, but the sample is not a benchmark for automatic extraction or a claim of a completed accounting import.

Should I append a corrected file to the previous import?

Usually inspect the correction separately first. Appending the same period again can duplicate records. Establish which rows the new copy replaces before combining anything.

Use what you learned.