What CSV is, and why convert a bank statement to it
CSV stands for comma-separated values — the simplest, most portable way to store a table. Each line is one record and each value is separated by a comma. A bank statement converted to CSV becomes a lightweight text file where every transaction is one row and every field — date, description, amount, balance — sits in its own column.
Your bank hands you a PDF, which is built for printing, not for data. Pulling transactions out of a PDF by hand means retyping or copy-pasting, and the structure collapses the moment you do. CSV is the opposite: it is nothing but structure. That is exactly why almost every accounting, bookkeeping, and finance tool accepts CSV as an import format.
Converting a bank statement to CSV is usually the first step in a workflow that finishes somewhere else — your bookkeeping software, an ERP, a budgeting app, or a script you wrote yourself. You are not converting the PDF to read it; you are converting it to feed another system cleanly, without manual data entry.
This converter does the structural work for you. Drop in your bank statement PDF and download a CSV with transactions already split into columns. The first 3 pages are free; after that, credit-based pricing starts at €0.06 per page, with no subscription.
Anatomy of a converted CSV file
Open the CSV in any text editor and you will see the structure plainly. The first line is a header naming each column; every line below it is a single transaction, with values in the same order separated by a delimiter.
The columns we output are Date, Description, Debit, Credit, Balance, and Reference. Keeping Debit and Credit as separate columns is deliberate — it mirrors how a ledger reads and avoids ambiguity about which transactions are money in versus money out. If your destination tool wants a single signed Amount column instead, that is a one-step transformation in any spreadsheet.
Text is UTF-8 encoded, so payee names with accents or non-Latin characters survive the round trip intact. Each row corresponds to exactly one transaction, even when the original PDF wrapped a long description across several printed lines — the converter rejoins those before writing the row.
Opening a CSV without breaking it
CSV is simple, but spreadsheet apps can mangle it if you double-click to open instead of importing properly. Two pitfalls catch people out, and both are easy to avoid.
Long numbers and leading zeros. If a reference or account number is long, Excel may show it in scientific notation (1.23E+10) or strip a leading zero. To prevent this, open Excel first, use Data → From Text/CSV, and set those columns to 'Text' during import rather than letting Excel guess.
Delimiters and locale. In regions that use a comma as the decimal separator — most of continental Europe — spreadsheet software often expects a semicolon as the field delimiter instead of a comma. If every row lands in a single column when you open the file, that is the cause: import via Data → From Text/CSV and pick the correct delimiter, and the columns will split correctly.
If you are importing the CSV straight into accounting software rather than opening it yourself, none of this matters — the import wizard handles parsing. These notes only apply when a human opens the file in a spreadsheet first.
How accuracy works, and when to double-check
No automated PDF parser is perfect. Knowing where ours is strongest, and where it can struggle, keeps bad data out of your books.
What works reliably. Native PDFs from major banks — Chase, HSBC, Barclays, ING, ABN AMRO, Rabobank, BNP Paribas, Deutsche Bank, and most others — convert with high fidelity. Dates, amounts, debit/credit direction, and running balances all come through cleanly because these banks publish consistent, structured statements.
Where to look closely. A few categories are inherently harder:
- Multi-line descriptions with a separate reference, counterparty, and memo can be rejoined differently than you expect when collapsed into one CSV field.
- Foreign-currency transactions carry an exchange rate plus a foreign and a local amount; the layout convention varies by bank.
- Credit card statements use cycle and posting dates rather than value dates, and some list interest charges with no counterparty.
- High-volume joint or business accounts sometimes paginate mid-transaction, splitting one transaction's description across two pages.
How to verify. Because CSV is plain text, there are no formulas to lean on — so open the file in a spreadsheet and check that the final balance in the Balance column matches the closing balance printed on the last page of your statement. If it matches, the transactions reconcile. If not, scroll to find the discrepancy and fix it before importing.
For high-stakes use — tax filings, audit prep, formal reconciliations — spot-check 5 to 10 transactions against the original PDF before relying on the file.
Importing your CSV into accounting software
CSV is the format these platforms expect for manual bank imports. The exact menu path differs by tool — here is where to start with the most common ones.
- QuickBooks Online
- Banking → Upload from File → CSV. QuickBooks expects Date, Description, and a single signed Amount column (positive for credits, negative for debits). Merge our Debit/Credit columns into one Amount column in a spreadsheet first, then upload.
- Xero
- Accounting → Bank Accounts → Manage Account → Import a Statement. Xero accepts CSV and lets you map columns during import, so you can upload our file without restructuring it first.
- Sage
- Use the Bank feeds / Import statement option and choose CSV. Sage maps Date, Description, and Amount; confirm the date format matches your region during the import step.
- Exact Online & Twinfield
- Both prefer CAMT.053 or MT940 for live bank feeds, but accept CSV for manual import from converted PDFs. Use the bank reconciliation import and map our Date, Description, and Amount columns to the expected fields.
Wave, FreshBooks, Moneybird, and Zoho Books all take CSV through their bank-import sections with similar column mapping: Date, Description, Amount.
Whatever the tool, do a dry run with one month of data before importing a full year — it is far easier to delete and re-import a single statement than to undo a year of duplicated transactions.