Flat organisations are usually presented as a philosophy. More often they are a headcount.
The literature on networked teams and post-hierarchical work is written, almost without exception, from firms small enough that everyone who matters can be got into a single room and the founder can still name everybody's children. At that size freedom is cheap, and the absence of an org chart costs nothing, because the org chart was doing no work in the first place. What the genre rarely supplies is an account of what is quietly performing the function the missing hierarchy used to perform—because something always is. Structure removed from the diagram does not leave the building; it relocates somewhere less visible.
If a firm runs on freedom, what is underwriting the freedom?
Chris McGowan is the founder and chief executive of Thunder Labs, an Australian firm working substantially in identity and access management, and at the time of our conversation employing somewhere between twenty and twenty-five people. He came to the recording arguing for the flattest possible reading of organisational life—freedom to act, deterministic chaos, minimal control—and finished it conceding something very close to the opposite. What makes the hour worth an article rather than a summary is that he noticed himself doing it, said so aloud, and did not attempt to reconcile the two positions on air.
What follows are the ideas from the conversation I have continued to turn over since.
The case for the Networked Firm is Epistemic, not Cultural
Asked to justify the structure to a sceptical finance executive in a sentence, McGowan did not reach for culture, engagement or belonging. He said that no individual is capable of solving the problem, and—the sharper half of the claim—that it is close to impossible for any organisation to acquire the thinking it needs and then hold on to it. The structure, on his account, is not a preference. It is forced.
This is a stronger argument than the one usually made on flatness's behalf, and it is an old one. Friedrich Hayek (1899–1992) argued in 1945 that the knowledge required to allocate resources never exists in concentrated or integrated form, but only as dispersed, incomplete and frequently contradictory fragments held by separate individuals. The constraint is not that central planners are lazy or ill-intentioned; it is that the knowledge is not available to be centralised. McGowan's version—that any room making a decision needs the technologist, the procurement person, someone with a view for law and someone with a view for the commercials, and that no individual holds all of it—is Hayek's problem restated at the scale of a meeting.
Where the analogy strains is instructive. Hayek's dispersed knowledge is coordinated by prices, which correct continuously and without anyone's permission. McGowan's is coordinated by an assembled group of advisers, which is a market in advice with no price signal in it. He raised the difficulty himself, in a different context, and did not connect it back: how, he asked, can you trust any organisation that is selling something, when the expansive conversation converts so reliably into the pitch? It is the right question, and it applies to his own arrangement. If the founder is, by his own description, not a domain expert in anything, and if every conversation he has therefore remains conceptual until a practitioner takes it further, what mechanism tells him when the practitioner is wrong?
His best answer is that the selection is the expertise—twelve years in recruitment, and by his own reckoning several thousand people hired, spent learning to identify who will disagree with him usefully. That is a serious answer. But it concentrates the firm's entire error-correcting capacity in one man's judgement of people, which is precisely the form of centralisation a flat structure cannot dissolve. The hierarchy was not removed. It was reduced to a single, invisible decision made repeatedly at the hiring gate.
Bureaucracy is not the opposite of Freedom; it is what Fills the Vacuum
McGowan's objection to structure is not to its existence but to its metastasis: get an administrator's hands on something and they will, in his phrase, admin it to death. He put it more carefully later—structures naturally assert themselves, and it is very difficult to be unstructured without being chaotic.
Michel Crozier (1922–2013) spent two French public bureaucracies establishing why, and published the results in 1964. His finding was that bureaucratic rigidity is not a failure of training or a defect of character. It is rational behaviour by people who control some zone of uncertainty and use procedure to defend it, producing self-reinforcing vicious circles that no one individually intends. I made a version of this argument on air—that bureaucracy is often heaviest at junior levels, because procedure is the only authority available to someone who has no other, and that the checklist derives its force from having been authorised elsewhere. Crozier's point is the one that ought to worry any founder attracted to flatness: removing the hierarchy does not remove the impulse. It relocates it.
The consequence is that the anti-bureaucratic firm is not less political than the hierarchical one. It is differently political, and in one specific respect worse. Formal authority is contestable. There is a person whose decision it was, a rule they applied, and—at least in principle—somewhere to take the objection. Influence exercised through a network of trusted relationships has no equivalent. There is no appeals process against not being one of the people the founder rings. McGowan did not address this, and I did not press him on it, which I now regret; it is the obvious question to put to any organisation that has replaced titles with trust. He may well have an answer. The conversation does not contain it.
Language is the First Work, and it Does not Scale
The most practically useful passage of the hour concerned vocabulary. McGowan's firm builds in modules—think Lego—and prefers two-way doors unless it decides otherwise with its eyes open, and keeps the blast radius low. He is explicit that the lexicon is made jointly rather than issued, and that it must be ubiquitous rather than cultish: language everyone can already use, including clients, not an in-group dialect that certifies membership.
Karl Weick (1936–2026) spent a career arguing that organisations do not merely use language but are constituted through it: organising is the ongoing reduction of equivocality through talk, and the vocabulary available to a group determines what that group is able to notice. On that reading, McGowan is describing the first work of the firm rather than a soft adjunct to it. A team with the phrase two-way door can see reversibility as a property of a decision. A team without it argues the decision itself, repeatedly, never naming what it actually disagrees about.
His own examples make the point better than he did. Two-way door is Amazon's; blast radius is borrowed from engineering. Both entered his vocabulary because they already existed in the wider vernacular—which was my argument to him: a term society already understands is adopted faster by whoever joins after the conversation ended.
And that is the limit. He named it himself: whoever joins next month was not present when the language was made, and will encounter it as inheritance rather than authorship. At twenty-five people the practice is available because everyone can be in the room. At two hundred it is not, and what remains is a vocabulary distributed downward from the founder—the precise thing he objects to when a chief people officer does it. The practice may be a function of size rather than of philosophy, and the honest version of his claim is narrower than the one he makes: this is how you build a lexicon in a small firm, not how you run a large one.
Necessity and Motivation Are not a Hierarchy
McGowan draws a distinction I have not seen elsewhere in quite this form, and it is his best original contribution to the hour. Motivation and necessity are different things. Necessity need not be financial—it can be emotional, or ego, or a burning platform—and the two behave badly on their own. Too much necessity is desperation. Motivation without it is unbridled ambition. He wants people who have both, held in tension, and reads the tension as the interesting part of a person rather than a problem to be resolved.
He is candid about his own case: a business idea he pursued for five years, his money gone, his family placed in financial jeopardy, and a landlord to pay ever since. His necessity is the rent. His motivation is building something good. He is aware that their present alignment is a happy accident rather than a design.
Edward Deci (1942–2026), with Richard Ryan, set out the case in 1985 that intrinsic and extrinsic motivation are not two grades of the same substance but distinct systems, and that external contingencies can actively corrode internal ones rather than simply adding to them. This matters because his tension is unstable in one direction: necessity does not merely sit alongside motivation but, left dominant, consumes it.
Which brings us to the hour's most awkward joint. For seventy minutes McGowan talks about freedom, security, families, meaning, and being a servant of the company rather than its owner. He describes a friend of twenty years asking him why he articulates his purpose in revenue when everything he says privately is about safety and families, and he says he wants to provide for fifty families. Then, in the lightning round, asked what single rule he would keep if the company had to be simplified, he answers without hesitation: absolute focus on growth and revenue as the priority. Asked what practice he would delete tomorrow: everything that stops him getting growth and revenue.
His best answer—and it is a good one—is that these are the same object described twice. Fifty families is a revenue figure. Revenue is a families figure. A business that does not make money provides for nobody, and he articulated the goal in revenue for four years because revenue is the thing an organisation can actually be held to. He is also disarmingly honest that his own team would find him more brutal on numbers than outsiders might expect.
But the two descriptions only coincide while the business is growing. They diverge the moment they cost each other something—a redundancy, a client one would rather not keep, a year in which the number can be made or the fifty can be protected but not both. On the evidence of the lightning round, we already know which one is load-bearing. The families are the espoused value; the number is the operating assumption. I do not think he is being dishonest about this. I think he has not yet had to choose.
The Freedom Was Underwritten by Controls He Calls Luck
Late in the conversation McGowan reversed himself, and did it well. He believes in freedom to act, he said, and in not putting controls on people—and yet he observes that where there are no controls, circumstance or temptation does the work instead. Then the concession: another twist of luck, that he happened to put good managers around himself. Brilliant accountants. A bookkeeper who chases a questionable receipt through his own company and asks what it was for.
He went further than most founders would, volunteering unprompted that he can see how embezzlement happens—that with the sums now moving through the business the opportunity is simply visible, and that perhaps seeing it so clearly is part of why it holds no appeal. He said plainly that he has never been tempted. What he was describing is not a confession; it is an unusually clear-eyed account of why internal controls exist.
Donald Cressey (1919–1987) interviewed convicted embezzlers in prison and reported in 1953 that three conditions had to coincide: a financial problem the person felt unable to share, a perceived opportunity, and a vocabulary of adjustment that allowed the act to be described as something other than theft. Cressey never called it the fraud triangle; that label was applied later. His central finding is the one that matters here—his subjects were not people of criminal disposition. They were trusted people whose circumstances changed. Which is precisely why I am an honest person is not a control, and why McGowan's instinct to hire the bookkeeper before he needed her is better governance than his stated philosophy can account for.
I quoted Acton at him—power tends to corrupt, and the worse heresy, that the office sanctifies the holder of it. McGowan's own formulation was blunter and, I think, more useful: the running internal dialogue, the fear of coming to believe his own publicity, the distinction between being entitled to the company's assets and being a servant of them. His worry is not that he lacks a conscience. It is that willpower wanes, something eventually happens, and structure has to hold when it does.
It is a version of something I have argued elsewhere: that risk discipline is largely a function of how close the consequence sits to the person making the decision, rather than of character or governance maturity. McGowan's bookkeeper puts the consequence in the room. The receipt is queried this week, by a person he sees, about a sum too small to matter. That proximity is doing more work than any values statement, and it was assembled almost by accident.
So the honest reading of the hour is this. The freedom is real, and he has built something genuinely unusual. But it is underwritten by controls he did not design, hired for other reasons, and describes as luck. He is right that it was luck. The uncomfortable implication—which he came close to naming and then let go—is that the flat firm's case for itself rests on a control structure its own philosophy gives it no way to specify. To his considerable credit, he said so at the point in the conversation where he had every opportunity to end on the flattering version instead.
Who Should Listen
Founders of firms between ten and fifty people who run on trust and have not yet worked out what happens when trust is insufficient. Executives in larger organisations who suspect their bureaucracy is a symptom rather than a policy, and want a diagnosis. Board members and audit committee chairs will find the final twenty minutes more useful than most governance training, being a founder reasoning his way toward controls rather than being subjected to them. Anyone who has ever written a purpose statement in one vocabulary and a target in another will recognise the problem in the lightning round.
Good night, and good luck.
