Still retyping invoices?
Five minutes per invoice, fifty times a month. Software reads the invoice itself and puts a posting proposal in front of you — you check it and approve.
From incoming PDF to posting proposal
Reading
Supplier, invoice number, date, due date, VAT number, amounts per rate and the structured reference are recognised — even when every supplier uses its own layout.
Matching
The invoice is linked to the right client file and the right supplier. If that supplier does not exist yet, it is proposed rather than silently created.
Proposing
Based on what was posted previously for the same supplier, the workflow proposes a ledger account and a VAT treatment. Anything that deviates from the pattern is flagged.
Checking
You see the proposal next to the original invoice, and approve or adjust. Only then is it posted and filed in the client folder.
What does not happen: the workflow never invents missing data, never posts when in doubt, and never lets through anything that does not add up. An invoice whose detail lines do not total the amount goes to the exceptions list, not into your books.
One batch of thirty invoices
Today
- Pull attachments from the mailbox and rename them
- Per invoice: open, read, retype the data
- Look up or create the supplier
- Determine the ledger account and VAT treatment
- Post and file it in the client folder
- Note the doubtful cases for later
± 2 hrs 30 per batch
With the workflow
- Invoices arrive and read themselves
- Supplier and file linked automatically
- Account and VAT proposed from history
- You run through the proposals and approve
- Posting and filing happen automatically
- The two doubtful cases sit apart, with the reason
± 30 min per batch
Indicative figures for illustration, not a measured result at a specific firm.
About invoice processing
What exactly is invoice recognition (OCR)?
Software that automatically extracts the amount, supplier, date, VAT number and reference from an incoming invoice, so you no longer have to retype them.
How accurate is the reading?
For suppliers that recur often, recognition is highly reliable because the workflow learns their layout. With an unknown supplier or a poorly scanned document, a field can be missing. That is why nothing is posted silently: uncertain fields are flagged and put to you separately.
From what invoice volume is this worthwhile?
The difference shows from a few dozen invoices a month, and the gain scales as volume grows. Just as important as volume is repetition: the more often the same suppliers recur, the more the workflow handles on its own.
Does this work with my accounting package?
With any package that allows an API or an import — including Exact Online, Octopus, Yuki, Adsolut, Expert/M, WinBooks, Sage BOB 50, Horus, Popsy, Unit4 Venice and Odoo. Document platforms such as Billit, Billtobox, ClearFacts and CodaBox are supported too.
What happens to invoices the workflow does not understand?
They land in a separate list with the reason attached — a missing VAT number, say, or an amount that does not match the detail lines. They are never half posted and never disappear silently.
Often in the same breath
Where does your time actually go?
We walk through your processes together, at your pace. You will know what can be automated in your firm, what it would save, and what is better left as it is. No strings attached.