The venture power law is the most accepted governing constraint in category creation investing. But accepted observations and governing causes are not the same thing. The power law may not describe the fundamental law of the domain. It may describe what happens when a field operates probabilistically before its deeper causal structure has fully surfaced.
Consider a company that, for three years, appeared to be winning. Revenue doubled two years in a row. Enterprise logos accumulated. Top-tier funds competed to enter subsequent rounds. Every metric confirmed the thesis. Institutional endorsement was arriving faster than the founding team could manage it.
Below the visible layer, a different process was operating.
The early deployments had been conducted inside organisations with dedicated innovation budgets, sympathetic executive sponsors, and pricing structures that would not survive normal procurement scrutiny. The founding team was present in every significant implementation. Customers had been selected for their receptivity to the new approach.
When the company scaled beyond those conditions — when ordinary procurement timelines replaced managed pilots, when unselected buyers replaced curated early adopters, when the founding team could no longer be present in every deployment — the demonstrated superiority did not reproduce. Not because the underlying capability was fraudulent. Because the conditions that had made it visible were not the conditions the full market presented.
By the time this became legible in the metrics, the window for meaningful structural correction had closed. The company had scaled a demonstration, not a transition.
This is the pattern the venture power law ultimately describes. Not the natural rarity of great companies. The recurring consequence of investing in demonstrations before they can be distinguished from transitions.
For most of its history, astronomy was not a science in any meaningful sense. It was an organised observation practice. Astronomers classified stars, tracked planetary motion, built increasingly precise correlations between celestial positions and predictable events. The correlations were real. They were practically useful. They could predict eclipses. They could guide navigation. But they could not explain why the planets moved as they did.
Newton changed this. Not by accumulating better data. By hypothesising a cause. Assuming gravitational attraction as the underlying mechanism, he could derive the planetary observations as necessary consequences rather than observed patterns. Three of Kepler's correlations were explained for the first time. Eight others were exposed as coincidences that had not been thoroughly checked.
The transition was not from wrong to right. It was from correlation to causation. From a field that organised observations to a field that understood the mechanism producing them.
Venture capital and category creation investing are at an earlier stage of a similar transition. The power law is real. It is consistent across geographies, fund vintages, and investment strategies. The industry has built its entire operating model around it. But the power law is an observation. It is not a causal account.
The Three Stages of Every Science
Eliyahu Goldratt, whose Theory of Constraints brought causal thinking to operations management, observed that every mature science passes through three distinct stages. The pattern is consistent across physics, chemistry, biology, and medicine — and it is diagnostic for understanding where any field currently stands.
Classification
Recurring phenomena are identified, named, and organised. Common terminology emerges. Classification is genuinely useful — it creates shared language and provides the first systematic view of a domain. But classification describes what exists, not why it exists. The question why is not yet available.
Correlation
Recurring relationships between phenomena are discovered and mapped. Predictive patterns emerge. Correlations enable forecasting and generate practical interventions. But they remain haunted by a fundamental limitation: they cannot distinguish genuine causal relationships from coincidences that have not yet been thoroughly examined.
Effect-Cause-Effect
A causal mechanism is hypothesised. Specific consequences follow necessarily from the mechanism. Those consequences are tested against observable reality. This is the stage at which a field becomes a science — not because it achieves certainty, but because it establishes the logical structure within which claims can be evaluated, corrected, and built upon.
The progression is not inevitable. Fields can remain at the classification and correlation stages indefinitely. Medical practice spent millennia classifying diseases by symptom and discovering correlations between treatments and outcomes — Jenner's smallpox immunisation worked for seventy years before germ theory arrived to explain why. The correlations were real. The causal mechanism was absent. And the absence mattered: without knowing why the immunisation worked, practitioners could not extend the insight to other diseases.
Pasteur's germ theory changed this. One assumed cause explained the existing correlations as necessary consequences and opened the possibility of creating immunisations rather than merely discovering them. The domain moved from managing observed phenomena to understanding the mechanism that produced them.
Where Category Creation Investing Currently Stands
Category creation investing has become genuinely sophisticated within both stages. Classification is advanced. The domain has developed precise terminology for founder archetypes, market timing, competitive dynamics, network effects, product-market fit, go-to-market strategy, and portfolio construction. Correlation is increasingly refined — pattern recognition across successful and unsuccessful category creation attempts has produced increasingly powerful heuristics.
But the central question — why do some category creation attempts successfully establish enduring new market organisations while others fail despite similar apparent conditions — remains primarily addressed through correlation rather than causal explanation.
The practical consequence is a domain operating probabilistically. Because the causal structure of category creation success remains incompletely surfaced, investors cannot reliably determine in advance which early-stage companies will achieve genuine enduring settlement. Because they cannot determine this reliably, they compensate through probabilistic diversification: many investments, extreme asymmetry, a small number of outliers.
The power law is not a cause. It is the statistical residue of operating probabilistically under structural opacity.
The nested nature of the power law confirms this. Within portfolios, a small number of companies generate the majority of returns. Across funds, a small number of firms consistently outperform. This recurring concentration at every level suggests that the mechanisms governing successful category settlement are insufficiently understood — not that extreme rarity is a fundamental and unavoidable feature of reality.
Rarity alone is not explanation. Before causal understanding surfaced in medicine, controlled flight, or semiconductor engineering, extraordinary rarity was the normal state. That rarity was real. It was not fundamental. It reflected operating before the underlying causal structure had been properly surfaced.
The Causal Hypothesis: Category Creation as Equilibrium Transition
The causal hypothesis proposed here begins from a specific observation about what category creation actually is. New categories do not emerge primarily by competing against incumbent companies. They emerge by competing against existing governing logics — the implicit rules through which entire actor populations organise how value is created, measured, and coordinated within a domain. These governing logics are not merely cognitive habits. They are structurally embedded: in incentive systems, evaluation criteria, institutional expectations, procurement processes, career structures, and measurement frameworks built around them over years or decades.
This means a genuine category transition is not primarily a product challenge or a market entry challenge. It is an equilibrium transition challenge. The new governing logic must displace an existing one that is still actively self-reproducing across every institutional layer of the domain.
This displacement creates a specific structural vulnerability — a fragile transition window — during which the new logic has demonstrated genuine superiority but has not yet become self-sustaining. During this window:
- The old governing logic remains gravitationally active — still organising what counts as rational, prudent, and safe for every actor in the domain.
- The new logic requires continuous active holding to prevent progressive reabsorption into the old equilibrium.
- Every pressure event — commercial, institutional, competitive, financial — is directionally toward the old logic, because the old logic is still ambient.
This is equilibrium gravity. It is not primarily a psychological phenomenon. It is structural. And under sustained pressure, the biological systems attempting to hold the new logic — founders, leadership teams, boards — face a specific and predictable set of failure mechanics.
The texture of equilibrium gravity is concrete. A procurement team requires integration with the incumbent system — which embeds the old governing logic in the technical architecture. A quarterly board presentation requires metrics calibrated to the old equilibrium's measurement standards — which trains the founding team to optimise for the wrong signal. An enterprise buyer with a budget built around the old solution's cost structure normalises pricing that makes the new logic appear economically similar to the old one — which eliminates the structural differentiation that justified the transition claim. None of these pressures are hostile. Each is individually reasonable. Together they are gravitationally directional — toward the old logic, from every institutional layer simultaneously.
The containerisation of global shipping illustrates the same mechanism at civilisational scale. The structural superiority of the standardised container over break-bulk shipping was demonstrably real from the mid-1950s. Yet the transition took decades longer than pure technical superiority would predict.
Port infrastructure had been built for break-bulk operations. Union agreements were structured around the labour requirements of the old method. Shipping companies had balance sheets organised around existing vessel designs. Customs and insurance procedures assumed the old handling logic. None of these forces were opposed to containerisation in principle. Each was simply optimised for the existing equilibrium. That optimisation constituted continuous gravitational pull against the new logic — not from malice but from structural embeddedness.
Category creation operates under the same mechanism compressed into a shorter timeline and higher stakes. The governing logic that a new category must displace is not entrenched by malice. It is entrenched by the same structural embeddedness that made containerisation take three decades rather than three years.
Time horizons compress. Near-term survival dominates. Local coherence becomes more urgent than long-range causal integrity. Small compromises accumulate. The narrative of the transition continues advancing — revenue grows, adoption expands, institutional endorsement arrives — while the governing logic underneath progressively loses fidelity to the structure required for genuine settlement.
The critical feature of this drift: it is invisible while it is occurring. Conventional metrics continue advancing long after the structural integrity underneath has broken. By the time the failure becomes visible in outcomes, the window for meaningful correction has often already closed.
Two Failure Modes, One Confused Solution
The causal account separates two failure modes that the correlation-based approach consistently conflates.
The overwhelming majority of current investment infrastructure addresses the first failure mode. Selection frameworks, due diligence disciplines, portfolio construction principles, pattern recognition systems — these are entry tools. They are designed to improve the quality of the initial investment decision.
The second failure mode is addressed almost nowhere systematically. Practitioners acknowledge it through concepts like founder resilience, coaching relationships, board quality, and execution discipline. These interventions operate at the level of human motivation and psychological support. They do not address the structural mechanism — equilibrium gravity — that produces drift regardless of motivation and discipline.
What Earlier Causal Visibility Changes
The causal account — if valid — changes the investor's relationship to uncertainty in a specific and important way. A causal account does not eliminate uncertainty. It changes its character. Instead of distributing probabilistic bets across many positions, the investor can ask structural questions that are answerable before outcomes accumulate:
- Is a specific structural constraint — not merely a preference or inconvenience, but a genuine operational limitation that the current governing logic cannot accommodate — actually becoming removable? Or is the company proposing to improve performance within a constraint that remains fundamentally intact?
- Does the demonstrated superiority hold inside organisations still operating under the full weight of the old governing logic — its incentive structures, evaluation criteria, institutional expectations — or only inside conditions that have been protected from that weight?
- Is the propagation of the new logic becoming self-amplifying — each new adopter making the next adoption more rational without additional central forcing — or does it require continuous investment of energy, narrative, and institutional relationship management to sustain?
These questions are structural rather than correlational. They do not ask what patterns the company resembles. They ask whether the specific causal conditions for genuine settlement are actually present. And they are answerable — through investigation of deployment conditions, customer selection methods, and adoption dynamics — before the outcomes that pattern recognition requires have accumulated.
The investor who can answer these questions operates with structural diagnostic capability rather than probabilistic selection. The portfolio they construct is not a portfolio of probability bets. It is a pipeline of structural settlement opportunities — each independently positioned in a different market, each with genuine structural conditions for enduring settlement, each progressing through its own transition on its own timeline.
The power law governs the probability-based portfolio because genuine structural positions are difficult to distinguish reliably from narrative positions at the point of investment. The structurally diagnosed portfolio is not subject to the same distributional constraint, because its selection is not probabilistic.
Multiple positions can settle independently. Multiple can return the fund. This is the first condition under which the distributional dependence created by probabilistic selection can begin to weaken.
The Holding Problem and a New Infrastructure Class
But selection alone does not close the equation. Selection is necessary. It is not sufficient. Even a portfolio constructed through genuine structural diagnosis — containing only companies with authentic settlement potential — still faces the holding problem. The fragility window operates regardless of how sound the initial structural conditions were. Equilibrium gravity does not distinguish between companies selected well and companies selected poorly. It acts on biological systems under pressure. And biological systems under pressure drift.
If the primary failure mechanism is structural drift occurring before conventional metrics reveal it, then any meaningful intervention must provide continuous visibility into governing-logic integrity — not periodic board reviews or episodic advisory relationships, but structural preservation that operates continuously through the fragility window.
This generates a specific implication that follows necessarily from the causal account: if equilibrium transitions are structurally fragile, and if biological systems cannot reliably hold governing-logic fidelity under sustained pressure, then infrastructure capable of preserving structural visibility through the fragility window would change how category creation outcomes distribute.
This describes a new class of infrastructure that does not currently exist as an explicit category: systems constitutionally oriented toward preserving structural visibility and governing-logic fidelity under the specific pressure conditions that produce drift. Not systems designed to optimise or motivate or automate. Systems designed to hold the structural thread when biological systems under pressure would naturally release it.
The investor holding a company with this infrastructure is holding a structurally different position from the investor holding a company without it. The transition window is the same. Equilibrium gravity is the same. What is different is the probability that the genuine structural conditions identified at entry will survive through the fragility window to produce the settlement they were positioned for.
The Scientific Test
Any causal account must be falsifiable. The claim that the power law is a diagnostic artefact of probabilistic selection rather than a natural law of the domain is not merely an interpretive reframing. It makes specific predictions that can be tested against observable reality.
The test is not whether the causal account elegantly reinterprets past outcomes. Any sufficiently flexible account can do that. The test is whether it predicts effects that were previously unexplained, exposes structural mechanisms that were previously invisible, and produces earlier visibility into transition fragility before conventional interpretation catches up. If the account cannot do this, it weakens. If it can, it earns the right to guide intervention rather than merely describe observation.
From Correlation to Causation
The venture power law will remain observed as long as category creation investing operates primarily within the classification and correlation stages. The observation is real. The question is what it means.
The Power Law as Natural Law
Extreme outcome concentration is an unavoidable structural feature of category creation. The power law is the natural law of the domain. The right response is to build investment models calibrated to it — accumulate more pattern recognition, refine heuristics, improve portfolio construction within the existing model. This is valuable work. It improves outcomes incrementally. It does not change the fundamental distribution.
The Power Law as Diagnostic Artefact
Extreme outcome concentration is the statistical consequence of probabilistic selection operating in a domain whose causal structure is incompletely surfaced. The right response is to surface the mechanism, test it against observable reality, and build infrastructure that operates on the mechanism rather than its statistical signature. This is harder — but it produces genuine advancement rather than better correlations.
These are not merely different framings. They produce different research programmes, different investment disciplines, and different priorities. The causal programme is harder. It requires asking why rather than how, and subjecting assumptions to falsification rather than protecting them through accumulating confirmations. But it is the programme that produces genuine advancement — the ability to explain, predict, and eventually intervene in category creation transitions rather than merely observing and classifying their outcomes.
The sequence does not care about narrative, conviction, or intention. It responds to structural position.
The central scientific question is no longer whether extreme outcomes exist. It is why genuine category settlements remain so extraordinarily difficult to sustain before the new equilibrium becomes self-reproducing.
What Remains Unresolved
Intellectual honesty requires stating what the causal account does not yet establish. The hypothesis that category creation is an equilibrium transition problem — and that the failure mechanics of biological systems under sustained pressure are the primary source of holding failure — is a causal account in the Effect-Cause-Effect sense. It is not yet a validated account. The prospective tests described above have not been completed at sufficient scale to distinguish the causal account from alternative explanations.
Several structural questions remain formally open: whether the conditions for genuine settlement can be diagnosed with sufficient reliability to justify concentrated positioning without portfolio-level protection; whether infrastructure that preserves structural visibility can be designed to be genuinely constitutional rather than subject to the same social pressures that produce drift in advisory relationships; and whether the power law distribution at the fund level reflects the same mechanism as the power law distribution within portfolios, or whether additional causal factors operate at that level.
The power law as currently understood is an observation requiring explanation, not an explanation in itself. The causal account proposed here — probabilistic selection under structural opacity, compounded by holding failure from biological drift under equilibrium gravity — is more specific, more falsifiable, and more productive of intervention than treating the power law as the natural law of the domain.
That distinction is where the work begins.
This paper is part of The Continuity Series — a four-volume work examining the structural mechanics of category transition, the biological limits of holding lawful continuity under pressure, and the institutional infrastructure required for endurance beyond those limits.
The Effect-Cause-Effect framework applied here draws on Eliyahu Goldratt's account of scientific development as presented in the Theory of Constraints Journal.
The framework presented here is an emerging falsifiable systems formulation, not a completed science. Claims are bounded to what has been demonstrated across multiple case reconstructions. Several structural questions remain formally unresolved and are submitted to adversarial testing.
The governing discipline: if observable reality contradicts the causal account, reconstruction becomes mandatory — not interpretive defence.