The Narrowed Corridor: How Harmony, Homogeneity and the Need for Consensus Can Threaten a Company's Future Readiness

Why leadership bodies need a higher tolerance for variance in times of disruption and crisis, and how regulation, paternalism and AI narrow the corridor even further

The Hypothesis

In a Western world that is losing its stable, rule-based order, where decades of economic growth and future prosperity are at risk, one of the biggest dangers for companies appears to be the homogeneity of staff and executive selection, stripped of any edges. For decades, leadership has relied so heavily on fit, consensus and risk avoidance that the capacity for justified deviation, contrarian thinkers and people who diverge from the norm, has in many cases simply been selected out. Yet that exact capacity could help a company navigate a crisis today, because nobody in the middle of the stream stands out.

Three further forces reinforce this narrowing: excessive regulation, which redirects leadership energy from shaping the business towards protecting it; technocratic and rule-driven leadership principles, where fear of conflict leads organisations to govern through regulations rather than leadership; and carelessly integrated artificial intelligence, which places the probable above the outstanding.

Force 1: Staff and Executive Selection

Nomination committees, executive search consultants and supervisory bodies increasingly select against the same criteria: cultural fit, the closest possible match to a profile, often AI-assisted, consensus ability, reputational safety, avoidance of exposure. The result is a narrow corridor of accepted profiles that may well exclude the most valuable diamonds. In stable growth phases, this is efficient, because whoever optimises wins. In times of disruption, it becomes a risk, because homogeneity reduces the range of thinking and behaviour patterns precisely when that range is needed most. It is of course comfortable to work with as little friction as possible. But can you win a competitive race without that inner friction?

Historical observation supports the suspicion. Many of the most defining companies were founded and led by people who would struggle to pass almost any nomination process today: school dropouts, eccentrics, single-minded obsessives. Werner von Siemens, Gottlieb Daimler, Carl Benz, Porsche, Otto, Jobs, Musk, Bezos or Huang were, and are, personalities who would barely be selected on profile today, though to be fair, we only remember the winners and have long forgotten the equally numerous characters of the same type who failed. The common denominator is not conformity. It is independent judgement, high stamina and resilience, and the willingness to be right against the consensus, with the risk of also being wrong; that comes with the territory. They stood against the wind and stood out because of it.

I remember an employee I hired more than twenty years ago: an unremarkable application with gaps, the cover letter handwritten with visible effort on lined notepaper and hard to read. An application like that would never make it into any process today. I invited the candidate anyway, precisely because that application did not spark curiosity, and I found a gem of a mathematically gifted software developer. Who still has that courage today?

Force 2: Excessive Regulation

Regulation acts as a second, independent amplifier of the narrowing. Where every decision must be documented, reviewed and hedged against hypothetical risks, the sought-after leadership profile shifts systematically: away from the builder, towards the compliance administrator. Over the past twenty years, I have encountered a number of companies where I wondered whether compliance was actually running the business. This approach rewards error and risk avoidance, which then dictates day-to-day operations. But does that make a company future-ready?

The scale of excessive regulation matters. The Draghi report on European competitiveness counts around 13,000 new EU legal acts between 2019 and 2024, against roughly 5,500 in the US. More than 60 per cent of EU companies and 55 per cent of SMEs name regulation and administrative burden as the biggest obstacle to investment. And roughly 30 per cent of European unicorns relocated their headquarters abroad between 2008 and 2021, mostly to the US. Switzerland has benefited from this too: European companies move to Switzerland despite higher wage costs, because the administrative cost of regulation is lower and productivity correspondingly higher.

The point is not that rules as such do harm. Rule of law, data protection and market trust are locational advantages, particularly for Switzerland. The point is the dose and the direction of the effect. Overregulation acts like a tax on unconventional action. It hits the new, the not-yet-proven, disproportionately hard, while sparing the established. In doing so, it reinforces exactly the homogeneity that becomes dangerous at a turning point in time.

Force 3: Technocratic Leadership

Leading an organisation is rarely simple and all too often a balancing act between guardrails and freedom. Adjusting those guardrails, and enforcing them where necessary, can take real effort. In my experience, though, the vast majority stick to the defined rules. Yet whole volumes of internal regulations often get written because of a handful of individual cases, in an attempt to govern every single one. Instead of leading the people actually involved more closely and directly, organisations regulate. Take Netflix’s expense policy as an example: “Act in Netflix’s best interests.” That is really all it takes.

The mechanism is insidious because it looks reasonable. Every incident produces a new rule, every rule stays in place, and over time the organisation optimises itself around the exception. The large majority, who behave responsibly anyway, carry the burden of controls that were designed for a small minority. Regulations are cheap to write and expensive to live with.

The cultural cost is more serious still. Anyone who can point to a regulation no longer has to exercise their own judgement. Responsibility shifts from the person to the document: “I followed the rule.” That is exactly what weakens the judgement that matters in unclear situations. At its core, technocratic leadership is often conflict avoidance dressed up as governance. It is more comfortable to write a policy than to confront a person.

And this closes the loop back to the first force. An excess of internal regulation signals distrust. It demotivates the self-directed, independently minded people organisations actually want to keep, and it attracts the pure rule-followers instead. Paternalism thereby reinforces the homogeneity of selection. But in times of disruption, which is what we are living through now, the decisive factor cannot be written down in advance: you cannot draft a rule for a situation that does not yet exist. What remains is judgement, principles over procedures, and the courage to have the direct conversation.

Force 4: Carelessly Integrated AI

Alongside these human factors comes a technical dimension: AI. With it, we bring a fourth, more subtle narrowing into the house. At its core, a language model is a statistical compression of past data and events. It answers with the most probable response, which is not necessarily the best or the outstanding one. By construction, it is therefore a corridor tool: it pulls towards the middle of the known. Yet success often lies in the unknown, outside the comfort zone.

This is made worse by how we try to tame it. To avoid hallucinations, we lower the temperature, meaning the spread of possible answers, and layer rulebooks, guardrails and approval processes on top. Both are operationally sensible. But both compress the output into a narrow band that again amounts to an average. With the best of intentions, we are building ourselves a “consensus machine” at industrial scale. The same force that protects us from errors also strips out productive deviation.

This is not an argument against AI. It is an argument against using it naively. The same technology can achieve the opposite when we deliberately deploy it as a divergence engine: to generate unfamiliar options, as devil’s advocate, for adversarial critique of our own strategy, with high rather than low variance where we are searching for ideas, and with human judgement at the point where decisions are actually made. The decisive question is not whether to use AI. It is what for: do we use it to generate deviation, or to compute it away?

The Behavioural Pattern: the Titanic

The pattern sharpens against a familiar image. The Titanic probably did not sink because the bridge failed to see the iceberg. It sank, in all likelihood, because of the way it chose to steer around the danger. The evasive manoeuvre turned the ship so that the iceberg tore open its flank and flooded several compartments at once. It is said that a head-on impact might well have been survivable. Avoiding the visible risk created the fatal situation.

Organisations optimised for two decades on fit, consensus and risk avoidance behave the same way. They swerve around the visible obstacle and, in doing so, expose their own flank: the blind spot, the question never asked, the bet nobody dared to back. And as the water rises, the band plays on. The rituals of normality, reporting, committee meetings, culture programmes, carry on in orderly fashion, precisely because they offer stability while the substance underneath is failing. This is how organisations cling to past patterns of success, even when those patterns genuinely deserve to be questioned. Whether sectors such as trust and fiduciary services, wealth management, banking or IT services will still operate the same way in five years is a fair thing to doubt.

Avoiding risk is itself a risk

The uncomfortable insight: avoiding risk is itself a risk. Whoever only swerves is not choosing between danger and safety. They are choosing between two dangers, and often the one they understand less well. A body trained exclusively to avoid mistakes will fail to recognise the one mistake that actually matters, because it does not appear on its checklist.

What This Means for the Board and the Executive Team

The answer is to build a degree of tolerance for variance into the leadership architecture, without breaching the duty of care. Dissent needs to be institutionalised, not left to individuals. Eight concrete starting points:

• Deliberately seek cognitive heterogeneity on the board. Productive friction, not fit, as a selection criterion. A board where everyone agrees is a cost centre, not a control body.
• Build dissent into the structure. A standing devil’s advocate, red-teaming and war-gaming for major decisions. Whoever argues the opposing position does so with a mandate, not at their own risk.
• Think in options and portfolios, not point forecasts. Deliberately allow a few asymmetric bets whose downside is limited and whose upside is large, ring-fenced from the core business both operationally and financially.
• Decentralise decision rights. Let variance talent operate where innovation happens, or where a completely different perspective can generate new momentum. Proximity to market and technology combined with innovation beats central control. And yes: it does no harm to bring someone with a genuinely different profile and skill set onto the executive team or the board either.
• Deliberately use AI as a divergence engine, not a consensus machine. Default settings, low temperature and tight guardrails must not quietly standardise strategic thinking. Where the goal is options, variance should be increased, not reduced.
• Treat regulatory burden as an active leadership issue. Measure, bundle and cap compliance effort, so leadership energy stays available for shaping the business instead of quietly draining into protection.
• At operational level, lead directly instead of regulating technocratically. Strip out unnecessary rules where possible and have the uncomfortable conversation with employees, rather than regulating the entire organisation because of a single incident.
• Build future readiness. Do we have key people, in leadership and across the organisation, who pursue and defend their goals with passion, outside comfort zones, regulations and narrowed opinion corridors?

Closing Thought

A narrowed corridor with low variance in selection, regulation, technocratic leadership and AI all pull in the same direction: towards the middle, towards the average, towards the proven. Each individual force is reasonable on its own. Together, they create a single current in an ever-narrowing channel, one that holds up in calm times and turns into a trap in times of disruption or crisis. Leadership’s task is to build a system that can once again tolerate independent thinking, productive argument and asymmetric bets.

Guiding question for discussion: Where in our company are we, consciously or unconsciously, selecting away justified deviation, in hiring, in compliance, in internal regulation, or in our AI tools? And which single measure from the list above will we test in the next ninety days?