I’ve written before about how many organisations don’t have anyone who is accountable for the full operating model. Fewer than 50% of the Fortune 500 have a Chief Operating Officer. It’s true that there could be other roles within the organisation that own pieces of operational excellence, but if there is no one at the C-suite who owns the overall operations of the business, it’s likely that there isn’t anyone who has both the accountability and the authority to manage the full operating model. There’s no one who is watching to see as things slip. There’s no one who can see that a planned change in one function is about to blow up a process somewhere else.
I watched this happen directly. At Visa, a decision was made to split one organisation into two. At a global level, the split made sense. It was two clean lines with two clear mandates. What the decision didn’t account for was that in many markets, a single person or small team was responsible for both sides of that line, and there was no way to separate the work between them without breaking something. The global view was clean. The operating reality underneath it wasn’t. No one had looked at both levels at once before the decision was made, so no one caught it until it was already a problem to solve rather than a risk to plan for.
That’s what happens when operational excellence is managed function by function rather than owned end to end.
Systems, processes, and tasks get duplicated. Things are being held together by heroics and individual experience. Incentives for different teams pull in different directions. Even where these issues don’t create critical failures, they are an ongoing drag on the organisation. Things are harder, slower, and less effective than they should be.
AI doesn’t create this problem, but as with many things, it raises the stakes.
On one side, with more and more talk about what can be automated or handed off to agents, this lack of a holistic view and accountability is a real risk. If no one is looking at how all of the pieces fit together, how different agents are or aren’t aligned, and what happens between individual initiatives, the problems that already exist will become worse. Fragmentation will become faster and less visible and the opportunity to mitigate those gaps with human workarounds will be gone.
On the other side, automation and agentic AI can be an incredible solution to the problems that have plagued operational leaders for years. There are options to solve entrenched problems that were avoided because it would be too slow or too expensive to solve the old way. Many of these problems are cross-functional and there are now real options to consider. But to be able to take advantage, leaders need to know about the problems and then understand their enterprise impact.
That’s the cost of the missing role and it’s got two sides. Without someone looking and managing across the whole, organisations keep making decisions that look clean, but aren’t, and end up breaking something somewhere else. At the same time, they miss the chance to fix problems that have been sitting there for years.
Every organisation needs someone who is looking across the whole. Someone needs to see how the pieces work together. Someone needs to be able to say when an individual team or function needs to compromise for the greater good. Without that, any transformation, AI-related or not, won’t fix the fragmentation and it might make it worse.