Performance management is the dimension that measures whether an organization sees what AI delivers, and whether that information lands anywhere. Not whether a dashboard exists, but whether someone looks at what an AI application changes about lead time, quality or error margin, and whether that outcome influences the next decision. A level up does not cost a fixed amount, because the distance between levels differs per organization. An organization that measures nothing yet needs to set up something different than an organization that already measures but uses the outcomes nowhere. What is certain: the costs rarely lie in software. They lie in time, repetition and the willingness to take results seriously even when they disappoint.
The dimension looks at three things. First, whether an outcome is defined before an AI application starts — what needs to improve, and how that is to be seen. Second, whether that outcome is actually measured after some time, instead of assumed. Third, whether the measurement has a consequence: is an application adjusted, expanded or stopped based on what was measured, or does the outcome remain a figure in a report that nobody opens again. Many organizations score low not because they don't measure, but because the measurement has no place in a decision.
At baseline level, there is no description of what success means before an AI application begins. Afterwards, people look at whether something "felt better", without a fixed standard. What that means in practice for an organization at this level is described on the page about what the baseline level means in practice.
At foundation level, an outcome is named, but the measurement happens once or by chance. Someone checks after a few weeks whether it works, but there is no repetition and no fixed form. The page about what the foundation level means in practice shows how that situation differs from baseline and from the level after it.
From activation onwards, measuring becomes a habit rather than an exception: the same standard is revisited at fixed moments. At insight, that outcome is linked to a decision — an application that does not deliver what was expected beforehand is adjusted or stopped. At intelligence, the measurement is embedded in a broader rhythm: performance of AI applications is placed side by side and compared, so the organization learns which type of application does and does not take hold.
Performance management depends on dimensions that come earlier in the sequence. Without an organization that knows who is responsible for an outcome, a measurement remains a number without an owner — that starts with the question of how mature the organization is when it comes to AI, to be answered via the organization dimension. Without infrastructure that makes it possible to record data about use and outcome, there is nothing to measure — see the IT infrastructure dimension. And without data management that determines which data is reliable enough to act on, a measurement may produce a figure but no footing — see the data management dimension. Anyone who wants to raise performance management without having these foundations in order measures something, but not something that can be built on.
The step from baseline to foundation mainly requires an agreement: name what success means before an application starts, and record that in a place where it can be found again later. That is not an investment in technology, but in habit.
The step from foundation to activation requires repetition: the same standard returning at fixed moments, instead of a one-time look. That costs the time of someone who puts it on the agenda and keeps asking about it.
The step to insight requires something harder: the willingness to roll back or adjust an application if the measurement disappoints, even if it has already been invested in. That is less a cost item than a cultural question.
The step to intelligence finally requires a comparison between applications, so the organization does not learn case by case but sees patterns. That requires a fixed place where that comparison happens, and people who are given time for it.
The hybridresourcing maturity measurement shows whether an organization is capable of carrying what AI delivers — whether the foundations are in place, and whether performance management can build on them. It says nothing about which tasks within the work itself are suitable to be taken over by AI, and to what extent. That question is answered by the work scan from FTE TO AI, which calculates per task which part of it can be transferred to AI. Anyone who knows where the organization stands on this dimension also has an indication of whether the outcomes of such a work scan will land somewhere later, or disappear into a report that nobody opens again.
The maturity measurement, with the plot round in which multiple people score separately and the spread between their answers becomes visible, is still under development. Anyone who wants to use the measurement as soon as it becomes available can sign up for the waiting list.
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.