There is a scene that recurs in weekly site coordination meetings with enough regularity to constitute a pattern rather than an anecdote. A non-conforming fabricated item comes to the table; the same manager is expected both to protect the programme and to enforce technical acceptance without concession; and the manager declines neither expectation outright. Nothing is minuted as a rejection, nothing as an approval. The item is carried to the following meeting, an information note is circulated to the relevant party, and the record is composed in a manner that preserves the author's position. Viewed from outside, this reads as indecision. Read against the instructions actually in front of that manager, it is something else entirely: both directives are defensible on their own terms, both originate from the same hierarchy, and no single action exists that satisfies them simultaneously. Under those conditions, the typical observed behaviour is not the making of a decision but the postponement of the moment at which one becomes unavoidable.
The pattern is not confined to the site. While the organisation chart shows a single unbroken reporting line, instructions in practice arrive from three sources at once — cost and schedule from the sponsor side, technical standard from the functional line, and contract administration together with client relationship management from the commercial side. A procurement manager is asked to drive unit prices down while guaranteeing continuity of supply; a finance director is asked to enforce group consolidation discipline while advocating a subsidiary's growth case. Because each expectation is separately defensible, none is ever formally withdrawn, and the contradiction, instead of being surfaced and settled at the level capable of trading one against the other, is left to be resolved at a single desk that frequently lacks the authority such a trade requires.
This pattern has a name — role conflict — and its mechanics operate in three distinct configurations. The first, and the most common in matrix structures, involves two separate instruction sources producing mutually exclusive expectations. The second involves a single source reversing its own directive over time, as when an acceptance criterion enforced strictly in early phases is quietly relaxed as closing pressure builds. The third is the tension between what the role requires and what the individual's professional standard permits, exemplified by the engineer asked to sign technical acceptance and run commercial closeout in the same week. All three are routinely conflated with role ambiguity, though the two are structurally opposite: in ambiguity the expectation is unknown, whereas in conflict the expectations are exceptionally clear, and it is precisely that clarity which makes them irreconcilable.
Generating contradictory expectations is not, in itself, a defect. In complex, capital-intensive projects the matrix is constructed deliberately so that the trade-off between cost and technical quality, between speed and risk, between commercial flexibility and contractual discipline, surfaces somewhere rather than remaining latent; the tension is the surface on which that trade-off becomes visible. The conditions under which the tension stays functional are narrow, and there are two of them: either the person carrying the contradiction holds the corresponding decision right, or, lacking it, has access to an escalation path that is short and explicitly defined. Where either component is absent, the structure loses its function, since the expectation has been delegated while the authority has not, and the resulting gap falls to individual judgement to close.
The behaviours that emerge when that gap is left to individual judgement are entirely predictable, and they are not irrational; each of them lowers short-term cost with considerable efficiency. Decisions are deferred, because any decision in either direction places the decider in opposition to one of the parties. Correspondence is drafted for defence rather than for resolution, since the written record is the only protection available when responsibility is later reconstructed. Visibility becomes selective, with reporting flowing toward the loudest party, the nearest deadline, or the party that in practice conducts the performance review, while the other side receives information only on request. And the individual who carries the contradiction longest tends, in the end, to leave the organisation rather than to renegotiate the role, taking with them not merely a headcount but the institutional memory of how that particular contradiction was managed under which circumstances.
The contractual expression of these behaviours appears most clearly in notice and notification windows. Claims for delay, non-conformity, employer-caused prevention, or force majeure are almost uniformly time-barred, typically at the scale of days rather than weeks; yet where the same individual is charged both with preserving the counterparty relationship and with reserving contractual entitlement, the notice is routinely held back for several days in the interest of not disturbing the former. Each deferral appears costless in isolation. Its cumulative effect is that a substantively valid claim becomes procedurally contestable. The same mechanism governs variation orders, where a scope divergence is recorded not at the moment it is identified but at the moment the relationship can accommodate it, with the intervening work proceeding as effectively unfunded.
The second cost item is operational and never appears in the budget under its own name. Where technical acceptance and commercial closeout converge in one role, the acceptance threshold drifts downward as schedule pressure rises, producing a rework burden absorbed under warranty in the first year and migrating into operating expenditure thereafter. The cost also accumulates in the distribution of approval durations rather than in their average — an average approval time may look entirely acceptable while specific categories of item run several times longer, and that divergence is the most reliable available indicator that an authority gap exists in precisely those categories. Attrition among qualified personnel is the latest-recognised and most expensive form of the same burden, since the individual who departs is generally the one who absorbed the conflict longest and on whom the organisation therefore leaned hardest.
The third cost, and the one that appears latest on most owners' agendas, materialises at the diligence table. The question asked in a due diligence process is not what the roles are but who, in practice, resolves conflicts between them, and where the answer converges consistently on a single name, the finding ceases to be an organisational observation and becomes founder dependency. That finding rarely translates into a headline discount. It is far more commonly priced through the mechanics of the transaction: the addition of a written delegation-of-authority matrix to the conditions precedent, an extended earn-out period, hardened key-person undertakings, and a broadened representation and warranty package on the contract administration heading. What determines a company's valuation is frequently not the performance itself but the demonstrability of that performance as repeatable independently of the founder, and where role conflict gets resolved is among the most direct evidence available on that question.
The mechanism that neutralises this tendency is not individual awareness but a four-component design. The first is the mapping of decision rights at the level of decisions rather than tasks; what is written down is not who performs which activity but who carries the final signature for which category of decision. The second is fixing the priority order by project phase, because where it has not been settled in advance which of technical acceptance or schedule governs in a given phase, the ordering is reconstructed each time in favour of whatever pressure is dominant at that moment. The third is binding the escalation path to elapsed time, so that a contradictory instruction unresolved within a defined interval moves upward automatically, which renders deferral structurally unavailable as a strategy. The fourth is maintaining conflicting instructions in a discrete register, converting the conflict from an interpersonal friction into a traceable governance item.
BEIREK's intervention on this problem begins, when the project governance framework is established, with fixing those four components in a single document. The instruction channel is made singular: multiple parties may offer direction, but binding instruction travels one line, and any direction issued outside that line is logged and not executed. The decision rights matrix is built around decision categories rather than the work breakdown structure — technical acceptance, scope variation, payment certification, contractual notification, and supplier substitution occupy separate rows, precisely because conflict originates in the blending of those rows. The role that runs commercial closeout is separated from the role that signs technical acceptance even where the team is small; where separation is genuinely not feasible, the combination is recorded as a risk item and offset by a compensating review step.
The second layer of intervention is rhythm. The conflicting-instruction register is opened at fixed intervals as the first agenda item of the existing project review, without creating an additional meeting, and each entry answers a single question: at which level, and within what elapsed time, did this contradiction close. Closure times exceeding a defined threshold are read as a signal that the authority definition for that decision category is incomplete, and the matrix is amended that week. Contractual notification windows are tracked on a calendar kept deliberately separate from relationship management, so that the decision whether to serve notice becomes independent of the prevailing temperature of negotiations. A by-product of this rhythm is exactly the evidence an acquirer looks for in a closing process: a retrospectively traceable record showing that decisions were bound to a defined structure rather than to particular individuals.
Contradictory role expectations measure an organisation's maturity not by whether conflict exists but by the level at which it becomes visible. A structure in which tension is absorbed quietly at the lowest desk looks untroubled from outside and, for exactly that reason, pays its costs on a delay; a structure in which tension travels upward looks noisier, but prices the trade-off at the level capable of pricing it. The question worth asking is not who is responsible for which task, but what authority the desk holds, and within what elapsed time it is required to release the matter, when two mutually exclusive and equally legitimate expectations land on it at the same moment.