The numbers tell a strange story. More than 90% of AP teams have digitised their processes in some form, yet over half remain stuck at what’s classified as medium automation. Another 30% are below 25% automation. The gap between ‘we have software’ and ‘it actually works end to end’ is vast.
In our SAP practice, we’re watching this play directly into hiring patterns. Finance directors aren’t just asking for SAP experience any more. They want people who can diagnose why their automation projects have stalled, and more importantly, who understand how AI agents might break the ceiling. That last skill is rare. Most candidates have implemented tools. Fewer have unstuck a process that was supposed to be automated but never quite got there.
The talent shift is subtle but real. A year ago, AP automation expertise meant knowing the right software and having a clean implementation under your belt. Now it means understanding where rules-based automation fails and how machine learning fills the gaps. The professionals who’ve actually done this, not in theory but inside a real finance function, are in short supply across the DACH region.
We’re advising clients to look beyond the obvious candidates. The person who fixed a half-finished automation project may be more valuable than someone who built one from scratch in ideal conditions. That problem-solving experience is exactly what AI-agent implementations will demand over the coming 12 to 18 months.
The automation ceiling is real. Breaking through it requires a different kind of hire.
Prompted by reporting from ERP Today.