The Concentration Doctrine

Four Dependencies, and Why Concentration Is Not a Risk to Be Reduced but a Structural Position to Be Priced and Chosen.

Concentration is not a risk to be reduced. It is a position to be priced.

The operator economy has one answer to concentration and delivers it with the confidence of arithmetic. Diversify. Add customers. Qualify a second supplier. Open another channel. Remove the key person. The answer arrives as though it were free, as though the only thing standing between the operator and a durable architecture were the discipline to spread the weight more evenly.

The answer is not free. Every diversification carries a cost, and the cost is almost never named in the room where the advice is given. The advice is given by parties who do not carry the cost.

This essay introduces the framework. The Concentration Doctrine. Four dependencies. Customer. Supplier. Channel. Key person. Each dependency is a position, not a hazard. Each has properties the architect audits. Each has a dominant failure mode that has nothing to do with the concentration itself and everything to do with the architect’s refusal to price it.

The doctrine’s central claim is simple and unpopular. Concentration chosen is a strategy. Concentration inherited is an exposure. The two look identical in the accounts and behave nothing alike under pressure. The difference is entirely a matter of whether the architect has priced the position or merely arrived at it.

This is Volume VIII’s foundational lens. Every subsequent essay this week refers back to it.

Why the diversification framing fails

Three reasons the diversification framing fails the architect.

First reason. Diversification is costed as though it were free. It is not. Each additional customer consumes account attention, pricing attention, contractual overhead and the architect’s own hours. Each additional supplier consumes qualification, negotiation, quality supervision and a second set of failure modes. Each additional channel demands a different operating discipline and a different economics. The operator who spreads across four customers, four suppliers and three channels has not reduced exposure. They have converted one legible dependency into eleven illegible ones and paid for the conversion in margin, attention and depth. The Pricing Architecture (Volume IV) established that price is a function of the position the architect occupies. Reflexive diversification degrades that position in every direction at once.

Second reason. Diversification treats all concentration as the same object. It is not the same object. A customer carrying seventy percent of revenue under a multi year agreement, embedded across four departments, with a replacement cost measured in years, is a structurally different thing from a customer carrying seventy percent of revenue on a purchase order that renews monthly through one procurement officer. Both read as concentration in the accounts. One is a fortress. One is a cliff. A framing that cannot distinguish between them is not a framing. It is a reflex.

Third reason. Diversification has no terminal condition. There is no point at which the operator is finished diversifying. Every remaining concentration reads as an outstanding task. The architect who accepts the framing is committed to a permanent programme of dilution with no defined completion, and the programme will consume the architecture’s depth long before it consumes its risk. The Risk Posture (Volume V) established that risk categories are chosen and held, not eliminated. Concentration is the clearest case.

The structural framing corrects all three failures. Concentration is a position. Positions are priced. Priced positions are either held deliberately or exited deliberately. Nothing is left to reflex.

Dependency One. The Customer.

The customer dependency is the degree to which the architecture’s revenue is carried by a small number of buying entities.

It is the concentration everyone measures and almost nobody understands. The percentage is recorded, discussed at board level, flagged in every diligence process, and treated as a single number that carries all the meaning. The number carries very little of the meaning. The structure beneath the number carries all of it.

Four properties to audit at the customer dependency.

First. Share and shape. Not only what proportion of revenue the customer carries, but what proportion of gross margin and what proportion of collected cash. These three figures frequently disagree. The largest customer is often the thinnest, and the concentration that reads as dangerous in revenue reads as trivial in profit. Audit all three.

Second. Substitution asymmetry. How long would the customer need to replace the architecture, and how long would the architecture need to replace the customer. If the customer can substitute within a quarter and the architecture would need three years, the concentration is priced against the architect regardless of what the contract says. If the reverse holds, the concentration is an asset.

Third. Depth of embedment. How many people, systems, processes and reporting lines inside the customer touch the architecture. A relationship held at one point of contact is a thread. A relationship woven through four functions is a fabric. The percentage of revenue is identical. The durability is not.

Fourth. Decision surface. How many people inside the customer would have to agree in order to end the relationship. A concentration that one procurement manager can terminate is a different instrument from a concentration that requires four departments to align. The Counterparty Stack (Volume II) treated this as counterparty quality. Here it is measured as structural durability.

The dominant failure mode at the customer dependency is gratitude pricing. The architect, conscious of the concentration, begins to treat the customer as a benefactor rather than a counterparty. Terms soften. Scope expands without price. Renewals are conceded before they are negotiated. The architect converts the strongest structural position they hold into the weakest commercial one, and does it out of a sense of obligation that the customer never asked for and frequently does not share.

Dependency Two. The Supplier.

The supplier dependency is the degree to which the architecture’s ability to deliver is carried by a small number of upstream parties.

The category is wider than most operators map it. It includes manufacturers and wholesalers. It also includes capital providers, software platforms the operation cannot run without, licensors, insurers, landlords, and the specialised contractors whose absence would stop the work. Anything that can withdraw and halt delivery is a supplier.

Four properties to audit at the supplier dependency.

First. Substitutability and lead time. Not whether a replacement exists in theory, but how many weeks from this supplier’s failure to a qualified replacement operating at standard. The honest number is usually three times the estimated one, because the estimate omits qualification, retooling, relearning and the period during which the architecture delivers below its own standard (see Volume III’s Standard).

Second. Specification lock. How much of what the architecture sells is defined by this supplier’s particular way of doing things. A supplier who provides a commodity is replaceable. A supplier whose specifications have quietly become the architecture’s specifications is not a supplier. They are a partner with none of the obligations.

Third. The supplier’s own concentration. The supplier’s dependencies are the architecture’s dependencies at one remove. A sole supplier who themselves depend on a single upstream source, a single facility, or a single financing line has passed that fragility through to the architecture without disclosing it.

Fourth. Information symmetry. Whether the architecture knows the supplier’s capacity, cost structure and financial condition, or only its price list. The architect who knows only the price list is negotiating blind and will discover the supplier’s condition on the day it becomes a delivery failure. This is the industry tier of Wednesday’s framework doing structural work.

The dominant failure mode at the supplier dependency is assumed continuity. The architect models supplier risk as price risk. What happens if they raise prices. The question the architecture actually faces is what happens if they cease to exist, are acquired by a competitor, or decide the architecture is no longer a customer worth serving. Price risk is annoying. Existence risk is terminal, and it is never on the agenda until the week it arrives.

Dependency Three. The Channel.

The channel dependency is the degree to which the architecture’s access to its buyers runs through a small number of intermediating paths.

A marketplace. A distributor. A broker network. A retail chain. A tender panel. A referral source. A search index. A platform that decides who sees what. The channel is not the customer and it is not the supplier, and treating it as either produces persistent structural error.

Four properties to audit at the channel dependency.

First. Ownership of the customer relationship. Whether the architecture knows who bought, or only that something was bought. The architecture that cannot contact its own buyers does not have customers. It has traffic that the channel has agreed, for now, to send.

Second. Terms mutability. How unilaterally the channel can change the economics. A channel that can alter commission, ranking, visibility or eligibility without negotiation is not a partner. It is a landlord with an unwritten lease, and the rent review happens whenever the landlord decides.

Third. Attribution honesty. What proportion of demand the channel genuinely creates against what proportion it merely intermediates. Most operators overestimate creation dramatically. The test is uncomfortable and simple. If the channel disappeared, how much of that demand would find the architecture by another route within a year. The honest answer reprices the channel immediately.

Fourth. Exit path. What remains if the channel is removed tomorrow. Not the revenue that would be lost, which everyone can estimate, but the operating capability that would be left. A direct relationship list. A brand the market recognises independently (see Volume VII’s Brand Architecture). A second path already warm.

The dominant failure mode at the channel dependency is mistaken authorship. The architect believes the channel is a distribution decision when the channel has quietly become the business. The architecture’s product development, pricing, packaging and hiring have all reorganised around the channel’s requirements. The operator still describes the channel as one of several routes to market. It has not been one of several for years.

Dependency Four. The Key Person.

The key person dependency is the degree to which the architecture’s judgement, relationships and operating method are carried by one individual.

In most architectures at structural altitude, that individual is the architect. This makes the dependency the hardest of the four to audit honestly, because the auditor is the exposure.

Four properties to audit at the key person dependency.

First. Judgement concentration. Which decisions genuinely cannot be made in this person’s absence. Not which decisions are habitually made by them, which is a different and much longer list, but which ones would be made badly or not at all if they were unreachable for six months.

Second. Relationship custody. Which counterparty relationships are held personally rather than institutionally. The Trust Architecture (Volume V) established that trust is engineered through four layers and carried by the person who engineered it. That is exactly why relationship custody is a concentration. Trust does not transfer because an email address changes.

Third. Tacit method. How much of the operating method exists only in this person’s practice and has never been written down. The architect who cannot describe how the work is done has not built an architecture. They have built a habit with staff attached.

Fourth. Absence tolerance. How long the architecture operates at its own standard without this person. This property is not estimated. It is tested, deliberately, on a schedule, in conditions the architect chooses rather than conditions that choose the architect.

The dominant failure mode at the key person dependency is the flattering dependency. Being indispensable is pleasant. The architect enjoys the calls that only they can resolve and mistakes the enjoyment for evidence that the concentration is necessary. The Talent Architecture (Volume III) named this as the ceiling that no amount of hiring resolves, because the constraint is not who has been hired. It is what the architect has been unwilling to release.

The pricing of concentration

The doctrine’s substance sits here.

Every concentration buys something. Every concentration costs something. Every concentration requires a reserve. The architect who can state all three for each of the four dependencies has priced their position. The architect who cannot has an exposure they are describing as a strategy.

What the concentration buys. Depth of relationship. Margin the diluted competitor cannot hold. Speed, because fewer counterparties means fewer approvals. Access, because a large share of one buyer’s spend produces standing that a small share of many buyers’ spend never produces. A standard that can be held, because the architecture is not spread thin enough to lower it.

What the concentration costs. Optionality. Negotiating position at renewal. Transferability at exit, which the Exit Doctrine (Volume V) established as the property most operators discover too late. And a quantity of attention that the architect will spend on the concentration whether or not they choose to.

What reserve the concentration requires. Capital sufficient to survive the loss for the length of the replacement cycle, which is the honest replacement cycle from the second property above, not the optimistic one. A warm second option that is genuinely warm, meaning contacted this quarter rather than identified once in a planning document. A written method, so the architecture survives the key person. And a named trigger, agreed in advance, at which the concentration converts from held to exited.

This is the entire doctrine in one sentence. A concentration with a stated purchase, a stated cost and a funded reserve is a position. A concentration without them is a bet the architect has not admitted to placing.

The integration of the four dependencies

The four dependencies are rarely independent, and this is where most audits fail.

The concentrated customer is frequently reached through the concentrated channel, served using the concentrated supplier, and held personally by the key person. Four dependencies that appear on the register as four separate risks are, on examination, one risk wearing four costumes. The architect who diversifies one of the four while the correlation remains intact has purchased comfort rather than durability.

The correlation test is a single question. Name the one event that would move three of the four at once. If the answer arrives quickly, the architecture has one dependency, not four, and it should be priced as one. If no such event exists, the four are genuinely separate and can be held at higher individual concentrations than the register would suggest.

The second integration point is sequencing. The four dependencies do not fail simultaneously. They fail in a characteristic order. The channel moves first, because channels are governed by parties optimising for themselves. The customer moves second, usually in response to something that happened in the channel. The supplier moves third. The key person moves last, and usually by choice rather than by event.

An architecture that reads its own sequence knows where to place its instrumentation. Watch the channel most closely, because the channel is the earliest signal, and the signal it gives arrives quarters before the revenue moves.

The audit cadence

The Concentration Doctrine is audited quarterly on the numbers and annually on the doctrine.

The quarterly audit is mechanical. Four dependencies. Share, shape, substitution asymmetry, embedment, decision surface, lead time, terms mutability, absence tolerance. Numbers, not impressions. Movement quarter on quarter, not absolute levels, because it is the direction that carries the information.

The annual audit is doctrinal and harder. For each of the four dependencies the architect states what it buys, what it costs, what reserve it holds, and at what trigger it converts. Then the architect asks the only question that matters. If we were building this architecture again with today’s knowledge, would we choose this concentration deliberately.

If the answer is yes, the concentration is a position and the architect stops apologising for it.

If the answer is no, the concentration is an exposure, and the work of the next year is either to price it properly or to exit it on the architect’s timetable rather than the counterparty’s.

The absence test is run annually and in advance. The architect removes themselves from the architecture for a defined period, chosen rather than forced, and the architecture is observed. What breaks is the key person dependency, measured rather than assumed.

A note on the catalogue

The Concentration Doctrine sits alongside three frameworks already in circulation and sharpens each.

The Counterparty Stack (Volume II) asked who the architecture is exposed to. The Concentration Doctrine asks how heavily, and whether the weight was chosen. The Risk Posture (Volume V) established that risk categories are held deliberately rather than eliminated. Concentration is the category where that principle is least practised and most necessary. The Customer Architecture (Volume VI) described the layers of the customer relationship. The Concentration Doctrine measures how much of the architecture rests on any one of them.

The doctrine also stands behind an older piece of substance. The Economics of Refusal (Volume I) argued that the architect’s refusals define the architecture more precisely than its acceptances. Concentration is what refusal produces. An architecture that refuses well will be concentrated, because refusal removes the marginal counterparty and leaves the ones worth carrying. Diversification, pursued as a virtue, is often nothing more than a failure to refuse, recorded on a spreadsheet as prudence.

Doctrine close

Concentration is not a risk to be reduced. It is a structural position to be priced and chosen.

Customer. Supplier. Channel. Key person. Four dependencies, each audited on four properties, each carrying a dominant failure mode, each correlated with the others more tightly than the register admits. The architect states what the concentration buys, what it costs, what reserve it requires and at what trigger it converts. The architecture is then held on purpose.

The Circle is a room of architects who carry their concentrations deliberately, who price the positions they hold rather than apologising for them, and who recognise one another by the quality of what they have refused.

Structure. Access. Positioning.

The Architect Circle is application only. Two hundred and fifty seats globally. createyourownmarket.com.au.