In a manufacturing review meeting, the statement that every part shipped during the quarter fell within specification limits is generally received as a performance result, while a second line appearing in the same deck is read with comparable calm: hundred-percent final inspection was applied before shipment, and a portion of the sorted parts was reworked and returned to the line. Placed side by side, the first item registers as an indicator and the second as an operational detail, though the second describes the mechanism by which the first was produced. To the extent that final inspection has been part of the routine for years, the question of how a lot would look with sorting removed does not arise, since an activity that has become routine falls outside the field of observation. The data carrying genuine information about the process is generated at the inspection station and consumed there, while only the binary outcome — conforming or non-conforming — travels to the decision table.
The same pattern presents a different face at the negotiating table where capacity is committed for a new program. Units per shift are calculated from cycle time, planned downtime, and a maintenance allowance, while sorting labor, the machine time a part consumes a second time when reworked, and the setup stops taken during the day to recenter the equipment typically fall outside that arithmetic. The small corrections an experienced operator makes when a material lot changes or as tooling wears are not recorded anywhere, and consequently reach neither the cycle time nor the cost standard. So long as that knowledge resides in a single person, the line runs; when the person takes leave or a second shift comes online, the cost of obtaining the same result from the same equipment rises appreciably.
The name for this configuration is process-capability shortfall — the condition in which the natural variation of a process does not reliably fit inside the tolerance band demanded of the part — and it is a statement of an entirely different kind than conformance to specification. Conformance is a retrospective judgment about a part already produced; capability is a prospective estimate about the process, a prediction of the band into which the next thousand parts will fall. A line can ship one hundred percent conforming product while sorting a material share of its output, a structure that holds the conformance indicator at perfection and capability well below it. The distinction also separates two different failures in practice: variation may be wider than the tolerance band, or variation may be sufficiently narrow while the center of the distribution has drifted off the middle of the band. The first demands investment in equipment, fixturing, and method; the second can often be closed through setup discipline, and collapsing both into a single scrap rate produces misdirected capital decisions.
Working through sorting is not an error but a rational shortcut that lowers cost under specific conditions. At low volume, while a product is still immature, with high unit margin and a limited customer base, the fixed cost of making the process capable — new equipment, refreshed fixturing, measurement infrastructure, engineering time — exceeds the variable cost of inspection labor, and choosing inspection is the correct decision. The difficulty lies not in the shortcut itself but in the shortcut persisting after the conditions that justified it have changed: when volume rises by an order of magnitude, when a second shift opens, when material begins arriving from a second supplier, or when the part stacks against other tolerances inside an assembly, the variable cost of inspection overtakes the fixed cost of investment. That crossover does not announce itself as a crisis; it becomes visible over months as a gradual climb in scrap, slippage in delivery dates, and a thickening stream of customer complaints.
How capability is measured carries a layer of its own. Capability indices calculated from a single material lot, on a single shift, with the most experienced operator on the line, and from a consecutively drawn sample, describe short-term variation, whereas the performance committed to a customer is subject to long-term variation across changing material lots, operators, ambient conditions, and tool life, and the distance between the two figures is typically not small. Added to this is the uncertainty of the measurement system itself: the repeatability and reproducibility allowance of the gauges and comparators in use consumes part of the tolerance band before measurement even begins, and where that allowance is unaccounted for, the process reads as more capable than it is. On tight-tolerance work, reviewing the share of the tolerance band absorbed by the measurement system is the step that precedes equipment investment and costs a fraction of it.
The first surface on which the institutional cost appears is the capacity plan. Because the difference between a line's nominal capacity and the capacity at which it produces capable output does not occupy a separate line in the investment budget, growth scenarios are generally built on the nominal figure; once new orders are accepted, the difference is closed through overtime, weekend shifts, and outside subcontracting. Since those three items disperse into manufacturing overhead in the accounts, the cost of the capability shortfall is never aggregated under a single heading and therefore cannot support any investment case. Where sorting labor is reported as indirect rather than direct labor, and rework time as efficiency variance, the magnitude of the problem remains invisible as a structural property of management reporting. A cost item that cannot be gathered into one line during the budget cycle predictably settles toward the bottom of the priority list.
The second surface is working capital. As process variation widens, the only practical route to preserving delivery reliability is to enlarge safety stock, which is why a capability shortfall usually reaches the balance sheet not as a quality indicator but as the level of finished and work-in-process inventory. When the slowdown in inventory turns is explained by supply chain conditions and the lengthening cash conversion cycle by collection terms, the link between first cause and final symptom is severed. Similarly, where stacked tolerances on an assembly line force selective assembly, the same part must be held in separate size groups, and the count of stock-keeping items grows in proportion to process variation rather than product variety. From outside, such an inventory structure reads as weak planning; from inside, it is the purchase of delivery performance with cash on behalf of a process that is not capable.
The third surface lies on the contractual and valuation side. Where PPM commitments, warranty scope, and delay penalties in customer agreements are signed against the conformance record of past shipments rather than against measured capability, a silent gap opens between commitment and capacity, and that gap realizes as cost at the first serious volume increase. In a sale or minority investment process, normalizing scrap, rework, and sorting in the quality-of-earnings review changes the margin picture; the buyer's typical response is not to reduce the multiple directly but to tie a post-closing capability program to an earn-out condition, or to raise the escrow percentage against warranty claims. Combined with a critical customer supplied from a single approved line, the same finding hardens how customer concentration risk is priced, since bringing a second line through customer approval requires a calendar measured in quarters rather than months.
A capability shortfall is a problem managed through decision architecture rather than individual vigilance or quality awareness, and it separates into four components. The first is tolerance provenance: a record of the functional requirement from which each critical dimension on the drawing derives, given that a meaningful share of tolerances observed in the field descend not from function but from a drawing written years earlier, and an unnecessarily tight tolerance is the cheapest closable cause of an apparently incapable process. The second is a critical-characteristic register: a limited set of dimensions defined at the characteristic level rather than the part level, each with a named owner and a fixed measurement method. The third is a capability gate ahead of commitment: before a volume commitment, price, or penalty clause is signed, a study representing long-term variation on the relevant characteristics is complete. The fourth is a change trigger: the capability record reopens automatically whenever fixturing, material supplier, machining program, or shift structure changes.
The mechanism BEIREK establishes in engagements of this kind is not a second audit layer running parallel to the quality function, but a recording and cadence discipline that carries capability information from the inspection station to the investment and contract tables. In practice this covers reducing critical characteristics to a bounded register, determining for each characteristic the share of the tolerance band absorbed by the measurement system, measuring long-term variation in a way that spans material lot and shift changes, and separating sorting and rework cost out of manufacturing overhead so that it accumulates in a single line. The cadence ties volume commitments and commissioning schedules to that record: each step of the ramp plan is read against measured capability on the relevant characteristics, and capacity increase is compared with the investment decision inside the same document. What emerges is not a quality report but a single basis on which both the investment committee and the party signing the supply agreement look at the same number.
The value of a manufacturing asset is measured not by whether the parts it produces conform, but by the cost at which and the predictability with which it can reproduce a conforming part, and sorting, rework, and safety stock are the missing portion of that repeatability purchased with cash. Once the sum of those three items becomes visible in one place, the discussion moves out of the heading of quality awareness and into the heading of capital allocation, which is where the question properly belongs. The answer to how capable a line actually is resides not in the records of the inspection station, but in which decision table those records reach and in what form.