The most revealing number in SAP's workforce research isn't the 62% of executives unhappy with their data integration, it's the gap between the 50% planning for AI's impact on productivity and the 21% planning for its impact on job design. That gap is where the real story lives. It suggests we're still treating AI as a faster way to do the same work, rather than as a force that changes what the work even is. A sharp point emerges: you can't decide where to automate without understanding how that decision reshapes the teams, roles, and skills around it. Yet most organizations are trying to do exactly that, and the pieces don't fit.
This connects to a theme we've been tracking closely: the gap between what AI tools can do and whether the humans using them actually understand the mechanics. As we noted in Verify Your AI's Understanding: A Simple Check for Tax Season, the challenge isn't always building the model, it's knowing what it's actually doing under the hood. The same logic applies at the organizational level. You can have a workforce plan that models automation, headcount, and contractor spend in perfect detail, but if your CFO and CHRO are working from different systems and different assumptions, the plan is fiction. The point about the "workforce" quietly expanding to include contractors and AI agents is critical here. Most planning models still assume a binary world: employees or machines. The reality is a spectrum, and the planning tools haven't caught up.
What's genuinely useful is that the fix isn't a software purchase. The hard part, as the authors state plainly, is a governance problem. CFOs and CHROs are being pushed into the same room not because of a cultural shift toward collaboration, but because the decisions genuinely require both perspectives simultaneously. That's a structural change, not a philosophical one. When we look at the shifting skill requirements in the AI job market, as we did in Navigating AI/ML Job Requirements: A Shift in Expected Skills, the same tension appears: the roles themselves are being redefined faster than the people defining them can adapt. The same is true for workforce planning. It's not just about tracking who has which skill; it's about understanding how those skills interact with intelligent systems that are now doing execution-layer work.
The practical takeaway here is direct: if you're a leader, stop asking whether to hire, automate, or reskill. Start asking how work should be configured across humans and intelligent systems, and at what cost. That question can't be answered by finance alone, or HR alone, or procurement alone. It requires a shared picture of capacity, skills, and cost, not reconciled after the fact, but visible from the start. The organizations that figure this out won't just have better data; they'll be asking better questions. The ones that don't will keep making workforce decisions in the dark, and the pace of those decisions is only accelerating. The specific thing to watch isn't the technology, it's whether the CFO and CHRO can agree on a single set of metrics before the next quarter's planning cycle begins. That's the real test, and it's one most leadership teams are still failing.