From one node to holacracy: how organisations actually grow
A one-person founder and the founder of a corporation with a coordination board answer the same question in opposite ways: what happens at the boundary between my business and the outside world? The whole structure of a company grows out of that boundary, one stage at a time, and no stage can be skipped.
By Andrey Belokrylov · October 1, 2026 · 9 min read

The founder of a one-person company and the founder of a corporation with a coordination board answer the same question in fundamentally different ways: what happens at the boundary between my business and the outside world? The first answer fits in one sentence. The second takes several layers of management, and the founder did not build those layers from a blueprint. They were built step by step, because each step was demanded by the boundary itself.
That is the ontogenesis of an organisation. Not a set of growth stages you can walk through with a workbook, but a history of how an entire structure grows out of a single node in its environment, provided you honestly follow the tension at the edges.
Below is the model, then its application to a company that sells into Russia from somewhere else.
One tension, many forms
Eight metaphors of organisation are familiar to anyone who has read Gareth Morgan: the machine, the living organism, the brain, the culture, the political system, the psychic prison, the ecology and the instrument of domination. Each looks at a company from its own angle. The ontogenetic structure is not a ninth metaphor. It is a way of seeing how a company moves through time across all of those forms at once, from birth to maturity.
There is one principle running through the whole model. Every new structure is born from tension on the membrane between the node and the environment.
The node is the point where the business is concentrated: one person with an idea, one product, one way of responding to a customer. The membrane is the boundary where that node touches the outside world: customers, competitors, resources it lacks. While the tension on that boundary is low, the node has no reason to complicate anything. As soon as the tension rises, and it rises inevitably because the environment is larger and more varied than a single point, the node needs a new form that will relieve it. That is how the next stage is born. Not because “this is how companies are supposed to grow”, but because the old form stopped coping with the pressure at the boundary.
The sequence, and why each link follows the last
The overall path looks like this: environment and node, contact, feedback, expansion, transaction centres, the dominant centre, hierarchy in the form of a coordination board, divisions, functional specialisation, delegation and the project, holacracy.
This is not a list of functions to implement. It is a chain in which each link is a direct answer to the tension left over from the previous one.
The node makes contact with the environment and immediately receives feedback, which either confirms the chosen way of acting or demands that it be revised. Positive feedback pushes toward expansion: the node starts occupying more space in the environment. But expansion cannot be run from one point forever. As soon as there are more deals, clients and directions than one node can hold, transaction centres appear: several points through which interaction with the environment passes. Between those centres a difference in influence inevitably emerges, and it takes shape as a dominant centre, the one around which the others begin to arrange themselves.
A dominant centre cannot coordinate a growing number of centres on its own. That is how hierarchy is born, and it takes the form of a coordination board rather than a single person’s will. Further growth demands division as well as coordination: divisions appear, each with its own part of the environment and its own focus.
Then functional specialisation switches on, and it works across divisions rather than inside them. While a function of the same type is duplicated in each division and does identical work everywhere, that is an intermediate state. Maturity pulls the function out into a separate unit that aggregates it for the whole company and holds the horizontal links between divisions. Specialisation in turn demands delegation: the head of a division can no longer personally control every function, and the project form appears as a way to temporarily assemble different specialisations under one task. And finally, when project intersections multiply beyond what any fixed hierarchy can untangle, holacracy emerges: a distributed structure where authority is attached to a role and a specific zone of responsibility rather than to a position.
Why a growth stage cannot be skipped
The main practical conclusion of this model is uncomfortable for anyone who likes ready-made templates. You cannot take a company at the single-node stage and install holacracy in it, because holacracy relieves a completely different tension, one that this company has not yet grown into. Just as you cannot hold a sprawling company with dozens of transaction centres in the logic of a single node, because the tension at its boundaries long ago exceeded what that form can bear.
This is where the model earns its keep for anyone who talks with entrepreneurs or builds an organisation of their own. Whoever runs a business right now lays under it exactly the level of awareness of the environment and of themselves that is available to them at this moment. No more and no less.
Every stage has its own set of questions, worth asking yourself or the person across the table, to find out where the tension on the boundary runs right now. What is the node or the division genuinely failing to cope with today, not in theory two years from now? The answer almost always points to the needed structure more precisely than any fashion in organisational design.
Each of the ten stages deserves its own conversation: its own questions, traps and moment to move on. What matters here is the logic by which they follow one another.
The Russian angle: a foreign company is opening a new node
A company that sells into Russia from another country rarely thinks of itself as being at any “stage”. It is a mature business with divisions, specialised functions and a project culture. At home that is true. In Russia it is not.
What such a company actually does when it enters Russia is open a new transaction centre in an environment it has never touched. That centre starts at the beginning of the sequence: contact, feedback, and only then expansion. And the environment on the other side of this membrane differs in ways a translated website hides. Search is dominated by Yandex rather than Google, with its own auction, ad formats and intermediaries between you and the buyer. Social advertising runs on VK and Telegram, because the Western platforms are unavailable. A large share of buyer contact happens in messengers rather than by email. The feedback loop at this boundary has to be built locally, from scratch, and read by someone standing at the boundary.
The mistake I see most often follows directly from the model. The parent company plugs the new centre into its mature functional structure. Marketing is aggregated at headquarters, pricing is set by a global team, and brand guidelines are enforced by people who have never seen a Russian search results page. In the language of the model, a function that should still live at the node has been pulled up into a company-wide unit before the node has produced any feedback worth aggregating.
The symptoms are recognisable. The Russian site is a translation. Campaigns copied from the home account quietly go dark. Reports go up the hierarchy and only budget cuts come back down. Then someone concludes that “Russia does not work for us”, when what did not work was the structure.
What local practice demands at each early stage
If you accept that the Russian operation is a young node inside a mature company, the model tells you what to protect at each of the first stages.
At contact, someone on your side has to be in the environment, not above it: research with real Russian buyers rather than a desk study, and a person who reads the Yandex auction and the messenger conversations first-hand. This is the work described on the Russian market entry page: customer research, positioning for this buyer, and a channel plan built from the demand that exists here rather than from the home-market media mix.
At feedback, the loop has to be short and local. If every change to a landing page or a bid needs approval three time zones away, from a team that does not read the language, the feedback arrives too late and the node learns nothing. Give the centre the authority to act on what it sees.
At expansion, resist scaling the structure before the contact is confirmed. Expansion is a response to positive feedback. Adding headcount, agencies and tooling before the first campaigns have produced a readable signal is expansion without feedback, and the model has no such stage. Most foreign companies never need the later stages inside Russia. The point is to recognise which stage you are at, not to reach the end of the list.
Reading the boundary instead of copying the org chart
The most useful thing about the model for a manager is a question, not a diagram. Where is the tension on the boundary right now, and what is the current form failing to do about it? For the Russian operation of a foreign company the answer is almost always the same: the node has not finished making contact, and the structure on top of it is stopping it.
That is the conversation I try to have before any talk about channels or budgets, and it is why a marketing strategy for Russia starts with the boundary rather than with the media plan. Which stage is this operation at? What feedback has it received? Who reads that feedback, and can they act on it? A structure that matches the stage will grow. A structure borrowed from a later stage, however well it works at home, will sit on the node like a roof on an empty lot.
Why an organisation grows like tissue, not like a building
An organisation does not grow like a building, raised floor by floor to an approved plan. It grows like living tissue, thickening exactly where the outside world presses on it. No one decides in advance how many levels there will be or what shape they will take. The form is dictated by the boundary: the place where the node touches the environment and either withstands the tension or stops being itself.
For a company selling into Russia from abroad, that boundary is further from headquarters than any other. It deserves its own node, its own feedback loop and its own stage, before it is asked to fit the shape of the whole.
Frequently asked questions
What is the ontogenetic model of an organisation?
It is a way of reading a company's history as a sequence of structures, each created to relieve a tension at the boundary between the business and its environment. A single node makes contact with the environment, receives feedback, expands, splits into transaction centres, forms a dominant centre, builds a coordination board, divides into divisions, specialises functions across them, delegates through projects and, at the far end, distributes authority into roles. Nothing in the sequence is prescribed. Each stage is a response to the pressure the previous one could no longer absorb.
Can a company skip a stage and adopt holacracy straight away?
No. Holacracy relieves the tension that appears when project intersections exceed what any fixed hierarchy can resolve. A company that is still one node, or a handful of transaction centres, does not have that tension. Installing a structure designed for a pressure you do not feel adds cost and confusion without adding capacity, and the reverse mistake, running a company of dozens of transaction centres as if it were a single node, is just as damaging.
How does this apply to a foreign company entering Russia?
A Russian operation is a new transaction centre, opened in an environment that is far more different from the home market than a translated website suggests. It has its own contact, its own feedback loop and its own tension. The frequent mistake is to plug it into the mature functional structure of the parent, so that marketing, sales and pricing decisions are aggregated abroad by people who never touch the Russian boundary. The centre then never gets the feedback it needs to grow.
What is the single most useful question this model gives a manager?
Where is the tension on the boundary right now, and what exactly is the current form failing to cope with? Not in theory, not in two years, but today. The answer points to the next structure far more reliably than any fashion in organisational design, and it works equally well as a diagnostic question in a conversation with an entrepreneur.
Sources
- Russian version of this article on belokrylovo.ru: От одной точки до холократии