Regenerative Business Ecosystems

A business ecosystem either renews the conditions it depends on or it consumes them. There is no third state, and almost everything built under the word is built in the second.

Future Proof Intelligence. Research. No. VIII. MMXXVI

Abstract

The word ecosystem entered management from biology in 1993 and, in the three decades since, quietly inverted. In ecology a keystone is defined by what its presence sustains in everything around it. In strategy a keystone became a firm that holds the central position and captures the network's value. This paper takes the real ecosystem strategy literature seriously, Moore on co evolution, Iansiti and Levien on keystones and dominators, Adner on the wide lens and ecosystem disruption, and reads it honestly against the ecological science it borrowed its name from and the regenerative design tradition it has not yet absorbed. We argue that the canon was more right than the practice that followed it. Iansiti and Levien named the extractive failure mode precisely, the hub landlord that pursues control of value extraction alone and leaves a starved and unstable network. The practice optimised for that mode while using the language of the opposite one. Underneath the borrowed metaphor sits a precise distinction the field stopped reading: Odum's separation of the young ecosystem that maximises throughput from the mature ecosystem that maximises renewal. Most business ecosystems are structurally young systems wearing the vocabulary of mature ones. We locate the exact point at which orchestration becomes extraction, ground it in the 2026 state of platform regulation, and set out what a genuinely regenerative ecosystem requires structurally rather than rhetorically: not values but design, drawn from Odum's mature attributes and Ostrom's principles for governing a commons that renews rather than depletes. The paper ends where regeneration actually has to be anchored, in an accountable identity layer over the orchestration layer.

1. The metaphor that inverted

1.1 A word borrowed in 1993, and what it was borrowed to mean

The vocabulary of business ecosystems has a precise origin. James F. Moore published "Predators and Prey: A New Ecology of Competition" in the Harvard Business Review in 1993, and the article did something the field has been living off ever since. It said that a company should not be understood as a member of a single industry but as part of a business ecosystem that crosses many industries, inside which firms co evolve around an innovation, cooperating and competing at the same time. The unit of competition, Moore argued, was no longer the firm. It was the ecosystem.

It is worth being exact about what Moore actually claimed, because the field kept some of it and dropped the part that matters most to this paper. He described an ecosystem as moving through stages: birth, expansion, leadership, and then a fourth stage he named self renewal or death. Read that fourth stage slowly. Moore did not present renewal as a virtue an ecosystem might choose to have. He presented it as the only alternative to ending. An ecosystem that reaches leadership has exactly two paths in front of it. It renews the conditions that allowed it to lead, or it declines and is replaced. There is no stable third state in which an ecosystem holds a dominant position indefinitely without renewing anything. The founding article of the field already contained the claim this paper exists to make. The field kept the first three stages, built an entire discipline around the third, leadership, and treated the fourth as an epilogue. This paper is, in one sense, an argument that the fourth stage was the point.

1.2 Why the borrowing was not innocent

A metaphor is never neutral. When a field borrows a word from another field it borrows a set of expectations along with it, and then it either honours those expectations or it quietly drops the ones that are inconvenient and keeps the prestige of the word. The word ecosystem carries, in its origin discipline, a very specific set of properties. An ecological system is defined by interdependence, by flows of energy and nutrients that cycle rather than run one way, by the fact that no part of it can prosper for long while degrading the conditions the rest of it depends on, and by a long literature on the difference between systems that renew themselves and systems that exhaust themselves. To call a commercial network an ecosystem is to invoke all of that. It is to claim, implicitly, that the network has the properties that make ecological systems worth admiring: resilience, diversity, the capacity to persist.

The claim is mostly not earned. Across three decades the word has been used to describe arrangements that have almost none of the properties of the systems they are named after. A platform that concentrates value capture at the centre while the participants who generate that value operate on thin and revocable terms is not, in the ecological sense, an ecosystem. It is a different kind of structure entirely, and the ecological sense of the word, the part that gives it its appeal, was precisely the part that did not transfer. This paper is not a complaint about loose language. The inversion is structural, not lexical, and tracing exactly where and how it happens is most of the analytical work. But the inversion begins at the word, and a reader who notices it at the word will recognise it everywhere after.

It is worth being precise about why a borrowed metaphor is dangerous in a way a plain false claim is not. A plain false claim can be checked. If a network calls itself the largest of its kind and is not, the claim is falsifiable and someone will falsify it. A borrowed metaphor does not work that way, because it does not assert a checkable fact. It transfers a set of associations and then lets the listener supply the conclusion. When an arrangement is called an ecosystem, no one has formally claimed that it is resilient, diverse, or self renewing. The word has merely arranged for the listener to assume those things without anyone having to defend them. This is the precise mechanism by which the inversion stays invisible for thirty years: there is no false sentence to refute, only a true sounding word doing the work a false sentence would have had to do, and doing it more durably because nothing was ever asserted that could be disproved. The whole of this paper is, in one sense, the act of converting that unstated borrowed assumption back into a checkable structural claim, so that an arrangement called an ecosystem can be tested for whether it is one rather than merely trusted because of what the word implies.

1.3 The shape of the argument

We will proceed in a deliberate order. First we read the real ecosystem strategy canon, Moore, Iansiti and Levien, Adner, honestly and without condescension, because it is better than the practice that followed it and one of its central distinctions is almost the distinction this paper needs. Then we read the ecological science the canon borrowed from and stopped consulting, in particular two papers from 1969, Paine on the keystone and Odum on the development of ecosystems, because together they contain the exact diagnostic the strategy literature lacks. Then we locate, precisely, the structural point at which orchestration becomes extraction, and ground it in the documented state of platform regulation in 2026. Then we show that the corrective vocabulary itself, the word regenerative, is being captured by the identical mechanism that captured the word ecosystem, which is why the standard that follows must be structural and not stated. Then we set out what a regenerative ecosystem requires, structurally and not as a values statement, drawing the design from Odum's mature ecosystem and from Elinor Ostrom's principles for governing a commons. Finally we show why regeneration cannot be anchored at the orchestration layer where almost all attention sits, and why it settles instead on the layer that holds still long enough to be made accountable. That last move is where the argument arrives, and it is where Future Proof has been operating the whole time.

2. The canon, read honestly

2.1 Moore: co evolution, and the stage everyone skipped

Moore's contribution was to move the unit of analysis. Before him, strategy largely asked how a firm positioned itself against rivals inside an industry whose boundaries were assumed. Moore said the boundaries were the wrong object. Value increasingly came from constellations of firms across industries that co evolved around a shared innovation, each making the others more valuable, competing for position inside an arrangement that none of them controlled alone. This was a real advance and it has held up. The smartphone, the cloud, the modern car, the payment rail: none of these is the product of a single firm and none of them is well described by industry level competition. Moore saw the shape early.

What the field did with the insight is the relevant thing. It took the third stage, leadership, and turned it into the entire game. An enormous body of subsequent work is, in effect, advice on how to reach and hold the leadership position in an ecosystem: how to become the firm around which others co evolve, how to make complementors depend on you, how to capture a durable share of the value the constellation creates. The fourth stage, self renewal or death, received far less attention because it is far less flattering. It says that leadership is not a destination. It says that an ecosystem in the leadership stage is not finished but at a fork, and that the only non terminal branch requires it to renew the very conditions that produced its position, including, uncomfortably, the prosperity of the participants it has learned to depend on and extract from. The canon's founder built the regenerative claim into the original architecture. The discipline professionalised the part before it.

There is a reason this happened, and it is not that anyone decided renewal was unimportant. It is that the leadership stage is the stage at which advice is purchasable. A firm trying to reach or hold leadership has a budget, an urgent question, and a willingness to pay for an answer, which is exactly the condition under which a body of practice grows. The self renewal stage poses a question almost no one wants to pay to have answered, because the honest answer is that the firm should accept less than the maximum it could currently extract in order to keep the system that produces the extraction alive, and that is not advice a firm at the height of its leverage is in the mood to buy. The asymmetry is not intellectual. It is commercial. The part of Moore's model that could be sold was sold, repeatedly and well, and the part that could not be sold was left as a closing paragraph. This is itself an instance of the paper's larger argument arriving early: even the literature about ecosystems was shaped by the same extractive selection pressure it describes, optimising for the stage with the immediate yield and leaving the stage that bears the long term load underdeveloped. The reader should hold that observation, because it will recur. A system that selects for immediate yield will hollow out its own account of renewal, including its scholarly account, and the hollowing will look like focus rather than loss.

2.2 Iansiti and Levien: the field's own name for the failure mode

The most precise instrument in the canon is Marco Iansiti and Roy Levien's work on keystones, set out in "The Keystone Advantage" in 2004. They distinguished three roles a firm can play in a business ecosystem. A keystone occupies a central, well connected position but holds only a small share of the network, and acts to improve the overall health of the ecosystem, creating and sharing value, on the reasoning that its own sustained performance depends on the health of the system it sits inside. A dominator integrates vertically or horizontally to own and control a large proportion of the network and to capture most of the value it creates, leaving little room for a meaningful ecosystem to form. And then there is a third role they name with unusual bluntness: the hub landlord, which they call the most anti social kind of dominator, a firm that gives up control of the network's activity but pursues control of value extraction alone, contributing little new value and leaving, in their exact phrase, a starved and unstable ecosystem around it.

Pause on this, because it is the most important sentence in the canon for the purposes of this paper. The field's own foundational text already contains a precise description of the extractive failure mode, and a precise prediction of its consequence. The hub landlord is not an outside critic's caricature of platform capitalism. It is a category inside the discipline's own taxonomy, defined by the discipline's own authors, with the consequence stated plainly: extract without contributing and the system around you starves and destabilises. Iansiti and Levien did not stop there. They proposed three determinants of business ecosystem health: robustness, the capacity to survive disruption; productivity, the efficiency with which the network converts inputs into outputs; and niche creation, the capacity to generate meaningful diversity and genuinely new capability. This is the closest the strategy literature comes to the ecological reading. A healthy ecosystem, on their account, is robust, productive, and generative of diversity, and a hub landlord predictably destroys all three.

There is a fourth element of their framework that the paper needs and that is usually passed over, which is the niche player. In Iansiti and Levien's account the niche player is the specialised participant that develops a focused capability and depends on the keystone's platform to reach the rest of the system. The health of an ecosystem, on their account, is substantially the health of its niche players, because they are where diversity and new capability actually live. This matters here for a reason that becomes central later. The niche player is the participant with the most to gain from a genuine keystone and the most to lose from a hub landlord, and it is also the participant with the least ability to verify which one it is dealing with from inside the relationship. A niche player sees its own terms, its own revenue, its own dependence. It does not see the architecture. It cannot tell, from where it stands, whether the centre's prosperity is coupled to its renewal or merely currently aligned with it, and that informational position, the participant who bears the consequence of the architecture and cannot see the architecture, is the precise problem the last section of this paper resolves. Hold it.

So the canon was right. The reason this paper is necessary is not that the literature failed to see the problem. It is that the literature named the problem precisely and the practice optimised for it anyway, because the hub landlord position, for the firm that holds it, is extremely profitable for exactly as long as the starvation takes to register. Iansiti and Levien's framework can describe the failure mode with clarity. What it cannot do, and what no strategy framework alone can do, is make the difference between a keystone and a hub landlord structural rather than a matter of the keystone's disposition. On their account a keystone improves ecosystem health because it is in its long term interest to do so. That is true and it is also exactly the kind of claim that does not survive contact with quarterly incentives, capital markets that discount the long term, and the simple fact that the starvation of a network is slow and the extraction from it is fast. A health that depends on the central firm choosing health, against its short term interest, is not a structural property. It is a hope. The rest of this paper is about what would have to be true for it to be a structural property instead.

It is worth stating the gap with full precision, because the precision is the contribution. Iansiti and Levien give a behavioural definition of the keystone: a keystone is a firm that behaves in a way that improves ecosystem health. A behavioural definition has a fatal property when the behaviour is costly and the cost is borne now while the benefit accrues later and diffusely. The same firm can satisfy the behavioural definition in one period and violate it in the next without anything observable changing at the moment the switch happens, because the switch is a change in disposition or incentive, not a change in form. A firm does not announce that it has stopped being a keystone and become a hub landlord. It simply begins, gradually, to price access higher, to widen the share it captures, to make the terms more revocable, each step individually defensible, and by the time the pattern is unmistakable the diversity it was sustaining has already thinned. A behavioural definition cannot raise the alarm until after the behaviour has changed, which is after the damage. What the paper is reaching for, and what the strategy literature structurally cannot supply because it operates at the level of firm behaviour, is a definition of the keystone that is architectural rather than behavioural: one under which a firm is a keystone because of how it is built and bound, not because of how it is currently behaving, so that the question of whether it is a keystone can be answered before the behaviour reveals the answer too late.

2.3 Adner: alignment is the precondition, and it is structural

Ron Adner's work supplies the piece the argument needs next, which is rigour about what an ecosystem actually is. In "The Wide Lens" (2012) Adner argued that innovators routinely fail not because their own innovation is weak but because they ignore the dependencies around it. He named two risks that live in the blind spot of conventional strategy. Co innovation risk is the extent to which your success depends on other parties successfully innovating in parallel. Adoption chain risk is the extent to which intermediaries have to adopt your innovation before the end customer can experience its value at all. His central tool, the value blueprint, is a map that makes the full set of dependencies explicit: who must do what, in what order, before the value proposition can materialise.

Adner sharpened this in "Ecosystem as Structure" (2017), where he defined an ecosystem as the alignment structure of the multilateral set of partners that need to interact in order for a focal value proposition to materialise. The word alignment is doing precise work there. An ecosystem is not, on this definition, a population of affiliated firms. It is a specific configuration of activities that has to hold together for value to exist at all. He distinguished this activity centred view, ecosystem as structure, from the looser actor centred view, ecosystem as affiliation, and the distinction matters here because it tells us that an ecosystem is, fundamentally, a structural object. It is held together by the arrangement of what must happen, not by the sentiment of the participants. In "Winning the Right Game" (2021) he extended this to ecosystem disruption: classic disruption happened inside an industry, modern disruption redraws the boundaries of the ecosystem itself, and a firm that wins the wrong game, the old industry game, while the ecosystem reconfigures around it, loses regardless of how well it played.

Adner's contribution to this paper is twofold and exact. First, he establishes that an ecosystem is a structural construct, which licenses the central claim that regeneration must be a structural property of that construct rather than a disposition of its central member. Second, his value blueprint is the closest the canon has to a diagnostic of whether an ecosystem can renew, and the precise way it falls short is instructive. The blueprint asks whether the activities are aligned well enough for the value proposition to materialise. It does not ask whether the delivery of that value depletes or renews the conditions the activities depend on. It is a map of whether the system works, not a map of whether the system, by working, consumes itself. That second map is the one the canon never drew, because the discipline it borrowed the word from had already drawn it, in 1969, and the borrowers had stopped reading the source.

The shortfall is not a defect in Adner's work; it is a boundary of what an activity map can do, and naming the boundary precisely is what licenses the rest of the paper. A value blueprint is a snapshot. It captures the configuration of activities at the moment it is drawn and asks whether that configuration delivers the value proposition. A configuration that delivers value beautifully in the period it is drawn can be, in the same instant, drawing down the conditions that will let it deliver value at all three periods later, and nothing in a snapshot can show that, because the depletion is a property of the system over time and the snapshot has no time in it. To see whether a system consumes itself you need not a map of its activities but a model of its trajectory: a way of asking not "does this work now" but "is the working of it now compatible with its still working later, or is the working now the mechanism of the later failure." Odum's model is exactly a model of trajectory rather than a snapshot, which is why the paper has to leave the strategy literature and go to the ecology to find it. The canon could describe the system. It could not, with its own instruments, describe the system aging. That is the instrument the next section retrieves.

3. The ecology the canon borrowed from and stopped reading

3.1 Paine's keystone, and the exact moment of inversion

The word keystone, which Iansiti and Levien made central to ecosystem strategy, has a precise scientific origin and a precise scientific meaning, and the meaning the strategy literature gave it is close to the opposite of the original.

Robert T. Paine coined keystone species in a 1969 paper, building on a removal experiment he had begun on the rocky shore of Makah Bay, Washington, in 1963. On an eight metre stretch of intertidal shore he removed the starfish Pisaster ochraceus and watched what happened to everything else. With the starfish gone, the mussel Mytilus californianus, freed from predation, monopolised the available space and crowded most other species out. Diversity collapsed. The point Paine drew from this is the point the strategy literature needed and did not take. The starfish was not abundant. It did not dominate the system by biomass or by occupying most of it. Its importance was entirely disproportionate to its share of the system, and it was defined not by what it captured but by what its presence sustained in everything around it. Remove it and the system did not lose a participant. It lost its structure. A keystone, in Paine's exact sense, is the element whose presence holds the diversity and persistence of the whole, and which is recognised precisely because the system cannot keep its shape without it.

There is a further detail in Paine's work that the strategy borrowing dropped and that the paper needs, which is the trophic cascade. In his 1979 Tansley Lecture to the British Ecological Society, Paine developed the idea that the removal or addition of a keystone does not produce a single local change but a cascade of changes propagating through the connected layers of the system, the rise of some populations and the collapse of others, rippling outward through the web. The cascade is the reason the keystone's effect is disproportionate: it does not act on one thing, it acts on the network of dependencies, and the network transmits and amplifies the action. This matters here because it tells us something exact about what happens when a central element in a business network shifts from sustaining to extracting. The effect is not a proportional reduction in the welfare of the participants directly connected to it. It is a cascade: the directly connected participants thin, which removes the conditions that supported the participants connected to them, which removes the conditions for the ones beyond, and the diversity of the whole collapses faster and further than any analysis of the central relationship alone would predict. The strategy literature's account of the hub landlord describes the centre starving the network. Paine's cascade tells us the starvation is not linear and not contained. It propagates, and it propagates through exactly the structure of dependencies that made the system worth calling an ecosystem in the first place.

Now read Iansiti and Levien's keystone against Paine's. The strategy keystone occupies the central hub, is recognised by its connectedness and its position, and is justified by the value it captures and the performance it sustains for itself. The biological keystone is recognised by the diversity it sustains in others and would not be called a keystone at all on the basis of what it holds for itself, since by Paine's measure it holds very little. The two definitions share a word and invert its content. The biological keystone is defined outward, by its effect on the persistence of others. The strategy keystone is defined inward, by its position and its capture, with the effect on others stated as a hoped for consequence of enlightened self interest. This is the inversion named in section one, located now at its exact point of occurrence. The strategy literature did not misuse the word casually. It took a concept whose entire content is "defined by what its presence sustains in everything else" and reattached it to a position defined by centrality and capture, then carried the prestige of the original across the gap. A keystone that has to be argued into caring about the system is not a keystone in the sense the word was coined to carry. It is the thing the word was coined to distinguish from.

The inversion has a property worth stating because it recurs through the rest of the paper. It is not symmetrical. The biological keystone, defined outward by what it sustains, can be verified from outside: remove it, or model its removal, and the effect on the diversity of everything else is observable and is the definition. The strategy keystone, defined inward by position and capture, cannot be verified outward at all, because its claim to be a keystone in the honourable sense is a claim about its disposition toward a system whose response to its absence is never tested. The biological concept is operational, in the strict sense that it specifies a measurement. The borrowed concept is aspirational, in the strict sense that it specifies an intention. Every time this paper meets the inversion again it will have this same shape: an operational, outward, verifiable property in the source, replaced by an aspirational, inward, unverifiable one in the borrowing, with the prestige of the first carried across to the second. Naming the shape once is enough to recognise it each subsequent time, and the recognition is most of what the paper is teaching the reader to do.

3.2 Odum's two ecosystems, and the diagnostic the field never had

The deeper instrument is older than the strategy literature by twenty four years and it is the structural spine of this paper. Eugene P. Odum published "The Strategy of Ecosystem Development" in Science in 1969. It set out, as a tabulated model, the systematic differences between an ecosystem in its early developmental state and an ecosystem in its mature state, across roughly two dozen attributes. The differences are not differences of degree. They describe two structurally distinct kinds of system, and the distinction is exactly the one the business ecosystem literature has been talking around for thirty years without naming.

A young or developing ecosystem is organised for throughput. Its rate of production exceeds its rate of respiration, so it accumulates yield faster than it spends energy maintaining itself. It has low standing biomass relative to its production, simple structure, low diversity, open and leaky nutrient cycling in which materials pass through and out rather than circulating, and low stability. It is fast, productive, and fragile. It is, in the economic vocabulary, optimised for harvestable output. A mature or climax ecosystem is organised for persistence. Its production and respiration come into balance, so it is no longer accumulating yield but sustaining a large, complex standing structure. It has high biomass, high diversity, high internal information, tight and largely closed nutrient cycling in which materials are retained and recirculated rather than lost, and high resilience and homeostasis. It is not optimised for output. It is optimised for the continued existence of the system that produces the output.

Odum then made explicit the conflict that this paper is, at the level of business networks, an extended treatment of. Human economic strategy, he observed, is geared to maximise production: high yield, rapid harvest, low retained structure, the deliberate maintenance of systems in their young, high throughput state because that state is where extractable output is highest. Nature's strategy, in the development of an undisturbed ecosystem, runs the other way: toward retention, complexity, closed cycling, and the protection of the system over the maximisation of its yield. The two strategies are not variants of one thing. They pull in opposite directions, and a system can be held in the young state indefinitely only by continuous external input and continuous prevention of the maturation that would otherwise occur.

There is a precision in Odum's model that the paper should not blur, because the precision is what makes it a diagnostic rather than a metaphor. His was a tabulated account of roughly two dozen attributes, sorted into groups covering the system's energetics, its nutrient cycling, its community structure, the life history strategies of its members, the kind of selection pressure operating on it, and the degree of internal regulation it achieved. The point of the table was that these attributes move together. A system does not get to have the high yield of the young state and the resilience of the mature one; the attributes are not a menu. Open nutrient cycling and high production and low retained structure are the same condition seen from three sides, and so are closed cycling and balanced production and high retained structure. This co movement is exactly what makes the model usable on a business network. It means you cannot diagnose an ecosystem by asking whether it is productive, because productivity in the high throughput sense is the signature of the young state, not evidence of health. You diagnose it by asking what the productivity is doing to the cycling and the structure: whether the output is being produced by drawing the system's retained conditions down and out, or within a system that is holding and recirculating them. A single number, output, is uninformative by itself for the same reason a single vital sign is uninformative without the others. The diagnostic is the pattern, and the pattern is what the strategy literature, lacking the model, could not read.

A second feature of Odum's account is decisive and is the one the paper most needs. The young state is not a phase a system passes through on the way to maturity unless something prevents the passage. It is, specifically, the state a system is held in by disturbance. An ecosystem matures when it is left undisturbed; it is kept young by repeated disturbance that resets it before it can accumulate the structure of maturity. This is why agriculture is Odum's central human example: a field is a permanently young ecosystem, and it is kept young deliberately, by ploughing, by clearing, by the continuous external work of preventing the succession that would otherwise turn it into something with high retained structure and low harvestable yield. The young state is not natural to a system that is let alone. It is manufactured and maintained, at cost, because the young state is where the extractable output is. Carry this directly across. A business network held permanently in the young, high extraction, low retained structure state is not failing to mature. It is being prevented from maturing, continuously and deliberately, by an operator for whom the young state is where the yield is, exactly as a field is prevented from becoming a forest. The starvation Iansiti and Levien predicted is not an accident of bad keystone behaviour. It is the maintained condition of a system that is being farmed.

Set this against the strategy canon and the entire picture resolves. Moore's fourth stage, self renewal, is the maturation Odum describes: the transition from a system organised for expansion and yield to one organised for persistence and renewal. Iansiti and Levien's healthy ecosystem, robust, generative of diversity, is structurally Odum's mature ecosystem: high diversity, high resilience, retained and recirculated value. The hub landlord that starves its network is a mechanism for holding the system permanently in the young, leaky, high extraction state, drawing yield out faster than the system can retain or rebuild it, and the predicted instability Iansiti and Levien named is precisely the fragility Odum attributes to the young state. The strategy literature has been describing Odum's distinction the entire time, in its own vocabulary, without the diagnostic that would let an operator say which of the two systems they are actually running. This is the missing map from the end of section two. The value blueprint asks whether the system works. Odum's model asks the question the blueprint does not: is this system retaining and recirculating the conditions of its own continuation, or is it running them out one way and calling the resulting yield health.

3.3 The shared network nobody owns

There is one more figure from ecology worth importing carefully, because it sharpens what circulation means in a system that is not a single organism. In many forests, trees of different species are connected belowground through common networks of mycorrhizal fungi, and carbon and nutrients move through these networks between trees, including between trees of different species. Suzanne Simard and colleagues documented such transfer between paper birch and Douglas fir in mixed forests of British Columbia. We state this carefully because part of the surrounding narrative is genuinely contested: the stronger claim that mature trees preferentially direct resources to their own offspring through these networks is disputed, and recent reviews describe the evidence for that specific claim as inconclusive or absent. We will not cement the contested part. The structural truth is not contested and it is the only part the argument needs. Resources move through a shared belowground network that no single tree owns, that connects organisms which are also in competition with one another, and whose presence is associated with greater resilience of the system as a whole than the individual organisms would have alone. The lesson for a business ecosystem is not the sentimental one about cooperation. It is the structural one. In a regenerative system the medium through which value circulates is not owned by the strongest participant. It is the substrate, shared, and its function is to keep value moving through the system rather than letting it pool at one node and leave the rest depleted.

Figure 1. The young ecosystem and the mature ecosystem. Picture two systems that produce the same visible output. The young system is fast and lean: it converts inputs into harvestable yield quickly, holds little structure, lets nutrients pass through and out, and is fragile to any shock. The mature system produces a comparable output but is organised differently underneath: it holds a large, diverse standing structure, recirculates its nutrients internally instead of leaking them, and is resilient because the structure absorbs shocks the yield figure never shows. From the outside, in a single period, the two can look similar or the young one can even look better, because it is not paying the cost of the structure that makes the mature one durable. The difference is invisible in the output and decisive in the persistence. A business ecosystem optimised for the output number is, by construction, being held in the young state. A regenerative one has paid for the structure that does not show in the number.

4. Where orchestration becomes extraction

4.1 Orchestration is not the problem

It would be a misreading of this paper to take it as an argument against ecosystem orchestration as such. Orchestration, the work of aligning a multilateral set of participants so that a value proposition can exist at all, is necessary and is not extractive in itself. Adner's value blueprint exists because alignment does not happen on its own and someone has to do the structural work of arranging it. A common network of fungi is, in its own way, an orchestration: it aligns flows that would not align without it. The keystone in Paine's sense is performing a kind of orchestration, since the structure of the whole community depends on its presence. The question is never whether there is a central organising element. The question is what that element's relationship is to the renewal of the system it organises. Orchestration becomes extraction at a precise structural point, and naming that point exactly is the work of this section.

4.2 The exact transition

Orchestration becomes extraction at the point where the orchestrator's capture of value becomes structurally decoupled from the renewal of the participants who generate it. That sentence is the hinge of the paper and every word in it is load bearing.

Decoupled means that the orchestrator can increase its own capture without the renewal of the participants improving, and crucially can continue to do so while their renewal declines. In a coupled system the central element prospers only as the system's capacity to continue prospers, which is the condition Iansiti and Levien described for a keystone and the condition Paine's starfish met by definition. In a decoupled system the central element's prosperity and the system's capacity to continue have come apart, so that the orchestrator's optimal short term move and the system's continued existence point in different directions. Structural is the operative qualifier. The decoupling is not a question of whether the orchestrator is greedy or benevolent. It is a question of whether the architecture of the system makes the orchestrator's capture depend on the participants' renewal, or merely makes it possible for a well disposed orchestrator to choose to support that renewal against its own incentives. Iansiti and Levien's keystone is the second case, which is why their healthy ecosystem rests on the central firm choosing health. A regenerative ecosystem is the first case, in which the architecture removes the choice by making the orchestrator's prosperity inseparable from the system's renewal, so that extraction at the expense of renewal is not a temptation resisted but a move the structure does not permit.

This gives a clean test, and it is the test the value blueprint does not perform. Take any arrangement called an ecosystem and ask one question. Can the central party increase its own capture while the capacity of the participants to continue and to renew is declining, and continue doing so for an extended period without the architecture itself imposing a cost on it for doing so. If the answer is yes, the arrangement is structurally extractive regardless of how it is described, how cooperative its rhetoric, or how benevolent its current operator. If the answer is no, because the architecture binds the centre's prosperity to the periphery's renewal, the arrangement is structurally regenerative regardless of whether anyone has used that word about it. The word in the marketing material is not evidence. The coupling in the architecture is.

4.3 The 2026 evidence: the hub landlord, regulated

This is not an abstract risk. It is the documented condition of the largest commercial structures currently called ecosystems, and by 2026 it is being treated as such by regulators in language that maps almost exactly onto Iansiti and Levien's own category.

The research literature on digital platform ecosystems through 2025 and into 2026 converges on a finding it sometimes calls the platform paradox. Value creation on these platforms is highly distributed: it is performed by large populations of complementors, sellers, developers, and creators. Value capture is highly centralised: it concentrates at the orchestrator, which exercises structural power over the participants through its control of access, terms, ranking, and the data the participants generate. This is the structural decoupling of section 4.2 described in another vocabulary. It is also, precisely, the hub landlord: control of value extraction retained at the centre, the activity left to the participants, and a network that is, in Iansiti and Levien's word, starved.

The regulatory record makes the mapping concrete and verifiable. The European Union's Digital Markets Act designates a small set of large platform operators as gatekeepers across their core platform services. As of 2026 the designated gatekeepers are Alphabet, Amazon, Apple, Booking, ByteDance, Meta, and Microsoft. On 23 April 2025 the European Commission imposed its first non compliance fines under the Digital Markets Act: a fine of 500 million euro on Apple, in connection with restrictions on the ability of developers to steer users to offers outside the App Store, and a fine of 200 million euro on Meta, in connection with its consent or pay advertising model. Enforcement and litigation around these designations continued through late 2025 and into 2026. We state these as facts and do not overclaim their outcome, because the structural reading does not need the litigation to resolve in any particular direction. The structural reading is simply this. A regime of ex ante regulation, imposing obligations on a designated few before harm rather than after, is what a society builds when it has concluded that certain central operators have become structurally decoupled from the renewal of the participants around them and will not recouple on their own incentives. The Digital Markets Act is, read structurally, public infrastructure for the proposition that the hub landlord exists, that its starvation of the network is real enough to legislate against, and that the coupling Iansiti and Levien hoped a keystone would choose has to be imposed from outside because the architecture did not impose it from within. The discipline named the failure mode in 2004. Two decades later the failure mode is large enough that the law is being rebuilt around it.

4.4 Why the choosing keystone is not enough

There is a tempting response to all of this, which is that the answer is simply better keystones: central firms that understand their long term interest, that choose to share value, that recognise the network's health as their own. This response is not wrong so much as insufficient, and seeing exactly why is necessary before the paper can say what regeneration actually requires.

The choosing keystone fails for the same reason a single beam chosen for its strength fails to make a building safe: it is a property of a part, held against load, with no structural redundancy if the part's incentive changes. Three forces reliably change it. The first is time horizon. The starvation of a network is slow and the extraction from it is fast, so any discount rate above a low threshold makes extraction the rational move even for an operator who fully understands the long term consequence, because the long term is discounted and the extraction is now. The second is succession. A keystone that chooses health is choosing with a particular operator's judgement, and operators change, are acquired, are pressured by owners who did not make the original choice and do not feel bound by it. A disposition does not survive a change of the person holding it; only a structure does. The third is capital. A central firm in public or institutional ownership is answerable to allocators who can move capital faster than a network can starve, which means the firm is structurally rewarded for the young, high extraction state and structurally punished for paying the cost of the mature, retained one, regardless of what its operators privately understand. Against all three forces, a health that depends on the keystone choosing health is a hope under load. The conclusion is not that good operators do not matter. It is that regeneration cannot be left as something the central party is trusted to choose, because every durable force in the system is pushing the choice the other way. It has to be built into the architecture, where it does not depend on anyone choosing it. That is the subject of the next section, and it is the whole of the difference between a regenerative ecosystem rhetorically and a regenerative ecosystem structurally.

5. The inversion recurs: why the word regenerative is already being hollowed

5.1 The same mechanism, one level up

A reader who has followed the argument should now be uneasy about the word this paper itself uses. If the word ecosystem was borrowed from a science whose admirable properties did not transfer, and the borrowing stayed invisible for thirty years because no falsifiable claim was ever made, then the word regenerative is exposed to exactly the same mechanism, and is in fact further along it than the field admits. This section makes that exposure explicit, because a paper that named one inversion and then walked its own key term into the next one would have failed its own argument.

The mechanism, recall its shape, is this: an operational, outward, verifiable property in a source domain is replaced by an aspirational, inward, unverifiable one in the borrowing, and the prestige of the source is carried across the gap with no falsifiable claim left behind to refute. Apply it to regeneration. In the ecological and the commons literature, as the previous sections established, regeneration is an operational property: it is closed cycling rather than throughput, retained structure rather than minimised standing crop, the observed design principles of commons that endured for centuries rather than collapsed. It specifies measurements and structural features. In the contemporary business vocabulary, regeneration is increasingly an aspirational property: a stated commitment to multiple kinds of value, a posture toward stakeholders, a narrative of giving back, a chapter in a report. The research literature on regenerative business models is, to its credit, candid that the field is still largely conceptual and that the gap between the aspiration and any structural specification of it is wide and open. That candour is the field's own confirmation that the inversion is underway. The word that was operational in the source is becoming aspirational in the borrowing, on schedule, by the identical mechanism, and a paper called Regenerative Business Ecosystems that did not say so would be performing the very move it diagnoses.

5.2 Why the second inversion is more dangerous than the first

The recurrence is not merely tidy. It is worse than the first inversion, for a precise reason. The first inversion, of the word ecosystem, hollowed a descriptive term: it let extractive arrangements borrow the prestige of resilient ones. The second inversion, of the word regenerative, hollows the corrective term: it lets the same extractive arrangements borrow the prestige of the correction. When the descriptive word is captured, the failure is at least nameable, because the corrective vocabulary still has its edge. When the corrective word is captured too, the system loses the language it would have used to name its own failure, and a failure that cannot be named cannot be acted on. This is the structural danger of the regenerative vocabulary becoming a posture rather than an architecture: not that it is insincere, much of it is entirely sincere, but that its sincerity is precisely what makes it effective cover. A hub landlord that calls itself an ecosystem is using a word. A hub landlord that calls itself a regenerative ecosystem, with a values statement and a multi capital report and a genuine internal belief that it is one, has captured the only vocabulary the system had for distinguishing it from a regenerative one, and has done so in good faith, which is the hardest kind of capture to reverse.

This is the exact point at which the paper's central methodological commitment becomes load bearing rather than stylistic. The brief under which this research library operates does not permit cementing anything that is not structurally solid, and that constraint is not modesty. It is the only defence against the second inversion. The moment regeneration is allowed to be asserted rather than demonstrated, the word is lost, because assertion is free and the parties with the most to gain from asserting it are the parties least subject to its discipline. The only form in which the word survives with its meaning intact is the form in which it is a checkable structural claim and not a stated value, which is why the next section refuses to define a regenerative ecosystem by anything a system could simply declare about itself, and defines it only by architecture a system either has or does not. The refusal to cement the non solid is not a tic of the house style. It is the structural firewall against the inversion the rest of the field is walking into.

6. What a regenerative ecosystem requires structurally

6.1 The standard to meet

A regenerative business ecosystem, stated precisely, is one whose architecture binds the prosperity of its central organising element to the renewal of the conditions and participants it depends on, such that extraction at the expense of that renewal is not a temptation the centre is trusted to resist but a move the structure does not permit. Everything in this section is an attempt to say what that architecture concretely contains, drawn not from values language but from two bodies of work that describe how systems actually retain rather than exhaust themselves: Odum's account of the mature ecosystem, and Elinor Ostrom's account of how a commons is governed so that it renews rather than collapses. We use these because they are the rare places where renewal has been specified structurally rather than asserted morally.

6.2 Odum's mature attributes, translated

Odum's mature ecosystem differs from the young one in attributes that translate directly into network design. Four of them are load bearing here.

Closed cycling rather than throughput. The defining structural difference between Odum's mature and young systems is that the mature one retains and recirculates its nutrients internally while the young one lets them pass through and out. Translated: in a regenerative ecosystem, value generated by participants is substantially recirculated to the participants and to the renewal of the system, rather than extracted out of it to the centre and its owners. This is not a redistribution preference. It is the single structural property that distinguishes a system that can persist from one that is being run down, and it is measurable in principle as the ratio of value retained and recirculated within the system to value extracted out of it. A system whose extraction ratio rises over time is maturing in reverse, regardless of how its output number behaves.

Retained structure rather than minimised standing crop. The young system minimises the structure it holds so that more of its production is harvestable. The mature system carries a large standing structure that does not appear in the yield figure and is exactly what makes it resilient. Translated: a regenerative ecosystem deliberately carries structure, redundancy, slack, diversity of participants, capabilities not currently being monetised, that an extraction optimised system would strip out as waste. The structure is the cost of durability and it is invisible in any single period's output, which is precisely why an output optimised operator removes it and why its removal is the mechanism of decline rather than its symptom.

Diversity rather than monoculture. Iansiti and Levien's niche creation and Odum's high diversity are the same property. A mature system holds many kinds of participant occupying many kinds of position, which is what gives it the capacity to absorb shocks and generate genuinely new capability. The hub landlord's predicted effect, a starved and unstable system, is the collapse of this diversity into a monoculture dependent on the centre. Translated: a regenerative ecosystem is structurally plural, and the test of this is whether participants can occupy positions the centre does not control and cannot unilaterally revoke.

Resilience valued over yield. The summarising attribute. The mature system is not optimised for the maximum extractable output; it is optimised for continuing to exist, and it accepts a lower yield to do so. A regenerative ecosystem is one whose architecture has, structurally, made the same trade: it does not maximise the output number, because the output number maximised is the young, fragile, declining state wearing the appearance of health.

6.3 Ostrom's principles, as the governance spine

Odum describes what a renewing system looks like. He does not describe how a system composed of self interested participants, including a powerful central one, governs itself into that state and stays there. For that the paper turns to Elinor Ostrom, whose "Governing the Commons" (1990) is the most rigorous account in the social sciences of how groups sustain a shared resource over long periods without either privatising it into extraction or having it imposed from above. Ostrom studied commons that endured for centuries, irrigation systems, forests, fisheries, pastures, and identified design principles their governance shared. They are not aspirations. They are the observed structural features of systems that renewed rather than collapsed, and they map onto the regenerative ecosystem with very little translation.

Clearly defined boundaries. It must be clear who is a participant in the system and who is not, because a resource open to unbounded extraction by undefined parties is the classic path to collapse. In an ecosystem this is the requirement that membership and the terms of participation are defined and stable, not arbitrary and revocable at the centre's discretion.

Rules matched to local conditions, and made by those they bind. Ostrom's enduring commons did not run on rules handed down by a distant authority. The participants affected by the rules participated in making and modifying them. This is the principle that most directly attacks the hub landlord, because the hub landlord's defining feature is precisely that it sets the terms unilaterally and the participants who generate the value have no constitutional role in the rules that govern their participation. An ecosystem in which the rules of value distribution are set by the party that captures the value, with no mechanism by which the participants who create it shape those rules, fails this principle by construction.

Monitoring and graduated sanctions, by accountable monitors. Enduring commons were watched, and the watchers were accountable to the participants, not only to the centre, and sanctions for violation escalated rather than being absent or absolute. Translated: a regenerative ecosystem has visibility into whether the coupling between centre and renewal is holding, and that visibility is not the private possession of the centre.

Accessible conflict resolution, and the recognised right to organise. Participants must have low cost ways to resolve disputes and a recognised right to organise their own collective arrangements. An ecosystem in which the only forum for dispute is controlled by the party one is in dispute with, and in which participants have no recognised standing to organise, is structurally a domain, not a commons.

Nested governance. For systems above a certain scale, Ostrom found that governance was organised in nested layers rather than centralised, so that local arrangements were embedded in and accountable to wider ones. This is the structural form of Carol Sanford's observation that living systems are nested: a regenerative ecosystem is not a flat field administered from one point but a set of nested arrangements, each accountable to the larger system it sits inside, which is also the precise structural reason a regenerative ecosystem cannot prosper while degrading the wider systems it is nested in, since by construction it is accountable to them.

Figure 2. The orchestrated platform and the governed commons. Two systems with a central organising element and many participants. In the first, the centre sets the terms, holds the data, controls access, resolves disputes, and captures the value, and the participants generate the value under terms they did not make and cannot change. It looks efficient and it is, for a while, until the network it depends on is starved. In the second, the centre still organises, but boundaries are defined, the rules are made partly by those they bind, monitoring is visible to the participants and not only to the centre, disputes have a forum the centre does not own, participants have the recognised right to organise, value substantially recirculates rather than concentrating, and the whole arrangement is nested inside and accountable to the wider systems it depends on. The first is an orchestrated platform. The second is a governed commons. Both can be called ecosystems. Only the second has the structure the word was borrowed to claim.

6.4 The five conditions, stated as architecture

We can now state, without metaphor, what a regenerative business ecosystem structurally requires. Not values. Architecture.

First, coupled prosperity. The architecture binds the central element's prosperity to the renewal of the participants, such that the centre cannot durably increase its capture while their capacity to continue and renew declines, without the structure imposing a cost on it for doing so. This is the non negotiable condition and the other four are how it is held in place.

Second, recirculation over extraction. Value generated within the system is substantially retained and recirculated within it rather than run out to the centre, with the extraction ratio a monitored quantity rather than an emergent one. This is Odum's closed cycling.

Third, constitutional participation. The participants who generate value have a real and recognised role in making and changing the rules that govern their participation, not merely the freedom to accept or leave terms set unilaterally by the party that captures the value. This is Ostrom's collective choice principle and it is the structural antidote to the hub landlord.

Fourth, retained structure and plurality. The system deliberately carries diversity, redundancy, and capability beyond what current extraction requires, and participants can hold positions the centre does not control and cannot unilaterally revoke. This is Odum's standing structure and Iansiti and Levien's niche creation made structural rather than hoped for.

Fifth, accountable nestedness. The system is organised in nested arrangements each accountable to the wider systems it depends on, with monitoring and dispute resolution that are not the private possession of the centre. This is Ostrom's nesting and accountable monitoring, and it is the structural reason a regenerative ecosystem cannot be healthy while degrading what it is nested in.

A system that meets all five is regenerative as a structural property and not as a claim. A system that meets the rhetoric of regeneration but fails the first is the young, extractive ecosystem wearing the vocabulary of the mature one, which is the condition this paper opened by naming and has now specified precisely enough to test.

7. The layer regeneration actually settles on

7.1 Why none of this can be anchored at the orchestration layer

There is a question the previous section's five conditions raise and do not answer, and it is the question the rest of the paper turns on. Each condition is a statement about the architecture of the system: coupled prosperity, recirculation, constitutional participation, retained plurality, accountable nestedness. But an architecture is not self enforcing. Each of these conditions can be written into terms, declared in a charter, asserted in a manifesto, and then quietly hollowed out while the language remains, which is exactly what the inversion in section one describes happening to the word ecosystem itself. So the operative question is not what the conditions are. It is what holds them in place over time, against the three forces, horizon, succession, capital, that section 4.4 showed will reliably push the centre to defect from them. The conditions are necessary. The thing that makes them durable is a separate question, and it is the one the paper now has to answer.

Begin with a distinction the FP research library has developed elsewhere and which this argument independently arrives at. There is a layer of any commercial system concerned with orchestration: the arrangement of activities, the platform, the terms, the flows, the machinery by which value is produced and moved. This is the layer the entire ecosystem strategy canon operates on. Moore's stages, Iansiti and Levien's roles, Adner's blueprint are all descriptions and tools at the orchestration layer. It is also the layer that changes constantly. Terms are rewritten, platforms are reconfigured, the central element is acquired, the participants change, the orchestration of a large system a decade ago is not the orchestration of it now. Anything anchored only at the orchestration layer is anchored to something that does not hold still, which is precisely why a charter written into terms can be hollowed while the terms remain: the orchestration moved underneath the language and the language was the only thing the commitment was attached to.

There is a second layer, quieter and slower, and it is the one every one of the five conditions actually depends on to mean anything. It is not concerned with how value is orchestrated. It is concerned with whether the system is what it claims to be: whether the coupling that was declared is the coupling that is operating, whether the recirculation that was promised is the recirculation that is happening, whether the participants who were given a constitutional role still have it in operation and not only in text, whether the entity at the centre is the entity that made the commitment and can still be held to it, whether there is continuity between what was attested about the system and what the system is actually doing now. Call this the identity layer: the attestable, accountable, continuous answer to the question of whether the system is, in operation, the system it represents itself to be, and who stands behind the gap when it is not. Every one of the five structural conditions is, on inspection, a claim that has to be true at the identity layer to be true at all. Coupled prosperity is a claim about whether the centre's incentives are actually bound, not whether a document says they are. Recirculation is a claim about where value actually goes, not where a report says it goes. Constitutional participation is a claim about whether the participants' role is operative, not whether it is written. The orchestration layer is where the system runs. The identity layer is where it is either true or merely claimed.

7.2 Identity is the only layer slow enough to hold a commitment

The reason regeneration settles on the identity layer rather than the orchestration layer is not a preference. It is the same property that made Odum's mature ecosystem settle on retained structure rather than on yield. A commitment can only be held by something stable enough to be held to. Orchestration is not stable enough; it changes faster than any commitment written into it can bind, which is exactly why commitments written only into terms are hollowed as the terms are rewritten and the operator changes and the capital reprices the horizon. The orchestration layer cannot carry a regenerative commitment over time for the same reason the young ecosystem cannot carry persistence: nothing there holds still long enough to take and keep an impression. The identity layer can, because identity is, by definition, the part that is supposed to remain continuous while the orchestration underneath it changes. Who stands behind this, what was attested about how it treats the system it sits inside, is the attested arrangement the operating arrangement, is there continuity between the claim and the conduct: these are properties of identity, not of orchestration, and they are the only properties slow enough for a regenerative commitment to be anchored to without being washed out by the next reconfiguration.

This is why the durable position in a regenerative ecosystem is not at the orchestration layer where the entire canon and almost all the capital and attention sit. The orchestration layer is where the network runs and it is therefore precisely where nothing settles, because everything there is being rewritten. The identity layer is where the question of whether the system is regenerative is actually decided, because that question is not "is the orchestration efficient" but "is the system what it claims to be, over time, and who is accountable when it is not," and that is an identity question by construction. An entity operating at the identity layer, holding the continuous, attestable account of whether a system's regenerative structure is real and is holding, is operating at the exact place a regenerative commitment has to be anchored to survive the forces that otherwise dissolve it.

7.3 What it means to already be operating there

To be at the identity layer of a regenerative ecosystem is not to publish a charter or to adopt the vocabulary. It is to be, in the present tense, the thing the five conditions resolve to when they have to be more than language: a continuous, accountable account of whether a system's coupling is real, whether its recirculation is happening, whether its participants' standing is operative, whether the entity behind it is the entity that committed and can still be held, treated as a reference the system is measured against rather than a claim the system makes about itself. This is the problem Future Proof was built underneath. Not as a commentary on regenerative ecosystems but as one of the things already operating at the identity layer where their regenerative structure either holds or is quietly hollowed: an AI era trust standard held as a foundational reference, an identity layer over the orchestration layer rather than a competitor within it, the continuity between what a system attests about how it treats the network it depends on and what it is actually doing. The argument of this paper has been describing the shape of that layer since its first page: the place where the difference between a regenerative ecosystem and a young extractive one wearing its vocabulary stops being rhetorical and becomes attestable. A reader who has followed the argument will recognise the shape. It is not a layer the strategy canon built, because the canon operates entirely at orchestration. It is the layer the whole argument has been reaching toward, the one slow enough for a regenerative commitment to be held at, and being there before a system needs it is not a claim one makes but a position one already holds or does not, on the day the system is finally tested against what it said it was.

8. Implications

The reader of a paper like this is usually doing quiet research before a decision. The implications below are written for that reader, by the kind of decision they face. They are concrete and they stay in the register of the argument. This is what the structure does, not an offer.

For institutions

If your institution participates in, depends on, or operates an arrangement called an ecosystem, the question that ages well is not whether the arrangement is cooperative or how it describes itself. It is the test in section 4.2 applied honestly to your own position. Can the central party in the arrangement you depend on increase its capture while your capacity to continue and renew declines, for an extended period, without the architecture imposing a cost on it for doing so. If yes, you are a participant in a young extractive system and the language of partnership around it is not evidence to the contrary; it is the vocabulary the inversion runs on. The institutions that will look prescient later are not the ones that joined the largest ecosystem. They are the ones that read the architecture rather than the rhetoric, required the five structural conditions rather than the values statement, and anchored their participation to something at the identity layer that can attest whether the coupling is real, so that when the arrangement is tested they are inside a structure that holds rather than holding a charter that was hollowed while the words stayed.

For investors

The diligence question that ages well is not whether a company sits at the centre of an ecosystem. On the strategy canon's own terms that position is desirable, and it is exactly the position that becomes a hub landlord and is then, as the 2026 regulatory record shows, the position public infrastructure is rebuilt to constrain. The question that ages well is structural. Is this company's prosperity coupled to the renewal of the participants it depends on, by architecture, or only by the current operator's disposition, which the three forces of horizon, succession, and capital reliably erode. A company whose returns depend on holding a network in its young extractive state is carrying an unpriced liability that hardens as the network starves and as ex ante regulation arrives, both of which are now observable rather than speculative. A company whose architecture binds its prosperity to the renewal of its network, and which can attest that binding at the identity layer rather than assert it, holds an asset that appreciates precisely as the extractive model becomes more regulated and less durable, because credible structural regeneration becomes scarce exactly when it becomes necessary. The thing to underwrite is not ecosystem position. It is coupling, attested rather than claimed.

For operators

The most expensive misreading available to you is the one the canon's own founder warned against and the field then professionalised away from: that leadership of an ecosystem is a destination rather than a fork. It is a fork. The branch that is not terminal requires you to renew the conditions, including the prosperity of the participants you have learned to depend on and extract from, against every short term incentive telling you that the extraction is rational because the starvation is slow. The operator's move is not to adopt the vocabulary of regeneration, which is free and worth nothing, but to build the five structural conditions into the architecture so that they do not depend on you continuing to choose them, and to anchor them at the identity layer so that they survive your successor, your acquirer, and your owners. A regeneration that depends on you choosing it every quarter is not a structure. It is a disposition under load, and the load does not relent.

For the people inside these systems

There is a reader this paper has not yet addressed directly: the participant who is not orchestrating, investing, or running the ecosystem, but generating the value inside it under terms set by a party that captures it. For that reader the argument is not abstract and the substrate register has to be most honest here. The difference between a young extractive ecosystem and a regenerative one is, from inside it, the difference between being a resource the system runs down and a participant the system is structurally required to renew. The five conditions are not governance theory to that reader. Constitutional participation is whether they have a recognised voice in the rules that govern their work or only the freedom to accept or leave them. Recirculation is whether the value they create stays in reach of them or is run out past them. Coupled prosperity is whether the system needs them to continue prospering or merely needs them to keep producing until they are exhausted and replaced. The reason to build regeneration into the architecture, and to anchor it where it cannot be hollowed, is not finally that the position is durable for the centre. It is that the architecture decides whether the people inside the system are something it sustains or something it spends. That is the load the layer actually bears. Everything else in this paper is the engineering around it.

9. Coda

The instinct, faced with the word ecosystem, is to trust it, because the word is borrowed from systems we admire and the borrowing carries their prestige across to whatever it is attached to. The argument of this paper is that the prestige is exactly the part that did not transfer. The strategy literature took the word from a science whose entire content is the difference between systems that renew themselves and systems that exhaust themselves, kept the admiration the word carries, and built a discipline mostly around the state the science calls young: fast, high yield, leaky, fragile, held in that state deliberately because that state is where the extractable output is highest. The founder of the field built the other state into the original architecture and called it self renewal or death, and the field kept the leadership stage and treated renewal as an epilogue. It was not an epilogue. It was the only non terminal branch.

A regenerative business ecosystem is not a kinder version of an extractive one. It is a structurally different system that can produce a comparable output while retaining and recirculating the conditions of its own continuation instead of running them out and calling the resulting yield health. The difference does not show in any single period's number, which is precisely why the number optimised operator strips out the structure that constitutes it and why the stripping is the mechanism of decline rather than its symptom. The difference is not rhetorical and cannot be made real by language, because language is exactly what the inversion runs on. It is architectural, and it is held in place not at the orchestration layer where everything is rewritten but at the identity layer where the question of whether a system is what it claims to be either has a continuous, accountable answer or does not.

There is a version of caring about regeneration that is just vocabulary, and it is now abundant, because the vocabulary is free and the architecture is not. This is the other kind. To build a regenerative ecosystem is not to declare one. It is to bind the centre's prosperity to the network's renewal in the structure, where no one has to keep choosing it, and to anchor that binding where it cannot be quietly hollowed while the words remain. When a system is finally tested against what it said it was, and they are all tested eventually, the test does not read the charter. It reads the architecture, and it reads it at the layer that held still. The only question this paper finally asks is whether, on that day, the regeneration was in the structure or only in the word.

References and Notes

The following are real, public, verifiable sources. Where a claim in the surrounding literature is contested, in particular the stronger interpretations of belowground resource transfer between trees, the paper states the structural truth that is not contested and does not cement the contested detail, as noted in the text.

1. James F. Moore, "Predators and Prey: A New Ecology of Competition," Harvard Business Review, May to June 1993, pages 75 to 86. The founding article of the business ecosystem concept, including the four stage model: birth, expansion, leadership, and self renewal or death. Extended in James F. Moore, "The Death of Competition: Leadership and Strategy in the Age of Business Ecosystems" (HarperBusiness, 1996).

1. Marco Iansiti and Roy Levien, "The Keystone Advantage: What the New Dynamics of Business Ecosystems Mean for Strategy, Innovation, and Sustainability" (Harvard Business School Press, 2004), and the related article Marco Iansiti and Roy Levien, "Strategy as Ecology," Harvard Business Review, March 2004. Source for the keystone, dominator, and hub landlord roles and for the three determinants of ecosystem health: robustness, productivity, and niche creation.

1. Ron Adner, "The Wide Lens: A New Strategy for Innovation" (Portfolio Penguin, 2012). Source for co innovation risk, adoption chain risk, and the value blueprint.

1. Ron Adner, "Ecosystem as Structure: An Actionable Construct for Strategy," Journal of Management, volume 43, number 1, 2017, pages 39 to 58. Source for the definition of an ecosystem as the alignment structure of the multilateral set of partners that need to interact for a focal value proposition to materialise, and for the distinction between ecosystem as structure and ecosystem as affiliation.

1. Ron Adner, "Winning the Right Game: How to Disrupt, Defend, and Deliver in a Changing World" (MIT Press, 2021). Source for ecosystem disruption and the reconfiguration of ecosystem boundaries.

1. Robert T. Paine, "A Note on Trophic Complexity and Community Stability," The American Naturalist, 1969, in which the term keystone species is introduced, building on the intertidal removal experiments reported in Robert T. Paine, "Food Web Complexity and Species Diversity," The American Naturalist, 1966. The concept of the trophic cascade was developed in Paine's 1979 Tansley Lecture to the British Ecological Society.

1. Eugene P. Odum, "The Strategy of Ecosystem Development," Science, volume 164, 1969, pages 262 to 270. Source for the systematic distinction between the young, high production ecosystem and the mature, high retention ecosystem, and for the stated conflict between human production strategy and the developmental strategy of an undisturbed ecosystem.

1. Suzanne W. Simard and colleagues, research on belowground carbon and nutrient transfer between tree species via common mycorrhizal networks in mixed temperate forests of British Columbia, including the work reported in Suzanne W. Simard et al., "Net transfer of carbon between ectomycorrhizal tree species in the field," Nature, 1997. The stronger preferential kin signalling interpretation of common mycorrhizal networks is contested in the subsequent literature and recent reviews describe the evidence for that specific claim as inconclusive or absent; the paper relies only on the uncontested structural point.

1. Elinor Ostrom, "Governing the Commons: The Evolution of Institutions for Collective Action" (Cambridge University Press, 1990). Source for the design principles of long enduring self governed common pool resource institutions, including defined boundaries, collective choice arrangements, accountable monitoring, graduated sanctions, accessible conflict resolution, the recognised right to organise, and nested enterprises.

1. John Fullerton, "Regenerative Capitalism: How Universal Principles and Patterns Will Shape Our New Economy" (Capital Institute, 2015). Source for the principles of a regenerative economy, including robust circulatory flow and holistic wealth, used at the level of structural principle.

1. Carol Sanford, "The Regenerative Business: Redesign Work, Cultivate Human Potential, Achieve Extraordinary Outcomes" (Nicholas Brealey Publishing, 2017). Source for the nestedness of living systems as a structural rather than rhetorical property.

1. Daniel Christian Wahl, "Designing Regenerative Cultures" (Triarchy Press, 2016). Source for the distinction between sustainability as maintenance and regeneration as the restoration of a system's capacity to renew itself.

1. Regulation (EU) 2022/1925 of the European Parliament and of the Council on contestable and fair markets in the digital sector (the Digital Markets Act). Source for the designation of gatekeepers across core platform services. As of 2026 the designated gatekeepers are Alphabet, Amazon, Apple, Booking, ByteDance, Meta, and Microsoft. On 23 April 2025 the European Commission imposed its first non compliance fines under the Digital Markets Act, of 500 million euro on Apple in connection with App Store steering restrictions and of 200 million euro on Meta in connection with its consent or pay model. Enforcement and litigation continued through late 2025 and into 2026; the paper relies on the fact of designation and the fact of the fines, not on the resolution of ongoing proceedings.

1. The 2023 to 2026 research literature on regenerative business models and on value creation and capture in digital platform ecosystems, in journals including Business Strategy and the Environment and the Journal of Business Research, which documents the platform paradox of distributed value creation and centralised value capture and characterises the regenerative business model field as a serious but still early and largely conceptual research frontier.

A note on method. This is a reference paper, not investment, legal, or organisational advice. The strategy literature is engaged on its own terms and is treated as more rigorous than the practice that followed it. The ecological sources are used for the structural distinctions they actually establish and not for metaphorical colour. Where the surrounding science is contested, the paper states the part that is not.

Future Proof Intelligence . Research . No. VIII . MMXXVI

Future Proof Intelligence . Research . No. VIII . MMXXVI