Banks give you statements as PDFs. Excel doesn't read PDFs. So if you need transactions in a spreadsheet — for tax prep, expense tracking, budgeting, or bookkeeping — there's a gap to bridge.
This guide covers the realistic options for crossing that gap, in order from worst to best for most people.
Why don't banks just give you Excel?
Some banks do, but only for recent transactions. Chase lets you download a CSV of activity from the last ~90 days; Amex offers an Excel export of the current statement period; most others have similar limits. For anything older — a year-end catch-up, an audit, a switch in accounting software — you'll be working from PDFs.
Banks could in theory expose every transaction back to year one as machine-readable. They don't, partly because the systems running statement generation are old and stable in a way no one wants to touch, and partly because most customers don't ask. So PDFs it is.
Option 1: Retype it by hand
Realistic for one statement with a handful of transactions. Catastrophic for a year of monthly statements at 30–80 transactions each. Anyone who's done this once doesn't do it twice.
Beyond the time, the bigger problem is errors. Manual retyping introduces transposed digits, missed rows, and miscategorisation. When the numbers eventually have to reconcile against the statement balance, finding which of 300 hand-typed lines is the wrong one is its own evening of work.
Option 2: Copy-paste from the PDF into Excel
Sometimes the PDF is “digital” (its text is selectable) and you can drag-select a table, copy, and paste into Excel. It almost works.
In practice, PDFs lay out tables visually, not structurally, so the paste comes out as a wall of text where the columns shift around. You spend the time you would have spent retyping instead fixing column alignment. Multi-page tables (most statements) compound this. Scanned or image-based PDFs (more common than you'd expect) have no selectable text at all and this option is off the table.
Option 3: Run it through a converter
This is the workable option. Specialised tools — including our free one, DocuClipper, BankStatementConverter, and several others — extract transactions from PDF statements into clean CSV, Excel, or accounting-software-ready files.
What to look for in a converter:
- Handles scanned/image PDFs, not just digital ones. Many old PDFs and most scanned ones don't have a text layer; if the converter is doing PDF text extraction rather than visual reading, it'll fail on those.
- Reconciles against the statement's own totals. The best signal that an extraction is correct is that the line-items sum to the statement's stated purchases total. A converter that shows you “✓ balanced” or “⚠ off by $X” before you trust the file is doing the check most others skip.
- Correct sign conventions for the export formats it supports. QuickBooks and Xero expect credit-card purchases as negative numbers; an Excel/CSV file for human reading typically has them positive. A converter should know which convention applies and flip signs accordingly.
- Privacy. You're uploading a financial document. Tools that store your files indefinitely, train AI on them, or have unclear retention policies are worth avoiding. Look for explicit statements about in-memory processing and no retention.
The workflow
- Download the PDF. Sign in to online banking and grab your statement(s) for the period you need.
- Open the converter. If you're using ours, drag the PDF onto the upload box (or click “Choose a PDF”).
- Review the extraction. Check the preview table — does every transaction look right? Pay attention to the ✓ balanced indicator; if it says off-by-$X, something in the extraction needs review before you trust it.
- Download in the format you need. Excel for general use, CSV for further processing or import into other tools, QuickBooks .qbo for QuickBooks, Xero CSV for Xero.
- Open the result.A real Excel file has bold headers, frozen header row, currency formatting. CSV is plain text — open in Excel or Numbers to view as a table, but it's mainly meant to be fed into another tool.
Scanned vs digital PDFs
A “digital” PDF is one generated programmatically by the bank — the text is real text, selectable with a cursor. A “scanned” or “image-based” PDF is one where each page is a picture; the text on it is pixels, not characters. Photos of paper statements, statements re-scanned by mail-room systems, and some older bank exports fall into this category.
Traditional PDF-text-extraction converters fail on scanned PDFs because there's no text to extract. Vision-AI-based converters (ours is one) read the page visually — like a human looking at it — so the distinction doesn't matter. If you have a stack of statements and aren't sure which are digital and which are scans, try one of each through a converter and see what comes back.
The bottom line
For one statement, copy-paste might be acceptable. For anything more, use a converter — preferably a free one with a ✓ balanced check and clear privacy policy. The 60 seconds of upload-and-download time replaces hours of typing or column-fixing, and the output is more reliable.