Preparing Your Files for the Best Results

DocuClipper reads almost any bank statement, credit card statement, invoice, or check, but the quality of the file you upload has the single biggest effect on accuracy. A few seconds of prep up front prevents the most common extraction issues - missing transactions, debits and credits flipped, or a statement that won't reconcile.

This guide covers the checks that matter most, and how to fix a file that needs it.

File types you can upload

DocuClipper accepts PDFs, common image files (JPG, PNG), and TIFF images (.tif / .tiff). Scanners and fax/document-management systems often save to TIFF, so you can upload those directly - DocuClipper normalizes a TIFF to a PDF automatically when it comes in, then extracts from it the same way. (For the bank-transactions and forms converters you can also upload CSV and Excel files; see Convert bank transactions CSV.)

Whatever the format, the quality guidance below still applies - a sharp, upright, full-page TIFF extracts well; a blurry or skewed one doesn't.

Start from the original PDF when you can

• Download the PDF directly from your bank's website rather than scanning or photographing a paper copy. Original ("text") PDFs are the most accurate input by far.
• Most banks keep 12-24 months of statements in online banking, so even older periods can usually be downloaded rather than scanned.
• If you only have paper, scan it rather than photographing it with a phone - scans are flatter, sharper, and properly aligned.
• Avoid screenshots and re-saved/"printed-to-PDF-from-a-photo" files. They lose the crispness DocuClipper relies on.

Scan paper statements before marking them up

If your team ticks off, highlights, or annotates statements while reviewing them, scan the statement first and mark it up afterwards (or scan a clean copy). Pen ticks, check marks, circles, and handwritten notes on or next to a date or amount are read as part of the text, so 01/22/2026 can come through as a01/22/2026 and 400.00 as 400:00. A row whose date or amount can't be read is left out, and the statement won't reconcile.

Make sure the pages are upright

If a page opens sideways or upside-down, rotate it the right way up and save before uploading. A statement that is rotated 90° puts the rows and columns in the wrong direction, which can cause whole pages to come back empty or with scrambled columns.

• Mac Preview: open the PDF, select the page thumbnails, Tools → Rotate Left/Right, then File → Save.
• Adobe Acrobat: Organize Pages → Rotate, then save.
• Free online tools (e.g. a "rotate PDF" tool) work too - rotate, download, then upload the corrected file.

One statement page per PDF page

DocuClipper expects one statement page per page of the PDF. Some exports - especially legal/DSAR "subject access request" packs and printed booklets - place two statement pages side-by-side on a single sheet ("2-up"), sometimes rotated to landscape. This layout interleaves two different pages' rows and is the hardest case to read accurately.

If your file is laid out two-up:

1. Re-download a standard, one-page-per-page copy from your bank if one is available - this is always the best option.
2. If not, split the sheets so each statement page is its own page (the "split/extract pages" feature in Acrobat, Preview, or a free PDF tool), and rotate them upright.
3. Then upload the cleaned-up file.

Keep the full page, in good quality

• Don't crop - the running balance column on the right edge is what DocuClipper uses to reconcile. If it's cut off, the statement won't balance.
• For scans, aim for ~300 DPI, straight (deskewed), and high contrast. A flatbed or a good sheet feeder in grayscale or black and white works best. Faint, skewed, or low-resolution scans are the most common cause of misread amounts and dates.
• Keep every page fully in frame, with no shadows, folded corners, or staples covering the text.
• Password-protected bank statements upload as they are: type the password in the Password required box under the file. For invoices, receipts, and tax forms, remove the protection first (open the PDF, print/save it as a new unprotected PDF). See Upload problems and errors.

Keep each file to one statement, complete and in order

DocuClipper reconciles a statement by starting from its opening balance, applying every transaction, and checking that it lands on the closing balance. Those balances are usually printed on the first and last pages, so the statement only balances when the whole of it is there, in order.

• Upload the full statement PDF your bank issues, with all pages in order from page 1. In online banking this is usually under Statements or Documents, not the transactions list.
• A browser "Transactions" printout won't balance. Printing or saving your online-banking transactions page gives a list of rows without the statement's opening and closing balances.
• A single page won't balance either. A file that holds only "page 2 of 2", for example, has the transactions but not the opening balance on page 1.
• Upload one statement per file (one account, one statement period) when you can, or at least one account per file. Multi-statement files work, but a year of statements, or statements from several banks, in a single PDF is harder to split into the right periods.
• Keep the pages in order, from page 1 to the last page. Pages in reverse or jumbled order won't balance, because the opening and closing balances end up on the wrong side of the transactions. When scanning a stack of statements, check that no statement went through the feeder back to front.
• Include every page, once. A missing page drops its transactions, and a statement included twice counts its transactions twice.
• Include only the statement itself - plus its check images if you want payees filled in (see Auto-fill payee from check images). Leave out cover letters, disclosures, deposit and withdrawal slips, and unrelated documents.

Choose the right upload type

Use Add Documents → Bank/CC Statements (PDF) for bank and credit card statements, including statements that print check images. Add Documents → Check Images is for files that contain only check images: it extracts each check (number, date, payee, amount) and does not read the statement's transactions or balances.

A few more tips

• Only need part of a large file? After you upload, you can select just the pages you want so DocuClipper processes only those - handy for very large packs. Untick Process this page under any page you want to skip (click the box or the words, or the page thumbnail itself); skipped pages show a red cross and the "This job will convert N pages" count drops. Uploaded the wrong file? Click the red trash icon on its row to remove it before you convert.
• Invoices and receipts: one document per file gives the most reliable line-item and total extraction. A file holding a stack of receipts works too, as long as each receipt is on its own page: scan or photograph them one receipt per page, not two side by side on one sheet.
• Statements of account go in as statements. A supplier statement, a bank or credit card statement, or a ledger is a list of transactions, not an invoice. Uploaded as an invoice it is labelled as a statement and is not synced to your accounting software; upload the individual invoices instead, and bank or card statements under Bank/CC Statements (PDF).

If something still looks off

Even with a clean file, occasionally a transaction is misread. DocuClipper flags these for you, and you can fix them in a click:

• Bank Statement Reconciliation Issues
• Fixing Missing or Wrong Transactions
• Debits and Credits Are Backwards

If a statement still doesn't extract correctly after cleaning up the file, send it to our support team - we tune DocuClipper for new statement layouts regularly and can usually turn a fix around quickly.