Your PDF bank statement will not convert: what to check

Symptom to cause

First: is it text or a picture?

This one distinction explains most outright failures. Open the PDF and try to select a line of text with the cursor.

• If text highlights, the PDF has a text layer and should convert directly.
• If nothing highlights, or the whole page highlights as one block, the page is an image. It was scanned, photographed, or produced by printing to PDF from a scanner.

Image-only pages are handled with OCR, which is included on every plan. Quality matters more than anything else here: a flat, well-lit, straight scan at 300 dpi reads well, while a photograph taken at an angle in poor light may not. There is more in converting scanned statements.

A hybrid document — some pages text, some scanned — is common when a statement has been partly re-printed. Each page is classified on its own, so the text pages convert directly and only the scanned ones go through OCR.

Encrypted and password-protected files

Banks frequently email statements encrypted with a password derived from your date of birth, postcode or account number. Two different protections exist and they behave differently.

• A user password stops the file opening at all. You must supply it.
• An owner password allows opening but forbids copying or printing. This is what stops you selecting text even though the page is clearly text.

Password-protected PDFs are supported on paid plans; the password is used once to open the document and then discarded. On the free plan, remove the password in a PDF reader first and upload the unprotected copy. The full explanation is here.

Documents that are not transaction lists

Some documents banks call statements contain no transactions at all, and no converter can produce rows that are not there.

• Annual summaries and interest certificates: totals only.
• Credit card statements showing just a minimum payment and a balance, with the transactions in a separate document.
• Mortgage or loan statements listing scheduled payments rather than an account's activity.
• Screenshots of online banking, which show a fragment of a list without the period or the balances that let it be checked.

Download the transaction statement for the period from online banking instead; see downloading statements.

When rows are missing

This is the failure worth taking seriously, because the output looks plausible. The checks exist to catch exactly this.

1. Read the Conversion Report. If the balance check failed, the difference is the size of what is missing.
2. Compare the row count against any count the statement prints, such as "28 Debit(s) This Period".
3. Compare the extracted totals against the statement's own totals for credits and debits.
4. Look at the first and last dates. Missing rows often sit at the start or end of a page.
5. If a whole section is absent, note which page it is on and report it.

Layouts that lose a block usually do so for a structural reason — a second table on a page whose first table had the only recognised header, or a heading that looks like a column label. Those are fixed in the extractor rather than worked around, so reporting one is worth more than editing around it.

Reporting a statement that will not convert

The most useful report says what the statement should have produced, because that is what makes a fix verifiable.

• The bank and the country.
• How many transactions the statement contains, and how many were extracted.
• Which page a missing block is on.
• The totals the statement prints for the period.

Send it through contact or the feedback control on the conversion itself, which attaches the conversion reference automatically. The document is not needed to start with — the structure usually is.