In product development meetings, cost tends to enter the conversation at almost exactly the same moment every time: after the design has been substantially locked, after the supplier has effectively been selected, and after the first samples have been reviewed. Everything discussed up to that point concerns function, delivery date, and customer expectation, while cost emerges only as the computed consequence of the configuration that has taken shape. Nobody in the room says anything incorrect, each function has done competent work within its own remit, and yet the figure placed on the table is not a target but the arithmetic result of decisions already taken. The significance of that distinction reveals itself over the following quarter, when a number that arrives higher than expected triggers a retrospective search for savings — a search that gravitates, predictably, toward material thickness, supplier substitution, or warranty scope, which are the items easiest to reverse and most expensive in their downstream consequences.
A second pattern is visible in the same room. So long as the aggregate margin of a product family remains acceptable, the cost performance of individual products goes largely unexamined, because a satisfactory consolidated figure offers no operational reason to open the underlying lines. That choice is rational to the extent that it lowers the near-term cost of management attention, which is finite and better spent where the aggregate looks distressed. The difficulty arises when the choice persists after the product mix has shifted: an average carried by a handful of profitable lines renders the loss-making lines beneath it invisible, and that invisibility converts, on the day the mix moves, into an abrupt margin compression that appears to have no proximate cause.
The mechanism underlying both patterns is the positioning of product cost inside the company as a **reporting item** rather than a **decision variable**. Treated as a decision variable, cost operates as a constraint that narrows engineering freedom in advance; treated as a reporting item, it is an outcome that narrates the past. What separates the two positions is nothing more than the moment at which the number is set. A ceiling established before the design decision compels the engineering team into genuine search — alternative materials, alternative production methods, alternative modularity are actually evaluated — whereas a figure computed after the design is frozen can only be accepted or negotiated downward, and negotiation of that kind is almost invariably externalized as pressure on supplier margin, eroding the quality of the supply relationship over successive cycles.
The second layer of the mechanism concerns ownership. A product cost target sits, by its nature, at the intersection of three functions: product management sets it, engineering renders it technically feasible, and procurement realizes it on the supply side. Where ownership at that intersection is not explicitly assigned, the target becomes nominally everyone's responsibility and practically no one's, since each function, when a deviation surfaces, holds a legitimate argument locating the cause outside its own boundary. Such a gap does not close on its own; it fills upward, and the decision arrives at the founder or the general manager. Companies frequently experience this not as a defect but as evidence of fast decision-making — a reading that remains accurate right up to the threshold at which institutional maturity is required.
Understanding how this structure appears at the review table requires attention to the question actually being asked. A diligence team does not ask what the product costs; that figure is already derivable from the financial statements and the cost accounting. What it asks is whether the figure is **predictable**: within what band a target set for a new product is realized once serial production begins, at what frequency and magnitude deviations occur, and what mechanism engages at the moment of deviation. Absent an answer, a buyer or investor constructs the margin projection not from the company's plan but from the average of historical outcomes, and an average — containing by definition both the good years and the poor ones — sits invariably below the plan.
The fate of an undocumented target is sharper still. Where a product cost target genuinely exists inside the company but resides solely as a spreadsheet in the product manager's working folder, the review will not treat it as a verifiable structure. The verifiability standard is strict and its logic is unremarkable: a document that is unapproved, undated, and unversioned may have been produced after the fact, and the reviewing party makes no allegation to that effect — it simply declines to rely on it. The practical outcome is close to identical with having had no target at all, the only difference being that the management team felt better prepared, a feeling that dissipates rapidly in the weeks before closing.
Measurement is the dimension most frequently skipped in this area, precisely because companies believe they are already measuring cost. What is typically measured, however, is **realized cost**, not variance against a target. The difference between the two is managerial rather than technical: realized cost is an accounting output describing what happened, while variance against target is a management signal indicating which product family, which cost line, and which direction of drift require attention. Where variance is not tracked by product on a settled cadence and under a stable definition, cost escalation tends to be noticed only after the price list has been printed, the quotation issued, and the margin already committed to a customer.
Continuity is the dimension that connects most directly to valuation. Where a cost target regime functions independently of the founder or of a single senior engineer, the acquirer is buying a capability, and a capability is transferable. Where the same result is generated through one individual's sector knowledge, supplier relationships, and negotiating skill, what exists is performance without capability; and once it becomes clear that the performance will not transfer, the transaction structure adjusts accordingly. That adjustment generally arrives not as a reduction in the headline price but as a deterioration in the shape of the consideration: the earn-out portion expands, the escrow ratio rises, the retention commitment of the key individual lengthens, and additional headings covering cost realization enter the representations and warranties.
Building this structure is a matter of institutional architecture rather than individual discipline, and it separates into four components. The first is a **timing lock**: the cost target is set before the design-freeze point in the development process, as an irreversible stage gate through which the program does not pass until the target exists. The second is **singular ownership**: the target is assigned to one role — typically product management — and that role carries not only the authority to defend the target but also the authority to relax it, since responsibility without authority has never produced ownership. The third is a **deviation record**: every occasion on which the target is changed is logged together with the rationale, the approving role, and the margin impact. The fourth is **cadence**: the variance report is produced on a fixed period, monthly or by product family, under the same definition and in the same format each time.
BEIREK's intervention in this area begins with identifying the layer at which the existing regime breaks, because in most companies it is not all four components that are missing but one or two, and the missing component differs from company to company. The structure established after that diagnosis consists of a decision record binding the moment of target-setting to a development stage gate, a written authority definition fixing both the owner of the target and the power to relax it, and a fixed-cadence reporting line tracking variance at product-family level. These three operating together produce a structurally different outcome from any one of them operating alone: the target ceases to be the conclusion of the argument and becomes the constraint under which the argument is conducted.
The second line of intervention concerns readiness for review, and here timing is determinative. A cost target regime becomes verifiable not through retrospective document production but through the genuine operation of at least several reporting cycles, which means the structure must be established well before a transaction process begins rather than at its outset. In the preparatory work we run, the target-versus-actual series by product family, the recorded rationales for deviation, and the corrective decisions taken are consolidated into a single record line, so that the reviewing party constructs its margin expectation from data the company itself generated rather than from the narrative the company presents. What this provides is not a story to be defended but a base that can be tested.
The real function of a product cost target inside a company is not to reduce cost; it is to determine who decides how far cost may rise, on what rationale, and against what record. In a company where that decision line has been built, the target is sometimes missed, but every miss is known, explained, and priced; in a company where it has not been built, the target appears to be met most of the time, because what the target was got settled after the deviation occurred. Distinguishing between these two companies at the review table takes a few hours, and the cost of the distinction propagates through the entire structure of the transaction.