The number of leaders an enterprise can grow is capped by the composition of its work portfolio rather than by its development budget, which means a capability shortfall can be simultaneously acute and unbuyable.
An enterprise short of leaders has an obvious response: increase the development budget. Approve the academy, widen the cohort, buy more coaching, add a tier to the emerging-leader course — the money is usually findable and the intervention visible.
The tension is that the enterprise already suspects this will not work, because it holds the other half of the proposition just as firmly. Ask any executive where they learned to lead and the answer describes a job, not a course — a first time carrying something they were not ready for, in front of people who would see the outcome. That conviction is near-universal, and at odds with what the budget funds.
Both beliefs are held at once and never reconciled, because reconciling them produces an uncomfortable conclusion. If leadership is built principally by assignments, capacity is set by the supply of suitable assignments. Assignments are not purchased; they are thrown off by the work the enterprise happens to hold. They arrive when the portfolio produces them, cannot be stockpiled, and develop one person at a time. The pipeline therefore has a ceiling, and the ceiling is a property of the portfolio.
The same contradiction sits in the field's own development material. It establishes that the assignments building leadership carry unfamiliar responsibility, genuine latitude and visible consequence — then cautions that some roles are too important to be used developmentally, and that identifying which is the hard part. Elsewhere it prescribes selecting by developmental need, and prescribes offering opportunities to whoever is ready at the relevant career point. Those rules choose different people: readiness is disqualifying under the first and qualifying under the second, and nothing prices the difference.
The Strategic Context
Boards treat leadership capacity as an input they can procure — reasonable for most inputs, wrong for this one. It is better understood as a manufactured good with one production line the enterprise does not control: the portfolio of work in flight.
A developmental assignment has four economic properties unlike those of a training place. It is indivisible: a first time holding a profit and loss cannot be shared between four people. It is rivalrous — one holder per period. It is non-storable, existing only while the work does, so a year with no new plant, region or service line generates none. And it is joint-produced with delivery: every developmental placement is also a delivery decision carrying a real chance of a worse outcome.
Together those properties mean a quantity constraint, not a price constraint: no budget converts a portfolio of repeat work into one that grows general managers. Unlike a shortfall in specialist engineering, which can be bought in a market, a leadership shortfall often cannot be produced at any price inside the enterprise that needs it.
What Leaders Commonly Misread
The first misreading is that a stretch assignment is the same as a demanding one. Enterprises with heavy workloads assume they are developing people continuously, since everyone is under pressure. Pressure is not the active ingredient; latitude is. The holder must be able to choose an approach, be wrong, and correct. Work that is demanding but tightly specified and intolerant of variation produces compliance, not judgement. A fully occupied enterprise can run at capacity for years and grow almost nobody.
The second misreading is that challenge and opportunity are the same good. Assignments consisting mainly of adverse conditions — an inherited dispute, a failing supplier, a programme already off track — are hard and build resilience. They do not reliably build what leads to advancement, which comes from creating something, carrying visible stakes, and influencing people over whom one has no authority. An allocation practice distributing difficult work evenly and consequential work narrowly will look fair and will not be. Expect that pattern in your own records: nobody designs it and nothing detects it.
The third misreading is that a deputy or shadow arrangement is a developmental assignment. Usually it is not: where the deputy carries no decision that would stand if the principal disagreed, the arrangement transfers information and withholds consequence, which is what develops people. Deputy structures can be made developmental, but only by allocating real decision rights within a defined boundary. The support around an assignment follows a different logic again: formalising those relationships widens access and lowers the return, which [Related article: Formalise the Relationship and You Widen Access and Lower the Return] examines and this article does not, being concerned with the supply of assignments rather than the relationships around them.
Reframing the Issue
Properly reframed, this is a supply question with a governance answer, not a budget question with a spending answer.
The supply side is the portfolio: what work will the enterprise hold, and how much can safely be held by someone who is not the most capable available person? The demand side is the succession plan: how many leaders must have been grown by when, and at what level? Almost every enterprise maintains the second half of that calculation and has never constructed the first.
The reframing also moves ownership. If leadership capacity is a by-product of portfolio composition, the decisions that most affect the pipeline are made not by the people function but in bids, capital approvals, divestments and operating-model changes — forums that never consider development, because it is understood to live in a different budget.
Why the Portfolio Sets the Ceiling
The assignment is a by-product, not a purchase
The enterprise chooses its work for commercial reasons; developmental capacity is what falls out. A business winning one genuinely new type of work every few years produces a handful of first-time transitions and no more, whatever it spends. One that has diversified its delivery model, entered adjacent segments or runs discrete bounded campaigns produces many. Neither outcome was a development decision, and in most enterprises neither was reported as one.
Consider a mining and mineral processing operator, hypothetically. Its portfolio is overwhelmingly steady-state operations and sustaining capital, with a major expansion once in seven or eight years. Demand for site and functional leaders is continuous; supply of first-time general-management assignments arrives in lumps, separated by years when the available work is high-consequence and low-latitude — a processing superintendent cannot experiment with recovery rates. The recurring exception is the shutdown or turnaround campaign: bounded, visible, consequential and recoverable. An operator that sees this treats turnaround leadership as its principal school; most treat it as an operational necessity given to whoever did it last.
Composition determines type, not only count
Counting is not sufficient, because different types build different capabilities: first transitions build the ability to hold a role rather than a task; scale increases build delegation; ambiguous work builds judgement under incomplete information; cross-boundary work builds influence without authority; recovery work builds composure.
An enterprise whose portfolio produces only one type will grow leaders capable in one dimension and untested in the others, discovering which is missing at the point of appointment. Fractional allocation does not solve it: the arithmetic of splitting people across concurrent commitments belongs to Article 23, which owns that calculation, and an assignment held at a quarter is not an assignment.
Two selection rules that choose different people
Every allocation is governed by two rules the enterprise holds at once: one selects the person who will develop most, the other the person most likely to deliver. In the ordinary case they choose different people, and the second wins silently, because it is the rule with a consequence attached this quarter.
The result is a pipeline that thins precisely when the enterprise is busiest — which is when the shortfall becomes visible. The honest response is not to harmonise the rules but to decide in advance, by name, which assignments are governed by which — accepting that on the developmental ones the enterprise has bought a lower expected delivery outcome for capacity it cannot obtain any other way. That trade must be written down, because a trade nobody records gets reversed under pressure.
Decision Framework
The instrument is the assignment inventory: a counted, classified and governed list of the developmental assignments the portfolio will actually generate over the next twenty-four months, refreshed at each portfolio decision and owned by the executive who owns the portfolio.
Step one: qualify. List every role or discrete piece of work commencing in the window. An item enters only if all three tests hold: the holder would carry responsibility they have not held before; would have authority to choose the approach; and the outcome would be visible above their line manager. Two out of three is not an assignment.
Step two: classify by type. Tag each item as first transition, scale increase, ambiguity, cross-boundary influence or recovery. This turns a number into a capability profile and exposes the dimension your future leaders will lack.
Step three: rate protectability. Score each item as protectable — a below-expert holder is survivable, failure recoverable and contained — or critical, where the most capable available person must hold it. Only protectable items are inventory. This is the step enterprises skip, and it determines the real number.
Step four: compute the yield. Divide protectable items by the leaders the succession plan requires in the same window. Below one, the shortfall cannot be closed by spending and the succession plan is a statement of hope.
Step five: choose the response explicitly. Four responses exist, each with a price. Change the portfolio, bidding for or retaining work that generates protectable assignments and treating that yield as part of the case. Manufacture assignments, carving genuine decision rights out of critical roles and funding the double-running required. Buy leaders externally, pricing integration and retention risk honestly. Or reduce demand, changing the operating model so fewer leaders of that type are needed. Doing none of these is also a choice, and the one most enterprises make.
The governance test. No item enters the inventory unless a named executive has recorded, before allocation, that they will let the holder choose the approach and absorb a defined failure. An assignment nobody has agreed to protect is not protectable, whatever the classification says.
From Strategy to Execution
Immediate. Build the inventory for the next twenty-four months and compute the yield. Expect the protectable count to be a fraction of the initial list; the gap is the most useful figure the exercise produces. Take it to the portfolio forum, not the people committee.
Medium-term. Put the inventory into bid and capital approval as a standing line: what developmental assignments does this work generate, of what type and level? Fund the deputy structures that manufacture assignments inside critical roles — noting that the double-running they create is systematically unfunded, a pattern examined in [Related article: Nobody Funds the Overlap] and left there, since this article concerns only the ceiling the portfolio sets.
Long-term. Report assignment yield as a portfolio characteristic alongside margin and risk. Where yield is structurally below one, decide whether to hold a smaller, less profitable line of work because it functions as a school — and price that openly rather than defending it as a marginal business.
Consider a network of community legal centres, again hypothetically. Its centre directors approach retirement together, its development budget is small but not the binding constraint, and almost all its work is casework: indivisible, high-consequence for clients, no latitude. Its only protectable assignments are a funding submission, a service-model pilot and a back-office consolidation. That is the true size of its pipeline, and no training expenditure alters it.
Signals to Monitor
Watch the ratio of external to internal appointments below the executive. A rising external share is usually explained as market competitiveness; more often it is the portfolio's yield showing up in the appointment record two or three years late.
Watch, too, which assignments went to the person who would develop most and which to the safest pair of hands: if the safe choice has won eight consecutive allocations, the enterprise has a delivery policy and no development policy.
Watch the type distribution: successors developed entirely through recovery work will be composed and unpractised at creating anything. And watch yield after a restructure, since consolidating roles removes the intermediate positions where transitions occurred. A reorganisation also destroys capability living between people rather than inside them — the subject of [Related article: The Capability That Lives in Pairs], not of this article, which counts assignments rather than relationships.
Questions for the Leadership Team
- Over the next twenty-four months, how many roles will commence carrying unfamiliar responsibility, real latitude and consequence visible above the holder's manager — and how many can we give to someone who is not our best available person?
- Which named executive has agreed, in writing, to absorb a defined failure so that a specific assignment can be developmental?
- Of our last ten significant allocations, how many went to the person who would learn most rather than the one most likely to deliver, and what did we tell ourselves each time?
- Which capability dimension is absent from every assignment our portfolio generates, and where will that gap first appear in an appointment?
- What did our last three bid, capital or divestment decisions do to our assignment yield, and was that effect stated anywhere in the papers?
- If our development budget doubled tomorrow, which specific leadership shortfall would it close — and can we name one?
Closing Perspective
The uncomfortable part is not that leadership is expensive, but that leadership capacity can be unavailable at any price, because the input is not money but a particular kind of work — and the enterprise may not hold enough of it.
That converts development from a spending decision into a portfolio decision, and puts it in a room where it has never been raised. Every bid won, every marginal line divested, every restructure consolidating two roles into one is a decision about how many leaders the enterprise can grow — taken by people who were never asked the question and would not have had the number if they were.
The responsibility this leaves is to make the number exist and put it before the forum that sets the portfolio. Until then the ceiling stands where the portfolio put it, and the development budget keeps buying activity against a constraint it cannot reach.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.