The evidence in Part Two does not lead to a new assessment methodology. It leads to a target. Across all four archetypes the failure was the same: the assessment was trained on the candidate and never on the thing the candidate had to fit. Part Three is about specifying that thing, the demand the situation places on whoever takes the role, with the same discipline already spent on the person, and about what changes once it is drawn.
A selection process measures one thing with care: the candidate. It measures the other thing, the target the candidate is being measured against, hardly at all. The competency framework, the structured interview, the psychometric battery, the reference process are all instruments trained on the person. Nothing in the standard process is trained with equivalent precision on what the person has to fit. The result is marksmanship at a target nobody drew. The aim is exact. What it is aimed at was never specified.
This is the question the whole process leaves unanswered: what does capable leadership actually require in this organizational situation? Not what capable leadership looks like in general, which the competency framework already answers. What this situation, with its specific ownership structure, its specific position in the business cycle, its specific organizational condition, demands of whoever takes the role. That question defines the target. Until it is answered, the assessment is measuring the candidate against a standard imported from somewhere else.
The objection to taking it seriously is intuitive and worth stating plainly, because it is the reason the question goes unanswered. The candidate is concrete. The demand is vague. You can put a person through a validated instrument and get a number; you cannot put a situation through anything and get a number. So the process anchors on the thing it can measure and treats the thing it cannot as background colour. The asymmetry feels like realism. It is the source of the error.
The demand can be specified. Not impressionistically, not as a paragraph of context in a role brief, but as a structured account of what the situation requires that is as disciplined as the assessment of the candidate. The demand profile is an object. It has components. It can be written down, argued over, tested against evidence, and found to be wrong. That last property is the one that matters: a demand profile that cannot be wrong is not a specification, it is a description. A real one makes claims specific enough to fail.
The result is marksmanship at a target nobody drew.
Four components give the demand profile its structure.
The first is the ownership and governance dynamic: who holds authority over the incoming leader, and what that authority does in practice. A PE board running a value creation plan, a family governance structure with informal authority that the org chart does not show, a dispersed listed company shareholder base, a founder still present in an advisory role each impose a different operating reality on the same nominal role. This is specifiable. You can state, before any candidate is seen, what the authority structure will demand of the person inside it, and you can be wrong in a way that later evidence will expose.
The second is the business cycle position: where the business sits in its development, and what that implies for the first eighteen months. A growth inflection, a turnaround, a steady state, a transformation each set a different opening demand. The cycle position at entry is knowable at the point of selection, and it determines what the leader will be doing before anything they were hired to do longer-term.
The third is the organizational health state: the condition of the team being inherited. Where the capability gaps sit, where loyalty is held and to whom, what informal authority will support or resist the new leader. A leader entering a stable, high-performing team faces a different demand than one entering a team depleted by a bad transition, and the difference is not a nuance. It is the larger part of what the first year will consist of.
The fourth, and the one most consistently left out, is the inflection point: the single most consequential thing the leader will face that is particular to this situation. Not the generic challenges of the role category, but the specific decision or pressure the organization is moving toward. The value creation milestone. The first test of independent authority in a succession. The shareholder challenge already visible on the horizon. The moment the founder's informal presence ends and the new authority has to stand on its own. Each defines what the leader has to be able to do before anything else, and each is specifiable in advance with more precision than the role brief ever attempts.
These four are not a checklist to be completed and filed. They are the anatomy of a target. Specified together, they describe a demand with enough definition that a candidate can be matched to it, missed against it, or found wrong for it on evidence rather than impression. That is what it means for the demand to be specifiable to the same standard as the candidate: not that it yields a number, but that it makes claims precise enough to be tested and precise enough to be mistaken.
Against this, the default process is exposed for what it produces. The role brief captures what the previous leader did, not what the situation now requires. The competency framework reflects capable leadership across organizations, not what this organization needs from this leader at this moment. The longlist gathers people who have held the role category, not people whose profile matches the demand. Each step is defensible on its own terms. The sequence assesses the candidate with precision against a target left undrawn.
When the demand profile is specified to this standard, the assessment is not replaced. It is aimed. The structured interview, the instruments, the references all keep their value, because the criteria they test against are now derived from the situation rather than imported from a generic standard. The same questions sharpen. Tell me about a time you drove significant organizational change becomes tell me about a time you established authority in an organization built around a single dominant figure. The capability under examination is the same. What counts as sufficient evidence of it is now set by the demand rather than by the descriptor. The instruments do not change. Their target does.
Evaluating against a defined target also changes who is qualified to do the evaluating. Recognizing whether a candidate has met the demand requires people in the room who understand the demand well enough to see it in an answer, who can tell the candidate who has navigated this specific challenge from the one who is merely fluent about it. That is not always the composition of a board selection committee, and it is the part of the change that is easiest to agree with and hardest to act on.
Two things stand in the way, and neither is technical. The first is that specifying the demand requires the board to say things about its own organization that are uncomfortable in writing: the governance tensions, the ownership pressures, the condition of the team, the risks that were invisible under the outgoing leader. A demand profile honest enough to be useful is honest enough to be awkward. The second is accountability. A demand that has been specified can be checked. If the appointment fails, the question becomes whether the demand was drawn correctly, which implicates the people who drew it. Individual shortcoming is a more comfortable verdict than a target specified wrong. Generic criteria protect the people who set them. A defined target exposes the judgment behind it, which is precisely why it is worth defining and precisely why it so rarely is.
None of this is intellectually demanding. It does not require a new instrument or a proprietary method. It requires drawing the target before taking aim, and accepting that a target drawn in the open can be judged. The objection that the demand is too vague to specify is not a finding about the demand. It is a description of the work not yet done.
That work raises a real objection of its own, that calibrating to context trades the dominant model's objectivity for judgment. Whether that is a loss is the question Chapter 8 takes up.
Valtteri Länsimies · FNDRY. Advisory