In a board meeting, the person expected to speak to a technical item on the agenda and the person who actually speaks to it are frequently not the same. The operations director prepared the deck and their name sits at the foot of the slide, yet the answer to the question comes from the founder, with the director adding one or two supplementary sentences before the item closes on the founder's formulation. The same pattern repeats in supplier negotiations, in pricing exceptions, in hiring approvals and at the escalation step of customer complaints. Inside that room the organisation chart is fully populated; no box is empty, each carries a name and a title. Asked where the decision authority those boxes are meant to carry actually resides, however, the answer is given not by reference to the distribution of boxes but by reference to a single person's calendar.
The party sitting at the diligence table never sees that room, but it sees the traces the room leaves behind. When the organisation chart shared in the data room is placed alongside the minutes, approval emails and signature authorities of the same period, the distance between the box and the signature emerges on its own. What the investor is looking for is not headcount but which decisions can be concluded without passing through a particular individual; the question asked is not whether the role is filled but which decisions would stall, and which would continue, were the person holding it unreachable for six weeks. That is generally the exact question the company has never put to itself, because from the inside the impression of an uninterrupted line conceals the fact that the line runs through a single node.
The mechanism beneath this configuration is not a management failure but a shortcut that genuinely lowers cost at a particular stage of growth. While the company is small, routing decisions through the founder is the fastest available path: the context already sits in the founder's head, the rationale is never committed to writing, approval arrives within seconds and the coordination cost is close to zero. Delegation, by contrast, is expensive at the outset, since writing the criteria, defining the threshold values and absorbing the first mistaken decisions all consume time and money. The persistence of the shortcut is therefore rational up to a certain scale. The difficulty lies not in the shortcut itself but in its remaining fixed when the conditions change — when headcount widens, geographies multiply, or decision volume rises by an order of magnitude.
What emerges when the shortcut stays fixed is the gradual emptying of titles of real authority. The role description has been written but the decision rights are not enumerated; a person has been appointed but the monetary approval limit is undefined; a team has been assembled but its output is re-evaluated at the same higher step every time. Under this configuration the person carrying the role predictably chooses one of two behaviours: either escalation becomes habitual, which loads the node further, or a parallel practice is quietly established within their own area, which moves institutional memory out of documents and into individuals. Neither behaviour alters the reported fill rate, and both empty the role in practice.
At this point the review dimensions separate from one another, and in most companies they break in the same sequence. The existence of the role — someone working under that title — is generally satisfied. Documentation of the role, meaning a current job description, defined monetary and contractual authority limits and an approved signature matrix, has typically either never been produced or belongs to an organisation that existed two or three years ago. Actual application, meaning the exercise in daily flow of the authority the document records, is a fact independent of the document's existence and usually lags behind it. The measurement layer — decision volume attributable to the role, cycle time, reversal rate — is rarely maintained. Ownership and continuity arrive last and are seen least often; where a second signature, a designated backup and a handover protocol exist for a role, the company has cleared both layers, and the number of companies that have is limited.
The institutional cost of this gap never appears as a discrete expense line; it distributes itself across other lines and accumulates there. An approval path running through a single node lengthens proposal cycle time on the sales side, and the lengthened cycle registers in the win rate; on the procurement side, approvals granted after the negotiating window has closed work into margin as forgone price advantage. In operations, the rework generated when decisions taken by an ambiguously empowered role are subsequently reversed at a higher step is not measured directly and therefore disappears inside general administrative expense. Staff turnover feeds from the same channel: the tenure of a senior employee holding an unempowered title falls appreciably below sector norms, and each departure carries away the undocumented portion of institutional memory.
At the valuation table this finding alters the transaction structure before it touches the multiple. Where critical decisions are found to depend on one individual, the buyer's typical reflex is not to cut price directly but to spread the risk across time: the earn-out period lengthens, earn-out thresholds are tied to the founder's continued involvement, key-person undertakings and non-compete periods widen, the escrow ratio rises, and the filling of specified roles or the approval of an authority matrix enters the conditions precedent. On the representations and warranties side, the scope of statements concerning management structure narrows, because the counterparty is pricing the acquisition of a person rather than a system. Under these conditions the present value of the cash reaching the founder falls materially below what the headline price implies.
The same finding has a debt-side counterpart, and it is frequently overlooked. Credit committees read operating risk through the continuity of management, and where no documented answer exists to the question of how cash generation would be preserved on the loss of a key individual, that risk enters the agreement either as a key-person insurance requirement, or as a covenant treating a change of management as an event of default, or as a clause restricting distributions. The result, if not an outright increase in the cost of capital, is a narrowing of flexibility in its use — which, for a company in a growth phase, amounts in practice to the same thing.
Correcting this area begins not with adding boxes to the organisation chart but with making visible the decision rights the existing boxes carry. BEIREK's intervention in structures of this kind proceeds through three separate records. The first is a decision rights matrix: monetary limits, contractual binding thresholds and exception authority are written role by role, fixing on a single page who may decide alone, who requires a second signature and who may act only by board resolution. The second is the decision log, and the critical detail is this: the record is kept at the moment of proposal rather than at the moment of approval — once it becomes traceable who brought a proposal forward, on what grounds, and at which step the decision changed, which roles are functioning and which are merely relaying becomes visible on its own within a few months. The third is the succession order: for each critical role, a first and second backup, a handover protocol and the thresholds triggering handover are defined in advance.
Once these three records exist, a rhythm must operate them, and without that rhythm the records age quickly. In the method applied, the authority matrix is tied to a fixed periodic review, and the review examines not the matrix as written but the distribution of decisions actually taken during the period; an authority assigned to a role in the matrix yet never exercised across the period indicates either that the role is effectively vacant or that the threshold has been miscalibrated, and each finding produces a correction. Measurement of critical role coverage is bound to the same rhythm: coverage is measured not by the assignment of a title but by the proportion of decisions defined for a role that are concluded by that role. The only verifiable evidence that founder dependence is receding is the direction of that ratio over time — a series, not an assertion.
One side effect of establishing this order arrives from a direction most companies do not anticipate: once authority boundaries are written down, certain critical roles turn out never to have existed. Different fragments of a function are distributed across three individuals, none of whom is responsible for the whole, and the owner of the whole is in practice the founding office. Where that finding surfaces from the counterparty during diligence, it weakens the negotiating position; where it surfaces within the company's own rhythm, sufficiently ahead of a transaction, the definition, cost and lead time of the role to be filled become a line in a plan. The difference between the two is that the same fact is priced once as risk and once as preparation.
What an investor is ultimately looking for in critical roles is not who produces the company's current performance but who the company can produce it without. A coverage table does not answer that question; what answers it is whether the consequence of interrupting the decision line has been tested at least once before and recorded. The question a company should therefore put to its own management structure concerns not the number of titles it carries but whether the owner of a decision and the record of that decision reside in the same place.