Four Horizons at Which Capability Supply Is Read, and Why the Longest of Them Is Published Years in Advance and Examined by Almost Nobody.
An architecture can only hire from a population that somebody else decided to train, and that decision was taken years before the architect noticed they needed it.
This is the structural fact the recruitment conversation cannot hold. That conversation is conducted entirely at one horizon, the current quarter. The candidate pool is thin. The salary expectations have moved. The good people are not available. All of it is accurate reporting about a single horizon, and all of it is downstream of decisions taken at three longer horizons the architecture has never instrumented.
Monday’s framework named capability as one of four constraints and separated depth from ceiling. Wednesday’s framework established that reserve is held in four currencies. Neither addressed where capability comes from, and the answer is that it comes from four places on four clocks.
This essay introduces the framework. The Capability Horizon. Four horizons. The Immediate. The Trained. The Taught. The Transmitted. Each is read with a different instrument and moves on a different timescale. Each has a dominant failure mode, and the failure modes run in one direction, which is that the architect reads the shortest horizon and infers the other three from it.
This is the research aligned essay of Volume X and the only one this week. We orient against publicly visible dynamics in the Australian environment and then bend the analysis into the framework. The sources are named in the editorial notes at the close, cited for direction rather than for precise figures. No number appears here that the architect should rely on without checking the current release. That discipline is the point rather than a caveat.
Why the recruitment framing fails
Three reasons the recruitment framing fails the architect.
First reason. It reads a stock as though it were a flow. The candidate pool is a stock, and a stock is the residue of years of inflow and outflow. An architect examining the pool is examining an outcome, with no visibility into whether the inflow that produced it is rising or has stopped.
Second reason. It is conducted at the one horizon where the architect has no instruments and no leverage. In the current quarter the architecture can pay more, lower its requirement, or wait. Those are the three available moves and all three are expensive. Every genuinely cheap move on capability sits at a horizon measured in years, and the framing never arrives there.
Third reason. It cannot distinguish a shortage from a disappearance. A shortage is a stock temporarily low against demand that will be replenished. A disappearance is an inflow that has stopped, which produces an identical looking shortage followed by nothing. The two require opposite responses. A shortage is waited out and paid for in the interim. A disappearance is answered by the architecture producing the capability itself, and the production takes years that are only available to an architect who identified the condition early.
The structural framing corrects all three. Four horizons, four instruments, four failure modes, one annual reading.
Horizon One. The Immediate.
The immediate horizon is the capability the architecture could acquire this quarter.
It is the only horizon most architectures read and the least informative of the four, because by the time a condition is visible here it has been in motion for years. The instrument is direct observation. Time to fill. Offers declined and the reasons. The rate as actually paid rather than as advertised.
Four properties to audit at the immediate horizon.
First. Time to standard, not time to fill. The interval from the vacancy to a person performing at the architecture’s own mark. This is the number that matters and almost nobody records it, because the role is treated as resolved at the signature.
Second. Decline reasons. Why the people the architecture wanted did not come. Recorded, not recalled. The pattern in declines is the clearest available signal about the architecture’s position in the capability market, and it is free.
Third. The actual rate. What the last four hires cost against what the architecture believes the rate to be. These figures diverge quickly and the belief is always the older number.
Fourth. The internal alternative. Whether the role could have been filled from inside. If the answer is never, the architecture has no replenishment, which is a finding about the fourth horizon dressed as a finding about the first.
The dominant failure mode here is paying for a structural condition. The architecture meets a shortage with money, which works, and continues to work, at a rising price, for as long as the condition persists. Nothing adjusts, because the problem is being solved every time. The expenditure is treated as a market rate rather than as the cost of holding no position at the three longer horizons.
Horizon Two. The Trained.
The trained horizon is the population currently in formation. People who are in a course, an apprenticeship, a traineeship, a degree or a structured development role, who will reach the market in one to four years.
This horizon is knowable and almost never consulted. Enrolment, commencement, completion and composition are all published. An architect who reads them holds a four year view of the supply of the capability their architecture runs on, which is longer than the recruitment conversation can produce by any means.
The Australian picture here rewards reading rather than assertion. The National Centre for Vocational Education Research publishes regular collections on apprentices and trainees by occupational group. The Australian Bureau of Statistics publishes Education and Work annually. Jobs and Skills Australia publishes occupation level shortage assessments. None is proprietary, none is expensive, and the proportion of architectures reading any of them on a schedule is very small.
Four properties to audit at the trained horizon.
First. Pipeline volume for the specific capability. Not for the sector. The occupation the architecture actually depends on, which is usually narrower than the category the data is reported in.
Second. Completion against commencement. A pipeline with strong commencements and weak completions is smaller than it appears, and the gap is where most misreadings of this horizon occur.
Third. Direction across three releases. One release is a level. Three are a direction, and the direction is the information. This is the most valuable habit available here and it costs an hour a year.
Fourth. The architecture’s position in the pipeline. Whether it has any relationship with the institutions producing the capability, or intends to meet the cohort for the first time at the point of hire, against every other architecture that also did nothing.
The dominant failure mode at the trained horizon is the absent relationship. The architecture waits at the end of the pipeline rather than standing beside it. The people in formation have no awareness the architecture exists, so the architecture competes for them entirely on price at the moment of entry, which is the most expensive point at which any capability can ever be acquired.
Horizon Three. The Taught.
The taught horizon is whether the capability is still being taught at all.
This is the horizon published years in advance and read by almost nobody, and it is the one with genuinely decisive consequences. A course closes. A trade qualification is restructured or retired. A licence class changes. A discipline is consolidated into a broader faculty. An apprenticeship pathway loses its funding arrangement. Each is announced in public, usually with years of notice, and each removes the second horizon some years later and therefore the first horizon some years after that.
This was identified once already, as the fourth property of a single signal, in the observation that a category whose skills are no longer taught has a terminal date calculable from the age distribution of its workforce (Volume VIII’s Obsolescence Audit). Here it is a horizon in its own right, because the taught horizon governs capability supply for every architecture, including the ones whose categories are growing.
The Australian environment supplies orientation here through several public bodies. The Productivity Commission’s five year productivity inquiry, published in 2023 as Advancing Prosperity, examined education and training supply against the economy’s capability needs. Treasury’s 2023 Intergenerational Report sets out the workforce’s demographic trajectory across decades. The OECD’s Economic Surveys of Australia return to skills formation regularly. These are frames rather than operational instruments, and that is the correct use of them.
Four properties to audit at the taught horizon.
First. Whether the pathway exists. Named specifically. The particular qualification, licence, apprenticeship or discipline the capability comes through, and whether it is currently offered.
Second. Announced change. Published timetables for closure, restructure, consolidation or funding change. These dates are public and are read almost entirely by the people directly affected, so an architect who reads them holds a view nobody in their market has.
Third. Who teaches it. The count of institutions offering the pathway, and the direction of that count. A capability taught in one place has a single point of supply.
Fourth. The age distribution of the existing workforce. If the holders are concentrated in one cohort and nobody is being taught, the capability has a date, and the date is arithmetic rather than opinion.
The dominant failure mode at the taught horizon is the policy posture. The architect, discovering a pathway is being withdrawn, joins the campaign to preserve it. The campaign is legitimate and frequently correct. It is not a capability strategy. Announced dates do move and the direction does not, and an architecture that spent four years on the response and none on its own transmission has converted a decade of warning into a delay.
Horizon Four. The Transmitted.
The transmitted horizon is whether the architecture can produce the capability itself.
This is the only horizon the architect fully controls and the only one most architectures have never attempted. It asks whether a less capable person can become a capable one inside the architecture, by design, through a method that exists independently of whoever happens to be teaching.
The four layers of the team inside an architecture are already established (Volume III’s Talent Architecture). This horizon is upstream of all four. It asks not how the team is structured but whether the architecture produces capability or only consumes it, and the distinction determines the architecture’s long position, because a consumer is permanently exposed to three horizons it does not control.
Four properties to audit at the transmitted horizon.
First. Whether the method is written. Not a procedure document. The reasoning behind the work, recorded in a form a capable newcomer can extend rather than only follow. A document recording the rule without the reason produces instructions that are followed until they stop working and then abandoned (Volume IX’s Institution You Are Actually Building).
Second. Whether anyone has actually made the transition. The count of people who entered below the standard and now hold it. If the count is zero across a decade, the capability does not exist, whatever the intention.
Third. Time to standard, internally. How long the architecture takes to produce a capable person from a promising one. This is the architecture’s most important operational figure and is almost never known.
Fourth. Whether transmission survives the transmitter. If one person teaches and nobody else can, the architecture has an apprenticeship with a single master, which is a key person dependency in the dimension that compounds most slowly.
The dominant failure mode at the transmitted horizon is the deferred investment. Transmission is always correct to begin and never urgent to begin. It produces nothing in the quarter it starts, costs capacity that is currently billable, and pays in year four. Architectures defer it, reasonably, every quarter, for a decade, and arrive at their second decade as pure consumers of a supply they have watched contract at three horizons.
The integration of the four horizons
The horizons are sequential and causal, which is what makes the framework diagnostic rather than descriptive.
The taught horizon determines the trained horizon some years later. The trained horizon determines the immediate horizon some years after that. The transmitted horizon is the only one operating independently of the other three, which is precisely why it is the one worth building.
The detection rule follows from the sequence. A condition visible at the immediate horizon has been present at the taught horizon for most of a decade, and the information was published and free throughout.
The framework also maps onto an instrument already in the catalogue, and the mapping is diagnostic. The four tiers of what an architect knows before the market does (Volume VIII’s Information Architecture) each read a different horizon here. The public tier reads the taught horizon, where the announcements live. The industry tier reads the immediate horizon, because the rate actually paid and the real reasons for declines circulate inside a trade and are never published. The relational tier reads the trained horizon, because a relationship with the institutions producing capability converts published enrolment data into named people. The proprietary tier is the transmitted horizon.
An architecture weak at a given tier will be blind at the corresponding horizon, and blind without knowing it.
What the reading produces
The reading produces one of three positions.
Purchase. The trained horizon supplies the capability abundantly, the taught horizon is stable, and the architecture can reasonably continue to buy. Most architectures are in this position for most of their capabilities. It should be held knowingly, with the current cost recorded, rather than by default.
Partner. The trained horizon is adequate but contested, and the cheapest move is to stand beside the pipeline rather than at the end of it. Relationships with the institutions producing the capability, work experience arrangements, visible presence in front of the cohort. Inexpensive, slow, and it produces a position at the one point where every competitor is bidding on price alone.
Produce. The taught horizon is contracting, or the capability is specific enough that nobody else will ever teach it. The architecture builds transmission. This is the expensive position, the only one that holds across decades, and it is funded from a bucket with a matching horizon (Volume IV’s Capital Allocation Doctrine).
The common failure is a fourth path, which is to continue purchasing while the taught horizon closes. It is not a decision but the absence of one, and it ends with an architecture bidding an escalating price for a shrinking stock of people, in a capability it could have begun producing a decade earlier.
The audit cadence
The Capability Horizon is read annually, in one sitting, against the published sources.
The reading covers four questions, one per horizon. What did the last four hires cost and how long did they take to reach standard. What is the direction of the trained pipeline across three releases. Has anything been announced at the taught horizon, and when does it take effect. How many people did the architecture itself move from below the standard to the standard this year.
It is performed for each capability the architecture genuinely depends on, usually three or four rather than the full list of roles. Secondary capabilities are read every second year, and they are where the taught horizon most often moves without notice, because small pipelines close quietly.
The output is one page. Four horizons, a direction for each, and one sentence naming which of the three positions the coming year executes. A reading that produces no position is not a reading. It is a document.
A note on the catalogue
The Capability Horizon completes something three existing frameworks each leave open.
The four layers of the team structure the people already inside the architecture. This instrument reads whether such people will continue to exist in the market, which is the question upstream of structure.
The four signal audit asks whether the ground beneath a category is moving, and touched the training pipeline as one property among sixteen. Here the pipeline is the instrument rather than an indicator, and it is read for healthy categories as well as declining ones.
Monday’s Constraint Architecture distinguished depth from ceiling. This framework supplies the supply side. An architect whose binding constraint is capability now has a second question. Is the constraint a stock condition they can purchase out of, or an inflow condition they will have to produce out of. The two answers lead to entirely different decades.
Doctrine close
An architecture hires from a population somebody else decided to train, and that decision was taken years before the architect knew they needed it.
The Immediate. The Trained. The Taught. The Transmitted. Four horizons, four instruments, four failure modes, read annually in one sitting against sources that are public and free.
The horizons are causal. The taught horizon sets the trained horizon, which sets the immediate horizon. A difficulty visible this quarter has been published for most of a decade. Only the transmitted horizon is under the architect’s control, and it is the one that is never urgent.
The reading produces one of three positions. Purchase, partner or produce. It never produces a plan to pay more indefinitely.
The Circle is a room of architects who read the long horizons while the short one still looks manageable, who stand beside the pipeline rather than at the end of it, and who recognise one another by the number of capable people their architectures have produced rather than purchased.
Structure. Access. Positioning.
The Architect Circle is application only. Two hundred and fifty seats globally. createyourownmarket.com.au.