In stressed companies, the most consistent observation we make is structural rather than personal: the management team's narrative is internally coherent, the explanations are individually plausible, and the underlying signals are describing a different company.
This is not deception. The management team usually believes the version it is presenting, and most of the explanations are factually correct in isolation. But the cumulative pattern — the indicators, the variances, the small exceptions that have become routine — is telling a different story. The Board, hearing the narrative, has to do something the system has stopped doing internally: separate the explanation from the evidence.
1. Every miss has a reason. The reasons add up to something else.
The first signal is the most subtle. Each individual variance has an explanation that is, on its own, reasonable. The customer who delayed an order. The supplier issue that pushed a shipment. The market softness in a specific segment. The team transition that slowed a project.
Listened to one at a time, the explanations dissolve concern. Listened to over six or twelve months, they describe a system in which something is consistently going wrong somewhere, and the boundaries between unusual circumstance and operating reality have begun to blur. The pattern is the variance. The variance is no longer an exception. But because each instance has been explained, the cumulative drift is not registered as one.
The macro environment makes the pattern harder to detect. Italy entered 2026 with industry described, in Confindustria's own reading, as volatile: weak export, fragile consumption, energy still elevated, partial relief from policy measures. When the macro produces multiple plausible explanations every quarter, almost any individual variance can be attributed to it. The discipline the Board needs is not to disbelieve those attributions, but to read whether the sum of them still describes a company in control.
2. Exceptions become the operating model
The second signal is process drift. A pricing exception that was approved once because of an exceptional customer becomes a standing arrangement. A reporting cycle that was extended once for a quarter-end issue stays extended. A capex authorisation that was waived in one project becomes the default in subsequent ones.
Each accommodation is rational at the moment it is made. Together, they describe an organisation that has gradually replaced its formal operating model with a layered set of informal exceptions — and a management team that is operating inside an exception structure it can no longer fully see. When the next variance occurs, there is no clean process to reference, only a precedent of accommodations that makes the next accommodation easier than the original discipline.
In production environments, the most common version of this drift is a quiet shift toward push-mode operating logic — exceptions to scheduling, replenishment, and prioritisation compounding into a system that runs on accommodation rather than design. A high-level lean read is usually enough to surface the gap between the operating model the company describes and the one it is actually running. The same accretion happens in finance, where cash-management practices that were once adequate become the basis for decisions the underlying tooling was never designed to support.
3. Reporting arrives after it would have been useful
The third signal is timing. In well-governed companies, variances are surfaced fast enough that they can still be acted on. In drift, reporting tends to lag — sometimes by weeks, sometimes by an entire reporting cycle. The variance becomes visible only after the moment for action has passed, at which point the explanation replaces the response.
This is rarely a deliberate choice. It is the result of reporting cycles that were calibrated to a previous level of stability, and that have not been retuned for the current rate of variance. The company is, in effect, governing itself in arrears — looking at last quarter's reality and explaining it, while the next quarter's reality is already being shaped by decisions that are not being made because the information they would require has not yet arrived.
What it takes to read the system, not the narrative
The Board's job, in this kind of situation, is not to disbelieve management. It is to read the system independently and let the evidence sit alongside the explanation. This requires three things in combination.
The reporting that the Board receives has to be the same reporting the operators are using, on the same cadence, without interpretation layered between them. The Board has to ask questions that test the system rather than the narrative — about the cumulative pattern of variances, the exceptions that have become routine, the timing between detection and response. And the gap, once it surfaces, has to be named on paper — in a brief, factual read of the system against a recognised frame, not as another general discussion. In finance, that usually means a direct read of how cash management is operating against the actual cycle; in production, a high-level lean read against the company's real flow logic. Putting the gap in writing, factually, is what allows a Board to act on it rather than to defer it once more.
This is uncomfortable, because it implies that competent management can be honest and still be wrong about the state of the company they are running. In our experience, this is the most common situation, not the rarest. The narrative and the system can coexist for years. They eventually converge — but the convergence is much more expensive when it happens through events than when it happens through decisions.
The most useful question a Board can ask in stressed situations is not is this true. It is do these explanations, taken together, describe a company in control? The Italian industrial map gives a way to test the answer empirically. Intesa Sanpaolo's research on industrial districts shows export resilience in 2025 driven by uneven performance — eighty-three districts growing, seventy-five contracting, often inside the same supply chain or region. The market and the district are no longer sufficient as references. The relevant question is how peers in the same conditions are reading the same evidence.
---
References
- Export e consumi zavorrano l'industria, Confindustria, February 2026.
- Distretti industriali: export resiliente nel 2025, incognite sul 2026, Intesa Sanpaolo Research, 2026.
- Replace, augment, disrupt: AI & organizational decision-making, Journal of Organization Design, October 2025.