A COO steers on operations: throughput, capacity, quality, continuity. AI pilots that turn out well in an isolated corner of the business say little about that. The question that matters is not whether something works in a test setup, but whether it keeps working once it becomes part of daily operations, with all the dependencies that come with it.
An operation that has been set up in a certain way of working is expensive to reverse. If an AI application is rolled out before processes, data and responsibilities are in order, the result is not acceleration but extra work: corrections, exceptions, people still checking what a system was supposed to have already done. That risk affects the COO directly, because the operation is what he has to account for. A pilot that felt good in a demo, and six months later leads to disrupted delivery or extra error handling, is a problem that lands on his desk.
The opposite risk also exists: waiting too long to scale up because no one can indicate whether the organization is ready. Then competitors or other business units keep getting ahead, without there being a substantiated reason for it. Both risks arise from the same gap: there is no shared picture of where the organization actually stands.
A COO wants a basis for building decisions that does not depend on the enthusiasm of one pilot team. He wants to know whether the organization, apart from individual projects, is set up to carry AI: are processes described and repeatable, is the infrastructure stable enough, is data available and reliable where it is needed. That is a different kind of question than "does this specific application work". It is the question of whether the foundation is right, independent of whichever application is built on it later.
That foundation consists of multiple layers, and the order in which they are questioned is not arbitrary. Organization, IT infrastructure and data management determine what is possible; dimensions that depend on these, such as decision-making or collaboration around AI, cannot get further than that basis allows. A COO who knows this asks his questions in the right order instead of starting with the application that is most visible.
A COO does not accept an answer that rests on a single opinion. "The team is ready" is a statement from one person, based on what that person sees from his position. An IT manager, an operations manager and a team lead often see a different organization, even when talking about the same department. That spread is precisely why a score from one person is not enough: it conceals disagreement that will surface later anyway, once scaling begins.
That is why a measurement that has multiple people score separately works, across five levels — from baseline to foundation, activation, insight and finally intelligence — spread across seven dimensions. Not to calculate an average, but to see where the answers diverge. A large spread on a dimension is itself information: it means there is no shared picture, and that is something different from a low level. An organization that scores low but consistently across all dimensions at least knows where it stands. An organization with a large spread does not yet know that, and must figure that out first before a level means anything.
A COO also does not accept an answer that suggests a guarantee. A measurement of readiness says something about the state of the organization at this moment, not about what an AI application will deliver later. That distinction is not subtle for someone with operational responsibility: he knows that readiness is a precondition, not a result.
Operations is not the only layer where these kinds of questions play out. What continuity of processes is for the COO is a different consideration for other roles in the business: what a CIO looks for in AI maturity describes the question from infrastructure and system architecture, while what a CHRO looks for in AI maturity addresses what this requires from people and roles. For companies where operations are built up from projects and locations rather than a central process, different patterns apply, as described in how far construction has come with AI maturity.
The measurement that makes this spread visible is under construction. Those who join the waiting list get access as soon as the tool is ready for use. There is currently no report to download and no session to book; there is an instrument in development that is built on the question a COO actually asks, and not on the question that is easy to answer.
The maturity measurement answers whether the organization can carry AI: whether the foundations are in place to build something sustainable. Once that picture exists, the question shifts from capacity to content: which part of the work can actually be taken over by AI, task by task, rather than organization-wide. That is a different measurement, with a different kind of answer, and a COO finds that in the work scan of FTE TO AI.
Vraag maar wat er moet staan voordat AI in uw organisatie kan landen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.