Decoding the DNA of Corporate Success and Failure
Most explanations for business success and failure contain enough truth to be convincing, and not enough truth to be useful. When we accept the headline explanation for a business outcome, we rarely ask what we are not seeing — but that is precisely where the real story and the profound learning lives.
We observe the surface. NVIDIA becomes the most valuable company on earth, and the story seems almost self-evident: they make the chips that power artificial intelligence. On the opposite side of the spectrum, Boeing’s 737 MAX crashed twice, killing 346 people, and the story is equally compact: there was a software flaw and a regulatory failure. Both explanations contain truth. Neither explanation captures thy dynamics behind. They describe the final frame of a long movie and mistake it for the plot. If we want to truly understand the DNA of success and failure, we must go deeper – beneath the surface.
I have spent years working inside organisations at the moments that define them — the strategic pivots, the leadership crises, the transformation programmes that succeed and the ones that quietly die somewhere between the boardroom and the front line. What I have come to understand is that business success and failure are never single-cause phenomena. They are the visible culmination of a long-drawn-out causal cascade — a sequence of dynamics operating simultaneously across multiple layers, most of which never make the headlines, and some of which are never examined at all.
Decoding that cascade — layer by layer – and understanding the full architecture of success or failure — this is what I mean by going beneath the surface.
What the Surface Cannot Tell You
Most leadership teams accept surface-level explanations for business outcomes. They read the analyst commentary, absorb the post-mortem, draw the obvious conclusion, and move on. This is not laziness. It is the natural human preference for simple stories over the messy, multi-layered reality.
But the cost of staying at the surface is significant. Harvard Business Review estimates that 67% of well-formulated strategies fail in execution. A McKinsey global survey found that only 27% of executives believe their organisations are genuinely good at translating strategy into action. These are not statistics about strategic incompetence. They are statistics about what happens when organisations focus on the visible layer and fail to address the invisible ones.
There are four layers of organisational reality. Understanding each one — and the way they interact — is the prerequisite for any serious diagnosis of success or failure. It also provides a blueprint for a true executive education that enables executives to transform systems, rather than just writing strategies or pleasing analysts.
The Four Layers of Organisational Reality
Layer One: The Surface — What the World Sees
The surface layer is what everyone observes. Stock prices. Market share. Earnings announcements. The headline that a company has become dominant, or the headline that it has failed. Alongside these outcomes, we typically receive a handful of explanations from analysts and commentators who offer two or three reasons why this happened.
This layer is real. These outcomes matter. But they are also radically incomplete — and the explanations offered at this level are, almost without exception, a gross simplification. The surface tells you what has happened. It tells you almost nothing about why. Staying at the surface level of analysis creates the illusion of understanding while foreclosing genuine insight.
Layer Two: Strategy — The Decisions That Shape Destinies
One layer beneath the surface, we find the sequence of strategic decisions — or the conspicuous absence of them — that set a company’s trajectory long before the outcome becomes visible.
In stable industries, short-term operational management competence can sustain performance for surprisingly long periods. But in environments characterised by technological change and continuous disruption, strategic choices become the primary determinant of long-term success. Empirical research consistently confirms what many leaders resist to admit: companies that successfully sense signals at the periphery early, are able to separate true signals from the noise, translate these signals into meaningful strategic direction, and allocate resources toward growth while systematically withdrawing resources from areas facing commoditization or extinction significantly outperform those that either have no coherent strategy or lack the capacity to execute it. And companies that are able to actively shape markets, based on a clear North Star and a long term strategy executed with conviction and tenacity outperform those who want to be everyone to everybody. Everyone wants to win – but many companies simply do not make the right strategic choices to be able to do so.
The deep challenge at this layer is not reading the future. It is building the organisation of the future while still operating the organisation of the present — moving decisively beyond the status quo before the status quo moves you. Analysing the sequence of strategic decisions a company has made, or avoided, or deferred, already reveals a great deal about why it finds itself where it does. But what determines why some organisations are able to define and execute a strategy, while others are just drifting along, or moving along?
Layer Three: The Human System — Where Strategy Lives or Dies
One layer deeper still, we encounter the network of human beings behind the decisions — and the interaction dynamics that determine whether strategy becomes reality or remains a beautifully written document.
Decisions in complex organisations are not made by the strategy. They are made by people, within power structures and hierarchical layers that have their own logic, their own interests, and their own gravitational pull. Whether a company can define and execute strategy depends far less on the brilliance of the strategy team and far more on what happens in the human layers of the leadership body. Many companies fail not because they lack strategic insight, but because alignment in the senior executive team is never genuinely achieved — with the natural consequence that pursuing a coherent strategic direction becomes impossible. Executing strategy successfully requires the redistribution of resources, which means dismantling influence in some places and building it in others. In organisations where kingdom builders can successfully defend their territory, strategic shifts either do not happen at all, or they happen far too late.
The problem compounds as it travels downward. In organisations whose leaders are driven by a need for harmony or an avoidance of conflict, change programmes are announced and not delivered locally, especially when they ask for sacrifices. When companies fail to bring leaders genuinely on board with strategic direction, programmes stall in the layers between the board and the people who must execute them. And when organisations fail to bring their people along — through clear communication of strategic direction, visible role expectations, and meaningful interaction platforms such as top-100 meetings, coffee chats and leadership development programmes — they fail to create the understanding, buy-in, and engagement that execution actually requires.
The key diagnostic question at this layer is deceptively simple: how quickly do ideas, strategies, and work actually move through the system — and what creates friction? The friction is rarely technical. It is almost always human.
Layer Four: The Inner Game — The Layer Ignored at Peril
The deepest layer of all is the one most difficult to grasp and most consistently overlooked. It is an invisible layer that directly shapes everything above it. The behaviour of every leader —every conversation they have or they avoid, every decision they make or stall, every signal they surface or ignore — originates in their inner operating system: their belief systems, their repertoire of learned responses, their level of self-awareness about their own inner operating system and behavior, and their ability to act consciously under pressure, rather than taking the easy way out.
In times of stability, the leadership behaviour of most leaders is shaped mostly by skills and values. Almost anyone can do the right thing under conditions of low pressure, provided they are skilled enough to do so. The real question is what happens under stress and pressure. Will leaders sustain the situational awareness and self-regulation required to act from principle — or will they be captured by the defensive autopilot: the self-protective mechanisms that trade what is right for what is safe, what is honest for what is comfortable, what the organisation needs for what the leader needs in that moment?
This is the layer where leadership development is most urgently needed and most commonly avoided. In times of change, leaders are asked to challenge their own belief systems, to unlearn and relearn, to rise beyond the logic of stress trigger and stress response, and then to enable others to do the same. They are asked to open themselves to genuine collaboration while remaining assertive on strategic goals — not for the sake of their own agenda, but in service of the organisation’s direction. They are asked to move into uncertainty, to execute decisions on strategies they may not fully understand themselves, and to challenge and support their people to grow rather than simply execute.
Whether leaders grow in terms of their inner operating system — developing the capacity to act from inner clarity and conscious choice under increasing amounts of complexity, uncertainty and pressure — or whether they merely perform their role, playing the game while protecting themselves, will ultimately determine whether the social system of the organisation moves forward or quietly stalls. This is the layer that is ignored by most organisations at their peril.
NVIDIA: How the Invisible Layers Created Visible Dominance
The visible story of NVIDIA is almost too obvious now. A company once associated primarily with gaming chips became the central infrastructure company of the artificial intelligence era — its chips powering large language models, data centres, robotics, and the new industrial architecture of intelligence itself.
But this surface story obscures a fascinating much deeper story that began a long time ago. NVIDIA did not simply benefit from artificial intelligence. It spent decades building the architecture that enabled it to take a central place when artificial intelligence finally arrived — and that preparation nearly destroyed the company first.
The Strategy Layer: Building Before the Market Existed
In 2006, NVIDIA was already a leading manufacturer of graphics chips, primarily used for gaming — but its vision extended well beyond that narrow domain. NVIDIA had identified a fundamental architectural difference between two types of processors. The CPU — the Central Processing Unit, the general-purpose brain of any computer — operates in a largely serial manner, working through one calculation at a time, with extraordinary speed and flexibility. The GPU — the Graphics Processing Unit, originally designed to render images and video — operates in a radically different way: it runs thousands of smaller calculations simultaneously, in parallel. That parallel architecture, NVIDIA realised, was not only useful for graphics. It was potentially transformative for any type of work that requires massive simultaneous computation — scientific simulation, financial modelling, weather forecasting, and eventually artificial intelligence. The obstacle was that GPUs could not yet be programmed directly for these purposes. In 2006, NVIDIA set out to change that, launching CUDA — Compute Unified Device Architecture — a software platform that allowed developers, for the first time, to programme GPUs directly for general-purpose computing tasks rather than graphics alone. The decisive move was to embed CUDA into every new GPU NVIDIA shipped — even though customers did not ask for it, were not yet willing to pay for it, and had no immediate use for it. NVIDIA built the infrastructure for a future its customers could not yet see. It was a quiet technical decision with consequences that would take a decade to become fully visible. The market’s response was silence. Jensen Huang has described the moment vividly: “Nobody wanted it. Nobody asked for it.” The consequences were immediate and severe. CUDA increased NVIDIA’s costs by approximately 50% at a time when the company carried roughly 35% gross margins. Its market capitalisation fell from approximately $8–12 billion to between $1.5 and $3 billion. By any conventional metric, the launch was a disaster.
What CUDA actually represented was much more profound than a mere technical and seemingly irrational decision: it was a strategic identity transformation. NVIDIA was no longer asking how to make better graphics chips. It was asking how to make the GPU programmable for every field that could benefit from parallel computing — scientific simulation, molecular modelling, finance, weather systems, computer vision, and eventually generative artificial intelligence. Huang has described how he recognised, after the CUDA launch, that it was going to add enormous cost to the business — but it was something he and his team believed in deeply, even as the market cap collapsed to around 1.5 billion dollars. This was not blind faith. Huang could foresee the emerging future, even if its precise contours were not yet clear. Nevertheless, he followed through on his convictions and persuaded his engineers, his partners, and his board to follow him — accepting a significant and prolonged profitability dip in service of a platform the world did not yet know it needed. This is what distinguishes genuine strategic leadership from competent administration: the willingness to invest in a future the market has not yet validated, and to hold that conviction through years of apparent failure.
The Human Layer: Co-Evolution as Competitive Advantage
NVIDIA’s advantage at the human layer was not that it had more brilliant engineers than its competitors. Many companies have brilliant engineers. The deeper advantage was the unique culture and operating model the company had created over time, which Huang calls extreme co-design: the disciplined, high-speed co-creation of hardware, software, networking, systems, manufacturing, and ecosystem partnerships as one integrated whole.
A modern AI system only works if every layer — GPU, CPU, memory, networking, power, software libraries, model architecture — evolves in concert. If one layer moves faster than the others, the whole system bottlenecks. NVIDIA deeply understood this, and it understood that if you want to evolve all of the elements in sync at high speed, you must create a culture of radical collaboration. A single kingdom builder would destroy the capability of the organisation to achieve its mission. Therefore, NVIDIA very deliberately built an organisation that is designed for co-creation and collaboration at scale. Huang deliberately avoids classic one-on-one management rituals. Problems are brought into group settings where people from different domains — CPUs, GPUs, algorithms, design — are expected to contribute across boundaries. This social architecture builds the kind of shared situational awareness that Amy Edmondson’s foundational research identifies as the bedrock of high-performing teams: a climate in which people will speak up, challenge assumptions, and surface uncomfortable signals without fear of retribution. In Stanley McChrystal’s Team of Teams framework, the common understanding and culture of radical transparency across functions that NVIDIA cultivates is what he calls shared consciousness — the condition that allows a complex organisation to act with the coherence and speed of a small, unified team. NVIDIA understood that shared consciousness does not emerge by itself – it must be built deliberately through the types of interactions described above.
The ecosystem dimension and the philosophy of extreme co-creation even extends beyond NVIDIA’s organisational boundaries. When the shift to full rack-scale AI systems required partner companies to invest ahead of certainty, Huang personally flew to suppliers, explained the strategic direction, and asked for major capital commitments. They trusted his reasoning enough to commit. This is not supplier management. This is ecosystem leadership — the capacity to build shared consciousness and conviction across an entire value system before the market makes the bet feel safe.
The Inner Game Layer: Identity That Evolves Rather Than Defends
At the deepest layer, the NVIDIA story is ultimately not a story about a company. It is a story about the inner operating system of one leader — and his ability to grow others around him.
Huang’s leadership demonstrates what I would call non-defensive conviction: the capacity to hold a long-term conviction under conditions of maximum external pressure without retreating into self-protection. CUDA consumed resources before the market valued them. It confused customers who did not yet understand it. It collapsed the company’s market capitalisation and generated every signal that the defensive autopilot most readily responds to: pull back, stay close to the core, do not risk what you have already built. Most leaders, under that kind of pressure, would have done exactly that — not out of weakness, but because the self-protective autopilot does not experience retreat as fear. It creates a convenient story and disguises fear-based retreat as pragmatism and flexibility.
Huang did not retreat. And what made that possible was not simply strategic clarity — it was a particular quality of inner freedom. He has stated that his mind does not begin with “what I am not” but with “what do we need to be.” That single sentence reveals an inner operating system that is oriented toward the future rather than anchored to the past — one that draws its identity from what it is becoming rather than from what it has already achieved. Most leaders — and most organisations — ask what they are, especially when there is already a great deal to protect. NVIDIA, even at its most vulnerable, asked what it must become.
But Huang’s inner game extended beyond his own psychology. The deeper leadership challenge was not holding his own conviction. It was growing others to the point where they could hold it with him. An organisation of 5,000 people cannot sustain a multi-year bet against market pressure through the conviction of one person alone. Huang had to develop leaders around him who were equally capable of operating from clarity rather than fear — who could withstand the pressure without falling back into defensive, self-protective behaviour. That is the inner game at organisational scale: not merely having an evolved inner operating system yourself, but creating the conditions in which others can develop theirs.
The result was an identity that evolved — from gaming company, to GPU company, to accelerated computing company, to AI infrastructure company — across two decades of continuous reinvention. That kind of evolution is not a strategic achievement. It is a psychological one. A defensive organisation, led by leaders whose autopilots are running on self-protection, preserves its current identity at the cost of its future. A learning organisation, led by people who are doing the inner work, allows the future to pull it towards a new identity. NVIDIA became the latter — not only because its strategy was brilliant, but because the people who carried that strategy had developed the inner architecture to sustain it under pressure.
Boeing: How the Invisible Layers Created Catastrophe
The Boeing 737 MAX case, well-documented through a congressional investigation, shows the same cascade operating in reverse — with dramatic consequences. Two crashes — Lion Air Flight 610 in October 2018 and Ethiopian Airlines Flight 302 in March 2019 — killed 346 people, triggered a roughly 20-month global grounding that became the longest grounding of a U.S. airliner, and cost Boeing more than $20 billion in fines, compensation, legal fees, customer payments and related direct costs, while inflicting reputational damage that continues to haunt the company to this day.
The Strategy Layer: When Commercial Logic Begins to Shape Technical Reality
At the strategy layer, Boeing faced intense competitive pressure after Airbus launched the A320neo. Boeing therefore chose to remodel the existing 737 NG rather than pursue a clean-sheet aircraft — a decision that saved time, resources and costs. From a commercial perspective, the logic was clear: a derivative aircraft could be brought to market faster and would preserve continuity for airlines already operating the 737. The commercial promise was powerful: better fuel efficiency without forcing airlines to absorb costly simulator retraining.
That promise, however, became more than a marketing advantage. The Congressional investigation found that avoiding simulator-based training was one of Boeing’s key goals for the 737 MAX program, and that this goal shaped design, training and risk communication decisions. Boeing had also agreed to a $1 million-per-aircraft rebate clause with Southwest Airlines if MAX pilots were required to undergo simulator training — creating potential financial exposure that could have reached hundreds of millions of dollars, depending on the number of aircraft covered. A commercial constraint had begun to shape technical reality.
The danger was not incompetence. It was the gradual dominance of a single commercial constraint that became increasingly difficult to challenge. The goal was no longer only to build a competitive and safe aircraft. It was also to protect a commercial promise that had become central to the MAX business case. When safety, speed, cost, competitive response and the no-simulator-training objective came into tension, the system repeatedly resolved that tension in ways that protected the program’s commercial goals: Describing MCAS — a new function within the existing Speed Trim System, capable of repeatedly commanding nose-down stabilizer movement without pilot input — as a minor extension of an existing system rather than treating it as a materially new pilot-facing risk; removing references to it from pilot-facing manuals and training materials; and managing information flows to regulators in ways that reduced scrutiny and protected the no-simulator-training objective Boeing had committed to avoid.
The Human Layer: What Happens When Reality Gets Filtered
At the human layer, pressure began to distort communication, engineering judgment and regulatory transparency in ways that would eventually prove fatal. The Congressional investigation described production pressure, countdown clocks and a constant organisational signal that schedule mattered. An internal Boeing survey, conducted in November 2016 during the MAX certification period, found that 39% of respondents perceived potential undue pressure and 29% were concerned about consequences if they reported potential undue pressure.
Here is the human-system question in its starkest form: what happens inside an engineering organisation when schedule, cost and customer commitments become louder than uncomfortable technical signals? The answer, in the Boeing case, appears to be that reality got filtered. People solved for what the system rewarded. They minimised what would slow down the program. They framed concerns in ways that did not threaten the dominant objective. The organisation became, in the precise language of the Congressional report, marked by a “culture of concealment” — withholding crucial information from the FAA, from airline customers and from pilots.
This was not merely a compliance failure. It was a psychological-safety failure of the most consequential kind. Amy Edmondson’s research is directly relevant here: teams learn more effectively and perform at a higher level when people can take interpersonal risks, raise concerns and speak openly about errors. In the 737 MAX case, the investigation shows what happens when a high-stakes technical system loses the ability to surface and metabolise uncomfortable truth.
The Inner Game Layer: the Defensive Autopilot at Organisational Level
At the innermost layer, the tragedy was not a lack of intelligence. Boeing was full of brilliant engineers carrying a legendary heritage of safety and technical precision. The deeper issue was the distortion of judgment under systemic pressure. The Congressional investigation described “clear resistance” by senior Boeing officials to acknowledging technical and managerial mistakes that, in the Committee’s view, had become obvious.
Under threat, individuals tend to defend themselves. Under threat, organisations defend themselves too. They protect the narrative. They protect the schedule. They protect the commercial promise. They protect the idea that nothing fundamental is wrong. The defensive autopilot rarely experiences itself as denial. It experiences itself as pragmatism. Inside the system, the rationalisations may have sounded entirely reasonable: this is manageable; pilots can handle it; we do not need to make this bigger than it is; changing course now would cost too much. That is what makes cultures governed by self-defensive autopilots so quietly dangerous. They do not collapse through one dramatic act of bad faith. They erode — one convenient rationalisation, one filtered signal, one avoided conversation at a time.
Years after the crashes, an independent FAA expert panel still found significant disconnects between Boeing senior management and other parts of the organisation on safety culture, ongoing concerns about open communication and non-retaliation, and inadequate or confusing implementation of key safety-culture components. Defensive cultures do not heal quickly or easily. They rebuild slowly, one honest conversation at a time.
Understanding the Hidden DNA of Success and Failure
These two cases work powerfully together precisely because they are not simplistic good-company and bad-company stories. They are pressure stories.
Both NVIDIA and Boeing operated in high-stakes technical environments. Both faced intense market pressure. Both had brilliant engineers. Both made strategic bets under uncertainty. Both depended on complex systems that no single leader could personally control. The difference was not raw capability. It was how each organisation metabolised pressure.
NVIDIA used pressure to sharpen a long-term vision and stick to it under pressure. The long and uncertain payoff period around CUDA did not produce a simple retreat to the core. It produced deeper integration, stronger ecosystem commitments and a clearer sense of what the company was becoming. Jensen Huang has described NVIDIA’s work as “extreme co-design”: optimising across software, chips, systems, algorithms and applications, with problems attacked collectively rather than through isolated one-on-one management. Pressure reinforced shared consciousness and strategic coherence.
Boeing allowed pressure to narrow the frame and compromise the integrity of decision-making. Competing with Airbus, preserving schedule, controlling cost and avoiding simulator training became dominant constraints that travelled downward into technical choices and outward into communication with regulators, customers and pilots. Pressure did not reveal reality. It concealed it — one small rationalisation, one filtered signal, one avoided conversation at a time.
In NVIDIA, the invisible layers created strategic coherence. In Boeing, the invisible layers created catastrophic disconnection. The surface looked entirely different in each case. But the causal mechanism — the cascade from inner operating system, through human dynamics, through strategy execution, to visible outcome — follows the same logic: a virtuous cycle in one case, a vicious cycle in the other.
What This Means for Leaders
I strongly believe that lasting organisational success can only be created when leaders are capable of understanding and influencing all four layers simultaneously: the surface outcomes they are accountable for, the sequence of strategic choices that shape the trajectory, the human system that carries or kills environmental awareness, goal setting and execution, and the inner operating systems of the leaders who create the framework and set the tone for everyone beneath them. Interventions at one layer that ignore the other layers produce, at best, temporary results and, at worst, new problems that compound the original ones.
At the strategy layer, the practical implication is demanding. Leaders must be willing to ask not what their organisation already is, but what it must become — and to hold that question seriously enough to invest where the future is built within their organisation before the market forces the answer. At the human layer, the evidence is unambiguous: execution fails most reliably not through strategic errors, but through leadership misalignment, political friction, and the inability of uncomfortable truths to travel cleanly through the system. The key question is not whether your strategy is brilliant. It is whether the human system beneath it is aware and capable of delivering it.
At the inner game layer, Buddhist philosophy offers a concept that is strikingly precise in its organisational relevance: non-attachment — not indifference, but the capacity to release grip on legacy mental models and beliefs, inherited identities, reactive emotions, and comfortable assumptions when reality demands evolution. The cognitive dissonance between personal belief systems and necessary corporate pivots must be named and worked through, not managed away. And every leader must learn to distinguish between performing leadership and actually executing strategy courageously — because the mere projection of competence, the empty language of transformation, and the hollow rituals of alignment can mask a system that is quietly losing touch with reality.
Questions Worth Sitting With
- What layer of organisational reality do you spend most of your time and energy on — and which layer do you habitually avoid?
- Are you asking what your organisation is now, or what it must become?
- Can uncomfortable truth travel cleanly through your organisation, or does it get filtered, softened, and managed before it reaches the people who most need to hear it?
- Where might the defensive autopilot be operating in yourself or in your team — and how might you surface it to enable your leaders to rise beyond it?
The surface will always capture attention. The headlines will always seem self-explanatory. But lasting success and lasting failure are both authored in the invisible layers beneath — in the cascade of strategy, human dynamics, and inner operating systems of its people.
NVIDIA held its nerve through near-destruction and built the future it had imagined before the market could see it. Boeing allowed commercial logic to colonise technical reality, one filtered signal at a time, until the system could no longer tell itself the truth. Both outcomes were authored long before they became visible. Both were built, layer by layer, in the dynamics and choices that the surface never shows.
Going beneath the surface is not optional for leaders who wish to build organisations that endure. It is the work itself.
For every complex problem there is an answer that is clear, simple, and wrong.
H.L. MENCKEN
