When a diligence session calls for the list of the company's active projects, what arrives at the table is rarely a single list. An operational tracker maintained by the delivery team, a separate extraction assembled by finance from budget lines, and a third record derived by the commercial function from customer commitments, placed side by side, will typically disagree on the count itself. The sharper question follows immediately: who owns each of these projects. The answer is characteristically not a name but a cluster of names — technical lead, commercial counterpart, budget approver and managing director cited within the same row. The cluster is not in itself a defect of organisational design, since genuinely shared responsibility does exist. What tends to be met with silence is the question that comes next: when scope and schedule collide, which of those names decides what is surrendered, and where is that decision written down.
A second pattern, less readily noticed, is that the author of the progress report and the owner of the delay are the same person. Where the delivery team reports its own advancement against a percentage of completion it has itself defined, a deviation enters the record only once the party producing it declares it — a lag generated by the mechanism rather than by intent, with the consequence that variance surfaces in the system well after the moment it occurred. The trace of this is easily read at the diligence table: the difference between project closeout cost and approved budget appears in none of the preceding progress reports. The structure is calibrated to defer bad news, and that calibration says nothing about the quality of the individuals operating within it.
What requires naming at this point is the conflation of project management ownership with execution responsibility. Execution responsibility concerns the performance of the work; ownership concerns the authority to decide among scope, budget, schedule and quality, and the obligation to answer for the consequences of that decision. In most companies the first is defined almost without exception — someone does the work — while the second remains suspended in the gap between formal authority and earned legitimacy: the person whose title confers the mandate does not make the call, and the person effectively making it lacks the formal ground on which to defend it. Viewed externally, the resulting arrangement resembles a matrix organisation; viewed internally, it functions as a routing table in which every consequential trade-off is directed to one superior node, usually the founder or the managing director.
The question cannot be framed correctly without first acknowledging that this configuration is rational at a certain scale. Where few projects run concurrently, where the resource pool remains tractable within a single person's memory, and where the customer relationship operates directly through the founder, formalising ownership raises coordination cost rather than reducing it, and centralised decision-making produces fast and accurate calls. The difficulty lies not in the shortcut but in its persistence after the underlying condition has changed. Once the number of concurrent projects crosses the threshold at which the same team must be allocated to two engagements within the same week, the central node ceases to accelerate and begins to form a queue, and the cause of slippage is no longer technical but the wait for an allocation decision.
The party conducting an investment review approaches this arrangement first through the question of existence, and locates the answer in availability rather than in verbal assertion: whether project ownership is defined, and whether that definition sits somewhere a middle manager inside the company can actually consult. Documentation follows. What is sought here is not a procedure manual but numerical authority thresholds for expenditure, scope change and schedule slippage, together with the decision records generated at the moments those thresholds were crossed. An undocumented practice may well be functioning admirably, yet a practice that cannot be verified is not credited in transaction pricing; so long as institutional memory resides in individual recollection, nothing guarantees that the memory remains with the company after closing.
The implementation dimension measures the gap between paper and behaviour, and it is typically the fastest finding to emerge: where an approval threshold has been defined yet the majority of the last twelve months' scope changes were passed beneath it in fragments, the mechanism exists without operating. The measurement dimension is the costliest in valuation terms. Absent regular tracking of schedule variance, cost variance, change-order volume and rework rate, successful past deliveries cannot be separated into those attributable to management quality and those attributable to favourable conditions, and a performance series that cannot be decomposed cannot support a projection. What the investor does at this juncture is not to assume the company is poorly run, but to price the uncertainty.
The channels through which that pricing occurs are well established and substitute for one another across the negotiating table. Where forecast reliability cannot be demonstrated, the first response is a direct reduction in the multiple; where the sponsor resists, the second is to make a portion of consideration contingent on delivery milestones — an earn-out constructed here not as an incentive instrument but as a bridge across the measurement gap. The third channel is the broadening of representations and warranties concerning cost-to-complete on projects in progress, matched by an increase in the escrow proportion. The fourth, applied wherever ownership is found to concentrate in a single individual, comprises key-person undertakings, non-compete periods and retention packages, the cost of which is borne almost invariably by the seller.
Continuity is the question standing behind all four channels. In modelling the first twelve to eighteen months after closing, a buyer or investor asks whether the existing project portfolio can be completed at the same margin without the founder's daily intervention; where the answer is uncertain, a transitional services arrangement, incremental management capacity and typically some form of external project management support are budgeted, and that budget is deducted from enterprise value directly. In a company where ownership has not been institutionalised, what is being sold is not the record of projects delivered but the likelihood that the person who delivered them will remain — and a likelihood is not priced the way an asset is priced.
What neutralises this tendency is not individual discipline or awareness but the architecture itself, and the intervention resolves into four separable components. The first is the assignment of decision ownership for each project to a single name, which need not be the same person as the executor. The second is the establishment of numerical authority thresholds for expenditure, scope and schedule, defined together with an aggregation rule that prevents threshold breaches from being passed through in fragments. The third is that the decision record is kept at the moment of proposal rather than the moment of approval, since where the rejected options and the reasoning behind their rejection go unwritten, no subsequent reviewer can assess the quality of the decision. The fourth is a fixed-cadence, threshold-triggered review regime that separates the reporting of variance from the party producing it.
BEIREK's intervention in this area typically proceeds not through the delivery of a procedure document but through building the mechanism and operating it directly for a period. For each project in the portfolio we separate the decision classes, define an owner for each class by single name and authority threshold, and redistribute according to threshold level the decisions currently routed to the founder's desk; thereafter we establish the monthly variance cadence — the arrangement under which schedule and cost differences between plan and actual, change-order volume and rework rate are reported in one format, on one day, through a record independent of the executor — and run the first several cycles ourselves. The closeout record, capturing what departed from plan at project completion and the decision point to which the departure can be traced back, is the only verifiable carrier of institutional memory, and its continuity is monitored beyond handover.
What this regime produces at the diligence table is not a stack of documents but a single demonstrable series: the narrowing of the gap between planned and actual across the last eight to twelve quarters, and the independence of that narrowing from whether a particular individual was assigned to the project. For the sponsor, the series does not so much reduce the discount as reduce the uncertainty allowance embedded in the counterparty's model; for the senior lender it loosens covenant calibration around completion risk; for the buyer it narrows the scope of the earn-out, since a structure erected as a bridge over what cannot be measured becomes redundant once measurement exists. The same series serves an internal function that is frequently more valuable than the transactional one: the resource allocation debate ceases to be a contest of persuasion and becomes a prioritisation grounded in data.
A company's project management maturity is measured not by observing how its best project runs but by observing when, and by whom, its worst project is declared. So long as no one owns the declaration mechanism, each delivered project stands in the investor's assessment as evidence of a singular outcome rather than of a repeatable capability. The operative question is therefore not what the company has delivered to date, but whether it is already known today which desk the first scope-change request will reach when the next project begins with the founder absent from the room.