I was asked to speak on invulnerability in front of a panel of manufacturers. The question put to me was direct: can French industry become invulnerable? It deserves to be taken seriously, and above all it deserves to be broken apart.
The energy transition, Asian competition, geopolitical instability and technological acceleration form a stack of shocks with no equivalent in recent industrial history. Faced with that, the word invulnerability comes up naturally in executive committees. It reassures more than it clarifies.
| The confusion | The scope | The evidence |
|---|---|---|
| Three questions asked as oneA leader who asks about industrial invulnerability is mixing together the survival of their organisation, their sector’s passage through the transitions and the very relevance of industry as an economic form. These three questions call for answers of a different nature. | The one they have power overAn industry does not become invulnerable by sector decree or by public policy. It becomes invulnerable one organisation at a time. The first question, the one about architecture, is the one that governs the other two and the one a leader can act on. | The gaps are architecturalWithin the same sector, exposed to the same constraints, some groups come through the transitions and others are subjected to them. The gap is explained by decisions taken one to two decades earlier rather than by size, resources or economic conditions. |
The three questions hidden inside one
When a manufacturer raises the question of invulnerability, they are in fact raising three, and it is rare to hear them stated separately. Confusing them produces answers that reassure and that deal seriously with none of them.
Three questions, three kinds of answer
- Can my organisation, already weakened, survive the disruptions that are coming? That is a question of organisational architecture, set out in the full definition of an invulnerable company.
- Can my sector come through the current transitions without breaking apart, at a time when it is struggling to attract people? That is a question of sector transformation.
- Does industry in general have the foundations it needs to stay relevant? That is a question of economic civilisation.
Why the first one governs the other two
This article concentrates on the first one, for two reasons. It is the only one over which an industrial leader really has power, and it is the one that governs the other two.
An industry does not become invulnerable by sector decree or by public policy. It becomes invulnerable organisation by organisation, architecture by architecture and foundation by foundation. A solid sector is the sum of solid companies, never the reverse.
What invulnerability is not in industry
Three confusions dominate the debate about industrial solidity and lead systematically to investments that are insufficient or poorly targeted. Clearing them up is the precondition for any serious conversation.
Resilience is not invulnerability
Industrial resilience has been a rallying cry since 2020. Once global supply chains had revealed their fragilities, the almost universal response was to invest in supplier diversification, buffer stock and the mapping of critical dependencies.
These investments are necessary and they do not build invulnerability. A resilient organisation absorbs the shock and returns to equilibrium. An invulnerable organisation has built, before the shock, the foundations that prevent it from being destabilised for any length of time. The difference is one of nature rather than of degree, and it is developed in the comparison of the five logics.
Digital transformation is not invulnerability
For the past five years, digital transformation accelerated by artificial intelligence has been presented as the answer to industrial vulnerability. The factory of the future, digital twins and predictive maintenance all have documented operational benefits.
They do not address the deep structural fragilities. A perfectly digitalised plant, with a governance that fails to protect long-horizon decisions and a confused identity, remains a fragile plant with attractive screens. Digitalisation improves efficiency and operational visibility, and it does not build the architecture that allows an organisation to come through a fundamental rupture.
Extreme efficiency is not solidity
This is the oldest and most deeply rooted misunderstanding in industrial culture. Just-in-time, theorised in the 1970s and adopted massively in the 1990s and 2000s, cut inventories and tightened flows to a level of efficiency that implicitly assumed a stable environment.
When that environment was destabilised on several dimensions at once, the model revealed its downside. Global carmakers collectively lost more than 210 billion dollars of revenue in 2021 because of electronic component shortages. That loss did not come from unpredictable demand. It was the predictable result of a supply structure that had eliminated all of its redundancies.
Ironically, it was Toyota, the company that invented lean flow, that came through this crisis best, precisely because it had maintained buffer stock on critical components against its own principles, a legacy of the way it handled the 2011 earthquake.
What European competitiveness forces you to face
A solid architecture does not remove the constraints of competitiveness. Most discussions of industrial solidity miss that distinction, and it is essential if the conversation is to stay credible in front of leaders who live with those constraints every day.
The cost of energy and the Asian differential
At the end of 2024, Michelin announced the closure of its plants at Cholet and Vannes, affecting 1,254 employees. Its chief executive attributed the decision to a slow degradation of European competitiveness, with production costs that had doubled relative to Asia between 2019 and 2024 and an electricity cost four times higher than in China.
No organisational architecture compensates for a differential of that size. Claiming otherwise would be dishonest towards a European industrial leader, and it would discredit the whole of the rest of the argument.
What the architecture changes all the same
What the architecture changes is the way the organisation responds to that cost gap, and the coherence of its trajectory while it responds.
With each site closure in France, Michelin puts a transformation plan in place to avoid leaving an industrial wasteland, following a repeatable pattern that mobilises private partners and local authorities. At the same time, the group confirmed 60 million euros of investment at the Roussillon chemical platform to produce a bio-sourced molecule replacing compounds derived from petroleum.
Read in isolation, these decisions look contradictory. They are coherent when they are read through a stable identity, that of a group which defines itself as a player in sustainable mobility, closes what is no longer competitive in its current form and invests in what will be relevant in ten years. In 2025, the group reported operating income of 2.9 billion euros and a rate of renewable and recycled materials of 32%, a gain tied directly to investments committed several years earlier.
From the field
ArcelorMittal, a plan with no alternative scenario
The group built its European decarbonisation projects on an assumption it had formulated itself: a favourable combination of regulation, technology and market movement was supposed to make those investments viable. That assumption was reasonable at the moment it was made.
It did not hold. The group acknowledges that the necessary regulatory environment is not in place, and the projects have been pushed back several times. No workable alternative plan had been built for the case where conditions failed to fall into place as expected.
In February 2026, the group finally confirmed the construction of the largest electric furnace in Europe at Dunkirk. The decision arrives several years late. In the same year, the group reported a net profit of more than 3 billion dollars and announced 600 job cuts in France.
An industrial strategy that rests on a single combination of external conditions is a bet rather than a strategy. The redundancy pillar covers more than suppliers and inventories, because it also covers scenarios. A decarbonisation plan with no workable alternative hands the company’s timetable to decision makers it does not control.
The eight pillars applied to industry
The invulnerable architecture rests on eight distinct and interdependent pillars, set out in detail in the overview of the eight pillars. What follows does not develop them again. It indicates the precise form each one takes in an industrial context.
The industrial translation of each pillar
Industry is particularly exposed on three of them, governance, redundancy and unlearning, for structural reasons: long investment cycles, assets tied up for decades and accumulated know-how that constitutes an advantage and an obstacle at the same time.
| Pillar | Its industrial form | The symptom of its absence |
|---|---|---|
| Long term governance | Protection of multi-year commitments against the quarterly horizons of the markets | Decisions about industrial capacity wait for results to deteriorate |
| Organisational identity | Coherence between closures and investments within a readable trajectory | Trade-offs that look contradictory even internally |
| Curiosity for weak signals | Picking up model ruptures, kept distinct from operational measurement | A data culture that is excellent and blind to ruptures |
| Business model reinvention | Questioning the model while the current plant is still profitable | A transformation started once the market has already imposed it |
| Structural responsiveness | Tested scenarios and pre-approved delegations ahead of a supply rupture | Continuity plans that are written and never rehearsed |
| Internal dissent | Technical alerts travelling upwards despite the hierarchical culture of production | Engineers’ signals absorbed before they reach the decisions |
| Strategic redundancy | Inventories, alternative suppliers and alternative regulatory scenarios | A strategy that depends on a single combination of conditions |
| The capacity to unlearn | Letting go of deep know-how that has become an obstacle to new architectures | Long-standing skills defended until circumstances force the issue |
What a complete architecture produces
Schneider Electric offers the best documented French illustration. In 2024, the group reported a record net profit of 4.27 billion euros on revenue of 38 billion, driven by organic growth of 10.2% in energy management, while its industrial automation business fell by 8.3%.
That asymmetry did not come from a recent change of direction. It is the consequence of a repositioning carried out gradually over a decade, before the energy transition became a regulatory or market requirement. In 2025, revenue reached 40 billion euros for the first time, with organic growth of 9%.
What this case demonstrates has as much to do with unlearning as with governance. The group gradually abandoned what it meant to be a manufacturer of electrical equipment in order to relearn what it means to be a player in energy efficiency. Those two definitions are not identical, and moving from one to the other required giving up certainties that had built its success.
Locate your organisation before you commit to anything
Do you want to know which of these eight pillars is giving way in your industrial organisation? Take the invulnerability diagnostic for your company, and compare your answers with those of your executive committee before you discuss them together.
Where an industrial leader starts
The answer to the original question is yes, provided you set aside the definition of the problem that is usually used and the tools that are usually deployed. What remains is knowing where to start, and the order matters.
The foundations before the mechanisms
Governance comes first, because it governs the capacity to hold every other decision over time. Identity comes next, because it provides the compass for the difficult trade-offs, and industry produces more of those than most sectors.
The other six pillars are built in that order, because each one rests on those that precede it. A continuity plan in an organisation whose governance systematically sacrifices the long term will be written, approved and never rehearsed.
Three signals that indicate the entry point
These three signals can be checked in an hour of committee time and they identify the priority workstream without any theoretical debate.
- Was your last decision about industrial capacity anticipated from a position of strength, or triggered by a deterioration in results?
- Do your multi-year commitments rest on a single combination of regulatory and market conditions, or on several workable scenarios?
- How long did your last significant technical alert take to reach the executive committee, and why did it not travel faster?
The complete construction sequence is set out in the implementation of the invulnerable architecture.
How I can help you in an industrial context
If you run an industrial organisation and the questions raised here touch you directly, three formats answer three different moments in this construction.
An architecture diagnostic
The diagnostic assesses the state of your organisation on each of the eight pillars, identifies the most critical structural fragilities and produces a map of what holds and what exposes you. It is the most understated starting point, because it works on what exists rather than on what would have to be created.
A keynote calibrated for an industrial audience
The keynote is built from cases documented in your sector or in sectors with comparable dynamics. Two hours that create the conditions for a strategic conversation which many organisations put off for want of a space to hold it in.
The content is set out on the page for the keynote on the invulnerable company.
An architectural construction seminar
The seminar brings your executive committee together for a day to ask the fundamental questions collectively: what is our real identity, which mechanisms protect our long-horizon decisions, which certainties do we need to unlearn, and which pillar do we start with. It produces a roadmap over eighteen months with priorities and follow-up mechanisms.
In person or remotely, these sessions are designed so that you leave with decisions to make rather than with ideas to explore. Book a session to discuss it.
Conclusion
The real question an industrial leader should ask goes beyond whether their sector will face major disruption in the next ten years. It will.
The question is whether their organisation will be the author of its response to those disruptions, or the subject of them. That question has only an architectural answer, which is to build, in the calm before the crisis, the foundations that will make those disruptions survivable.
Industrial invulnerability is not immunity from reality. It is the capacity to absorb, to rearchitect and to keep creating value while the scenery shifts.
Michelin is closing plants in France. ArcelorMittal is accumulating delays on its decarbonisation. Neither of these organisations is invulnerable in the absolute sense, and that is not what the term means. Industrial invulnerability describes the capacity to correct a trajectory, to modify a structure and to stop a local fragility from becoming an existential crisis.
Frequently asked questions about industrial invulnerability
Can French industry become invulnerable?
Yes, one organisation at a time rather than by sector decree. Industrial invulnerability is an architectural property that is built company by company. A solid sector is the sum of solid companies, never the reverse.
Does a solid architecture compensate for the Asian cost differential?
No, and claiming otherwise discredits the whole approach. Production costs that have doubled and electricity four times more expensive than in China are not offset by governance. The architecture changes the way you respond to them and the coherence of the trajectory while you respond.
Is digital transformation enough to make a plant solid?
No. Digital twins and predictive maintenance improve efficiency and operational visibility. They address neither the governance of long-horizon decisions, nor identity, nor the capacity to unlearn, which are the decisive dimensions during a rupture in the model.
Has just-in-time become a handicap?
It is when it eliminates every redundancy in the name of efficiency. The model implicitly assumes a stable environment. Toyota, which invented it, maintained buffer stock against its own principles and came through the component shortage better than its competitors.
Which pillar should a manufacturer start with?
With long term governance, because industrial investment cycles are long and the assets are tied up for decades. Without a mechanism protecting multi-year commitments against quarterly pressure, no other workstream survives a bad financial year.
