You can predict which parts of a multi-part capability programme will survive by asking which have an internal buyer with a commercial reason to keep delivering them — and this is a designable property, not a failure of will.
The components of a large capability programme still running five years after approval are rarely the best-designed ones, nor the ones with the most senior sponsor, the largest allocation or the cleanest line to strategy. They are the ones somebody other than the programme office had a reason of their own to keep delivering once the central money stopped.
That inverts the order in which programmes are assembled. Design quality, sponsorship and funding are the three things a programme director controls directly, and all three are legible at approval. The property that actually predicts survival — whether a component has a buyer whose own interests are served by its continuation — is almost never assessed, and appears in no stage gate most enterprises run.
This matters at design rather than in recovery. While the work is still being scoped, a component without a buyer can be given one cheaply: change who delivers it, change what the deliverer receives, or fold it into something that already transacts. Once delivery starts, the same change means reopening the design in front of the people who approved it, by which point the component is usually in difficulty and the conversation becomes about competence rather than architecture.
There is a structural precedent worth describing, though not for its numbers. A government once commissioned a national review of management and leadership capability, which produced a large interlocking set of recommendations spanning schooling, enterprise support, workforce diversity, frontline supervision and management education; long afterwards it was formally revisited to establish what had been implemented. [FACT CHECK REQUIRED] What took hold were the recommendations where employers wanted the resulting qualification and providers could earn revenue supplying it. What went inert were those requiring coordination no single party profited from. And the coordinating body the architecture depended on was never established. The revisit read this as a stakeholder and governance failure; the structural reading is stronger, and the argument below does not need the case in order to stand.
The Strategic Context
A capability programme is not a large project. A project ends when its output exists; a capability programme is a set of interdependent components whose value arrives only once each has been absorbed into somebody's ordinary operating routine. That absorption happens in the portfolio, not the programme — which is why closure and survival are separate events, often years apart, and why a programme can be delivered in full and leave nothing behind.
Every component therefore has two funders in sequence: the programme funds construction, somebody else funds continuation. Most business cases model the first with precision and the second not at all, and the transfer between them is where components die. This article addresses whether a continuing funder exists, not the cost of the period in which both are notionally responsible and neither is paying: [Related article: Nobody Funds the Overlap] owns that double-running problem — transition programmes budget the state they are leaving and the state they are entering and fund the overlap out of neither.
What Leaders Commonly Misread
The first misreading is that survival is a question of will. When a component lapses, the enterprise reaches for commitment: sponsorship weakened, attention moved, the executive who cared was promoted. Such accounts are accurate in their particulars and useless as guidance, because they name the trigger rather than the condition. A component with a buyer survives the departure of its sponsor. A component without one survives nothing.
The second misreading is that market discipline substitutes for design. Devolving capability components to internal demand gets described as letting the business decide what it values. What happens is a sorting, and not by value: internal demand funds what somebody can price and capture, and does not fund shared goods — the common standard, the cross-unit forum, the reference data everybody uses and nobody owns — however much those are worth in aggregate. An enterprise that leaves its programme to internal demand is not testing value; it is filtering for appropriability and calling the result a verdict.
The third misreading is the most expensive: that the components most central to the strategy are the safest, because centrality attracts protection. Centrality attracts attention, which is a different currency. The most central components are frequently those whose benefits are widest and most diffuse — precisely the profile with no single buyer. Breadth of benefit and strength of buyer pull in opposite directions, so the programme's own logic of importance points away from what lasts.
Reframing the Issue
Reframe the question from what should survive to what will be transacted. Every component that outlives its programme does so because, in the ordinary course of business, one party delivers it and another gives that party something for doing so. The consideration need not be money: budget, headcount, priority in a queue, a credential that changes what somebody may be rostered to do, a licence condition satisfied, a customer requirement met. What matters is that the exchange occurs without anybody remembering the programme.
Take a hypothetical network of vocational training providers implementing a multi-part capability programme: a new supervisory qualification, a shared assessment standard, a cross-provider moderation forum, and an employer-facing skills brokerage. Three of those four have obvious buyers. Employers want supervisors who hold the qualification, so providers can sell it; the brokerage earns a fee. The shared assessment standard — the component that makes the other three defensible — is bought by nobody, its benefit distributed across every provider and captured by none. The outcome is predictable: the qualification thrives, the standard drifts, and within a few years the qualification means different things at different providers.
Where the Buyer Sits, and What They Are Buying
The buyer is a role, not a function
"The business" is not a buyer. Nor is "operations", "the network" or "our people". A buyer is a role occupied by a person who could describe in one sentence what they get and what they give up to get it. That is the only threshold separating a genuine transaction from an assumption. In a hypothetical regional freight rail operator, the buyer of competency verification is the depot manager who cannot roster a shift without certified crews; that manager keeps the assessment cycle running through a funding freeze because the alternative is idle rolling stock. The buyer of the cross-depot lessons-learned forum is nobody. Both were approved in the same programme, and only one is a going concern.
What the buyer is buying is usually permission, not capability
Enterprises design capability components around the skill being built. Buyers rarely purchase skill. They purchase what the skill entitles them to do: bid for a class of work, sign off a design, operate without a supervising engineer present, satisfy an insurer. The durable components are those where somebody's permission to act depends on the capability existing, because that permission is checked by a party outside the programme's control. Attaching a component to an existing external check — a customer's prequalification, a regulator's condition, an insurer's schedule — converts an internal preference into an external requirement, and the component acquires a buyer it did not previously have.
Some components can be given a buyer; some cannot
Three moves create a buyer where none exists. Change the deliverer, so the work sits with a party that captures value from doing it. Change the consideration, so delivering releases something the deliverer wants: access, priority, a rate, a lighter oversight burden. Or bundle it, so the orphaned component becomes an inseparable condition of one that already transacts — as an assessment standard becomes inseparable from a credential that cannot be issued without it. Bundling is the underused move and the strongest, because it survives personnel change.
Some components resist all three, and coordination is the standing example. Whether the enterprise has the leadership capacity to run the survivors is a separate constraint, and [Related article: Your Pipeline Is Capped by Your Portfolio, Not Your Budget] owns it — that the number of leaders an enterprise can grow is capped by the composition of its work portfolio rather than by its development budget.
Decision Framework
The internal buyer test. Run it on every component before approval, and retrospectively on any programme in delivery. One page per component, and it produces a decision rather than a discussion.
Name four things. The deliverer: the role that performs this work in the first ordinary operating period after programme funding ends. The buyer: the role that receives something it would otherwise have to obtain another way. The consideration: what passes from buyer to deliverer — money, budget, headcount, priority, a permission, a credential. The independence date: the date after which that consideration flows without programme money. Then classify against three thresholds.
| Classification | Test result | Required action before approval |
|---|---|---|
| Self-funding | Deliverer, buyer and consideration all named as roles; consideration continues past the independence date | Approve, and remove the component from the programme's forward cost base at that date. Hold the sponsor to it. |
| Subsidised | A buyer is named, but the consideration is programme money | Convert it — change the deliverer, change the consideration, or bundle it — or approve it as a permanent operating cost with a named budget holder and an annual review. |
| Orphaned | No buyer can be named as a role, or the buyer named is the programme itself | Do not approve as a standalone component. Fund it as declared central overhead with an owner and a sunset review, or cut it and record why. |
Two rules make the test bite. First, no component may be scored self-funding on the strength of a function: if the buyer cannot be named as a role held by a person who could state what they get, the score is orphaned. Second, coordination components always score orphaned — the coordinating body, the shared standard, the common register, the cross-unit forum. That is a finding, not a failure. Every component benefits from coordination and none will pay for it, leaving two honest options: fund coordination permanently from the centre with a named owner, or design the programme so its components do not require it. Approving a coordination component and expecting it to find a buyer is the specific error that hollows out capability programmes, and it is avoidable at design.
From Strategy to Execution
Immediate. Take the capability programme consuming the most executive attention and score every component against the three thresholds in a single session, with the programme director and sponsor present and named roles required for every buyer. Expect a material share to score orphaned, coordination components among them. Publish the scores unaltered.
Medium term. Add the test to the stage gate: no component enters delivery without a named buyer, a named consideration and an independence date in the business case, and the gate paper carries the classification on its front page. Where a component scores subsidised, require the conversion attempt to be documented — which of the three moves was tried, and why it failed — before permanent overhead is granted.
Long term. Change what programme directors are assessed on. A director assessed on delivery within programme funding has no reason to care whether anything transacts afterwards; one assessed on the proportion of components still transacting two operating periods after closure designs differently from the first week. That change in the assessment basis does more for capability survival than any amount of governance. It also alters how developmental and mentoring components are structured, though not whether formalising them is wise — [Related article: Formalise the Relationship and You Widen Access and Lower the Return] owns that trade, showing that an enterprise formalising its developmental relationships buys wider access and worse outcomes.
Signals to Monitor
The clearest evidence a buyer exists is the deliverer beginning to charge for a component, or defending its schedule against competing demands. Watch whether the component appears in an operating budget the programme did not write, and whether the capability shows up unprompted in the enterprise's own job advertisements. Measure attendance at shared forums through a peak trading period rather than a quiet one, since attendance under pressure is the only meaningful reading. And watch the minutes: a coordinating forum that has stopped recording decisions and started recording updates has ceased to function and is waiting to be cancelled.
Questions for the Leadership Team
- For each component of our largest capability programme, which named role delivers it in the first operating period after funding ends, and which named role gives that person something for doing so?
- Which of our components are coordination goods, and have we funded them centrally or quietly assumed the business will carry them?
- Of the capability programmes we closed in the last three years, which components still transact today, and what do the survivors have in common?
- Which existing external check — a customer prequalification, a licence condition, an insurance requirement — could be attached to a component so continuation stops being discretionary?
- What are our programme directors actually assessed on, and does any of it extend past programme closure?
- Which component of our current programme is most central to the strategy, and can anyone name its buyer without saying "the business"?
Closing Perspective
The choice here is not which parts of a capability programme matter most, but which will still be transacting when no one is watching — and that choice is made in the design room, not the recovery review. Every component approved without a named buyer is a decision, usually unrecorded, to fund a capability for the duration of the programme and no longer.
An executive who approves a programme knowing which components are orphaned has committed to funding them permanently or to losing them, and either is a legitimate position honestly taken. An executive who approves the same programme without asking has made that commitment without noticing, and will find out years later, in the form of a capability everyone assumed was embedded and nobody can locate.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.