As automation absorbs the transactional layer of procurement, the relational layer it leaves behind turns out to have been deciding the outcome all along — and pressure quietly bends it in ways no dashboard catches.
Elvire Regnier, a former CPO, made an observation on LinkedIn this week that is worth sitting with. As AI absorbs the transactional work of procurement, she argued, the capability that will define the function is emotional intelligence — the reading of what a supplier does not say, the trust built over years that holds when a crisis arrives, the judgment of exactly when to give ground. Most organizations, she noted, know precisely how to measure cost savings and cycle times, and have almost no way to measure any of that.
She is right, and the gap she names is about to get expensive. But I would put the point a little differently, because I think something more specific is happening than “human skills matter more now.”
AI is not raising the value of emotional intelligence. It is revealing a value that was always there and never showed up on the dashboard. For decades, procurement optimized the transactional layer because it was visible, measurable, and controllable. The relational layer was none of those things, so it was treated as soft — a nice-to-have around the edges of the real work. Then disruption arrives, and the outcome turns on trust, credibility, and who anticipated the problem before it surfaced — none of which were ever on the report. Automation simply strips away the measurable layer faster, and leaves the determining one exposed.
Here is the part that should worry a CPO more than the measurement gap, and it is where Elvire’s own argument cuts deepest. She observed that pressure reshapes behavior. It does — and it does so invisibly, rationally, and against the organization’s own interest.
A documented case from my own work makes it concrete. Years ago I was brought into a field-service operation running at 51% next-day parts delivery against a 90% contractual requirement, and asked to fix procurement. But the breakdown did not start in procurement. The technicians were measured on service-call volume, and stopping to order parts after each call slowed them down — so they batched their orders to the end of the day to protect their numbers. Entirely rational, inside their own frame. Except the late orders missed customs windows and arrived too slowly, which is what wrecked delivery — and the parts that never came on time left open the very calls the technicians had rushed to close. Their own call-closure rates were being destroyed by the behavior they adopted to protect them. No one saw it, because no one was looking at the line connecting the technician’s incentive to procurement to the technician’s own result.
That is what “pressure reshapes behavior” looks like with the numbers attached. It is not a soft observation. It is a hard, measurable loop in which every agent acted reasonably and the system quietly defeated itself — and no dashboard in that organization would ever have surfaced it, because each metric looked fine on its own.
This is why Elvire’s point matters beyond procurement, and why it is converging with what people in very different disciplines are starting to say in their own languages. The trust, the relationships, the reading of the room she describes are not separate from the operating mechanics — they are the mechanics, operating in the layer no one instruments. As AI puts more agents into that layer — human and increasingly non-human, each optimizing its own objective under its own pressure — the scarce capability is the one that can see across them: locate what each is really responding to, and align them before the loop closes on itself. Emotional intelligence is one name for part of that capability. There are others. They are circling the same thing.
The technologies keep changing. What decides the outcome — and what pressure keeps quietly bending out of view — does not.
With thanks to Elvire Regnier, whose LinkedIn post and video prompted this.
Truth Is Believing. Accuracy Is Knowing.
-30-
Related
AI Isn’t Replacing Emotional Intelligence. It’s Exposing What You Never Measured.
Posted on June 23, 2026
0
As automation absorbs the transactional layer of procurement, the relational layer it leaves behind turns out to have been deciding the outcome all along — and pressure quietly bends it in ways no dashboard catches.
Elvire Regnier, a former CPO, made an observation on LinkedIn this week that is worth sitting with. As AI absorbs the transactional work of procurement, she argued, the capability that will define the function is emotional intelligence — the reading of what a supplier does not say, the trust built over years that holds when a crisis arrives, the judgment of exactly when to give ground. Most organizations, she noted, know precisely how to measure cost savings and cycle times, and have almost no way to measure any of that.
She is right, and the gap she names is about to get expensive. But I would put the point a little differently, because I think something more specific is happening than “human skills matter more now.”
AI is not raising the value of emotional intelligence. It is revealing a value that was always there and never showed up on the dashboard. For decades, procurement optimized the transactional layer because it was visible, measurable, and controllable. The relational layer was none of those things, so it was treated as soft — a nice-to-have around the edges of the real work. Then disruption arrives, and the outcome turns on trust, credibility, and who anticipated the problem before it surfaced — none of which were ever on the report. Automation simply strips away the measurable layer faster, and leaves the determining one exposed.
Here is the part that should worry a CPO more than the measurement gap, and it is where Elvire’s own argument cuts deepest. She observed that pressure reshapes behavior. It does — and it does so invisibly, rationally, and against the organization’s own interest.
A documented case from my own work makes it concrete. Years ago I was brought into a field-service operation running at 51% next-day parts delivery against a 90% contractual requirement, and asked to fix procurement. But the breakdown did not start in procurement. The technicians were measured on service-call volume, and stopping to order parts after each call slowed them down — so they batched their orders to the end of the day to protect their numbers. Entirely rational, inside their own frame. Except the late orders missed customs windows and arrived too slowly, which is what wrecked delivery — and the parts that never came on time left open the very calls the technicians had rushed to close. Their own call-closure rates were being destroyed by the behavior they adopted to protect them. No one saw it, because no one was looking at the line connecting the technician’s incentive to procurement to the technician’s own result.
That is what “pressure reshapes behavior” looks like with the numbers attached. It is not a soft observation. It is a hard, measurable loop in which every agent acted reasonably and the system quietly defeated itself — and no dashboard in that organization would ever have surfaced it, because each metric looked fine on its own.
This is why Elvire’s point matters beyond procurement, and why it is converging with what people in very different disciplines are starting to say in their own languages. The trust, the relationships, the reading of the room she describes are not separate from the operating mechanics — they are the mechanics, operating in the layer no one instruments. As AI puts more agents into that layer — human and increasingly non-human, each optimizing its own objective under its own pressure — the scarce capability is the one that can see across them: locate what each is really responding to, and align them before the loop closes on itself. Emotional intelligence is one name for part of that capability. There are others. They are circling the same thing.
The technologies keep changing. What decides the outcome — and what pressure keeps quietly bending out of view — does not.
With thanks to Elvire Regnier, whose LinkedIn post and video prompted this.
Truth Is Believing. Accuracy Is Knowing.
-30-
Share this:
Like this:
Related