SAP has started folding AI token consumption into standard finance planning cycles. The catalyst was straightforward: as employees and automated agents scaled up their AI use, the bills arrived with no clear owner and no visibility into what drove them. Token caps, model routing rules, and forecasting are now part of the picture.
CFOs want to know what the AI initiative will cost next quarter. SAP teams talk in tokens and inference calls. Nobody in the room can translate one into the other, and the conversation stalls. Companies need finance professionals who can sit in that space and make the numbers legible to both sides.
A Frankfurt-based manufacturer approached us in early spring looking for a finance business partner with hands-on SAP analytics experience and enough technical literacy to build cost models around AI usage. All three requirements in one spec. We placed someone from a consulting background who had worked on cloud cost optimisation. She had never touched token economics before, but she understood consumption-based pricing and could hold her own with the data science team.
AI workloads scale with adoption, which means the spend is a variable operating cost. Finance teams will need people who can forecast that spend, allocate it to business units, and defend the assumptions to leadership. The SAP talent pool has not historically produced many candidates with that profile, so companies are pulling from adjacent disciplines: cloud FinOps, SaaS commercial management, energy trading analytics.
A finance manager who can read a token consumption dashboard and turn it into a quarterly budget line has an immediate advantage over one who cannot. The companies investing in AI want governance, and governance means finance involvement from the start.
Prompted by reporting from ERP Today.