Automating Invoice Coding in IFS: Why Most Teams Still Do It Manually

IFS is a powerful ERP – so why are 80% of non-PO invoices still coded by hand?

If your finance team is using IFS, chances are you’ve made big strides in ERP-led digitalisation. But for many IFS customers, one surprisingly manual process still eats up time, budget and team focus: coding non-PO invoices. Despite the tech stack, despite the workflows, most finance teams are still manually entering GL accounts, cost centres, VAT codes and more. And they’re doing it for thousands of invoices every year.

This is the automation gap – and it’s bigger than you think.

The IFS Automation Gap: Where Digitisation Stalls

Invoice automation inside IFS isn’t broken, but it’s not complete either.

Most companies use a combination of OCR tools and hardcoded rules to try and speed things up. But here’s the reality:

  • OCR can help extract data from PDFs or scanned invoices, but it can’t decide where that invoice should be coded. It still needs someone to fill in the blanks.
  • Rules, on the other hand, are rigid. They’re time consuming to build, hard to maintain, and they break the moment something changes – like a new supplier, a missing reference, or an unusual invoice type.

The result? 80% of non-PO invoices are still coded by hand. That means:

  • Slower approvals
  • Human error and inconsistencies
  • €4–8 per invoice in admin cost
  • A finance team stuck doing work your ERP should already be handling

Why the Status Quo Persists

It’s not for lack of trying. We’ve seen IFS customers pour time and budget into building better rules or updating templates. But the reality is this:

Most companies only have a small percentage of repeatable, rule-based invoices. The rest are too variable.

And building logic to cover every edge case just doesn’t scale. Especially when the AP team is already under pressure to get invoices out the door.

Enter Snowfox: AI That Understands Your IFS Data

This is where Snowfox comes in.

Snowfox is a native AI add-on built for IFS Finance. It doesn’t need rules or templates. Instead, it learns directly from your historical invoice data, understanding your actual coding and approval logic.

From the moment it goes live, Snowfox can predict how to code and route over 90% of your non-PO invoices. That means:

  • No more manual coding
  • No rules to maintain
  • No interface changes or new systems
  • Just a smarter layer that works inside IFS and improves with every invoice

Want to See It in Action?

If you’re still manually coding invoices inside IFS, you’re not alone. But there is a smarter way.

We’ll take 6–12 months of your historical invoice data and show you exactly how accurately Snowfox could automate your non-PO invoice coding. No guesswork, just real insight based on your real data

About the author

Miikka Savolainen

Miikka Savolainen is the COO of Snowfox, where he’s responsible for driving operational excellence and aligning strategic initiatives across teams to support growth and scalability. With a deep background in finance automation and a sharp focus on AP innovation, Miikka is passionate about using AI and fintech to streamline operations, reduce friction, and future-proof the finance function.