Invoice to Excel: One Row Per Line Item, Not One Blob Per PDF
An invoice is not a table — it is a small document with a header block, a line-item table, a tax section, and a total that sits somewhere unpredictable. Extracting it well means knowing which region is which, not just OCR-ing the page.
Convert one file free — see the table land in Excel
Free daily conversions · scanned invoices supported · files shredded after processing.
Why invoice extraction goes wrong
Every vendor has a different layout
A template-based parser breaks the moment a supplier changes their PDF generator. You need layout understanding, not coordinates hardcoded to one form.
Line items versus totals
Subtotal, discount, VAT, shipping, and grand total all look like table rows. Mix them into line items and your spend analysis double counts.
Multi-page invoices
Line items continue on page 2 with the table header repeated. Pasted text arrives with the header rows embedded mid-data.
Wrapped descriptions
A long description wraps to a second visual line and is parsed as a second line item, throwing off quantities and unit prices.
Currency and tax conventions
European invoices use comma decimals and tax-inclusive amounts. Treating 1.234,56 as 1.23456 ruins totals.
Scanned supplier invoices
Many arrive as scans or phone photos and contain no text layer at all.
What to extract from each invoice
- Header fields: vendor name, invoice number, invoice date, due date, purchase order reference, currency.
- Line items: description, quantity, unit price, tax rate, line total.
- Charges below the table: shipping, discount, VAT, rounding.
- Totals: subtotal, tax amount, amount due — kept distinct from line items.
Practical workflow
If you are processing a stack of invoices, do not try to build one giant sheet in a single pass. Convert in batches of one vendor or one month, confirm the columns line up, then append.
Keep the source filename in a column. When a total looks wrong three weeks later, the only thing that saves you is being able to find the original PDF in seconds.
Convert one invoice first
Check header fields and the line-item block on a representative sample from each vendor.
Normalize in Excel
Add your own columns for cost centre or category; keep vendor fields untouched so you can trace back.
Batch the rest
Once the layout for a vendor is known, the same document class converts consistently — verify totals, not every row.
Invoice OCR is not enough on its own
Plain OCR gives you text in reading order. For an invoice you also need to know that the block of numbers below the table is not another set of line items, and that the number after "Total due" is a total rather than a repeated subtotal.
That is a document-understanding step. It is why two invoices with identical content but different supplier templates can produce identical structures here — the layout is interpreted, not memorized.
Frequently asked questions
Can I convert a scanned invoice to Excel?+
Yes. Image-only invoices are OCRed first and then structured. Sharp scans and flat phone photos work best; crumpled paper or heavy stamps may need a manual check.
Does it work with any vendor format?+
No parser is universal, but this one is layout-based rather than template-based, so it adapts to unfamiliar invoices far better than fixed-coordinate tools. Always spot-check a new vendor format.
Can it read invoices in other languages?+
OCR supports many languages, including accented European text. Number conventions such as comma decimals are the more common source of error — check one invoice per country.
What about invoice images (JPG/PNG)?+
Images work the same way as scanned PDFs. Fill the frame with the page and avoid shadows.
Is there a free tier?+
Yes — free daily conversions with no account required. Volume packs are available when you need to process batches without a subscription.
Keep reading
Done reading? Convert a document and check the output yourself.
Extract an invoice free