--- title: "The First Hour of Production: Is Start-Up Loss a Deviation or a Fixed Cost?" description: "Start-up loss is the time and material consumed each time production is initiated, before the process reaches steady state, and it behaves as a fixed cost per launch rather than a variable cost per unit. Averaged into scrap rate or overhead, it misprices batch size, quotations, and capacity plans simultaneously. The corrective is to make each launch a cost object with its own record." url: https://www.beirek.com/en/blog/production-startup-loss-cost canonical: https://www.beirek.com/en/blog/production-startup-loss-cost published: 2026-01-12 modified: 2026-01-12 category: "Operations & Supply Chain" category_url: https://www.beirek.com/en/blog/category/operations-supply-chain language: en-US reading_time_minutes: 7 publisher: BEIREK LLC publisher_url: https://www.beirek.com license: "© BEIREK LLC — citation with attribution and link permitted" keywords: ["start-up loss","batch sizing economics","commissioning curve","scrap rate reporting","manufacturing cost allocation","quality of earnings discount"] topics: ["Operations cost measurement","Batch size and working capital trade-offs","Commissioning and ramp-up modelling","Quotation and customer mix discipline","Manufacturing due diligence"] alternate_language_url: https://www.beirek.com/tr/blog/production-startup-loss-cost --- # The First Hour of Production: Is Start-Up Loss a Deviation or a Fixed Cost? > **In short:** Start-up loss is the time and material consumed each time production is initiated, before the process reaches steady state, and it behaves as a fixed cost per launch rather than a variable cost per unit. Averaged into scrap rate or overhead, it misprices batch size, quotations, and capacity plans simultaneously. The corrective is to make each launch a cost object with its own record. *Every production start consumes time and material until the process settles into steady state, and that consumption becomes invisible the moment it is averaged into a rate. An unmeasured start-up cost quietly miscalibrates an entire chain of linked decisions — batch sizing, quotation, working capital, and commissioning schedule alike.* --- In a capacity planning session, when two scenarios producing identical annual volume are placed side by side — one built on a small number of large runs, the other on a large number of short runs — the unit cost shown in the model is typically the same in both. The realised unit cost the plant reports at month end, however, tracks not the total quantity produced but the number of times production was started. That this difference goes undiscussed in the meeting is not an oversight; the standard cost card simply carries no line indexed to launch frequency, which means the single most decisive variable separating the two scenarios appears nowhere in the arithmetic being debated. The same pattern is observable at shift level. Scrap measured in the first hour after an extended stoppage runs materially above the shift average for the same line and the same part; the first components off a die change are typically consumed as measurement material; the opening cycles of a newly assigned operator take longer than the same operator's cycles three hours later. Because production reporting presents these differences already averaged into the shift total, the loss does not disappear — it loses its name, settling into a generic scrap percentage where it can no longer be interrogated, attributed, or planned against. The pattern has a name: start-up loss — the time and material systematically consumed at every initiation of production until the process reaches a stable regime. Its mechanism is physical and largely unavoidable. An injection machine produces parts outside dimensional tolerance until it reaches thermal equilibrium; a plating line sends its first surfaces to rework until bath concentration stabilises; a filling line operates below nominal rate until its buffers are charged; an assembly cell remains hostage to the slowest station until line balance is re-established. The loss is the price of the process passing through its own transient regime, not evidence of a process performing badly. There is a domain in which that price is entirely functional, and it should not be skipped over. An operation carrying broad product variety, promising short lead times, and shipping to customers in small and frequent consignments will, by definition, start frequently. Frequent starting is rational to the extent that it lowers inventory carrying cost and shortens response time to shifts in demand; start-up loss is the price of that flexibility rather than an indicator of poor discipline. The difficulty lies not in the loss itself but in how it is recorded: when a cost that arises per launch is booked as though it varied per unit, every decision sensitive to batch size errs in the same direction and by the same magnitude. On the balance sheet, that error usually hides not in the scrap line but in an inventory balance that has risen relative to the prior period. For a plant manager measured on scrap rate, the fastest available lever is to enlarge the batch; as the batch grows, the start-up loss is divided across more units, the ratio improves, and the improvement is reported. The same move lengthens the working capital cycle, depresses inventory turns, and enlarges obsolescence exposure whenever the product mix shifts — yet none of these three consequences surfaces in the scrap report. What has gone missing from reporting is the launch count, not the volume. A second layer opens on the commercial side. When a quotation is priced off average unit cost, the low-quantity order is priced below what it actually costs to produce, while the high-quantity order is priced to carry somebody else's start-up loss; the customer that appears to carry the strongest margin is frequently the one eroding it most. The same gap reopens at the negotiating table, where agreeing without a price adjustment to convert a single monthly shipment into four weekly ones follows directly from a quotation model that carries no per-launch line. Absorbing a counterparty's inventory cost means absorbing not warehouse space but launch frequency. In capital-intensive projects the identical mechanic reappears at a different scale. Demonstrating nameplate capacity for a defined test duration and sustaining that capacity across a month are two distinct claims; the commissioning curve carries a start-up loss expressed through material consumption, rework, and unplanned downtime in the months following first production. A financial model that assumes steady-state operation from the commercial operation date will predictably overstate first-period DSCR, and lenders requiring a reserve account and an initial covenant holiday are pricing precisely the curve the model omitted. Mobilisation and remobilisation belong to the same family: after every demobilisation, site productivity returns to its uninterrupted trajectory only after a measurable interval. The most expensive consequence of this invisibility surfaces at the transaction table. A buyer observing unexplained volatility across a three-year gross margin series will attribute it to product mix or pricing pressure and price it as uncertainty, when a meaningful share of that volatility may originate in the batch plan of those periods — that is, in fluctuation of the launch count. A seller able to demonstrate that distinction establishes that margin is normalisable against the batch plan and therefore manageable; a seller unable to do so watches the gap convert into a quality-of-earnings discount, an earn-out structure, or a widened representations and warranties package. What determines valuation here is not performance itself but the demonstrability of performance being reproducible independently of planning decisions. The mechanism that neutralises this tendency is measurement architecture rather than individual attentiveness, and it separates into four components. The first is converting the launch into a cost object with its own record: elapsed time, measurement material, reworked quantity, and the number of cycles to first acceptable part are captured per launch event rather than folded into the scrap pool. The second is recomputing economic batch size against measured rather than assumed start-up cost, and holding that computation on a single sheet alongside inventory carrying cost. The third is binding pre-launch preparation to a protocol with a named owner — tooling and fixtures staged in advance, material presented kitted at line side, first-article approval scheduled before the shift rather than inside it. The fourth is a quotation model carrying a per-launch component, which shapes mix decisions correctly even where it is never invoiced. In industrial facility, data centre, and process industry projects, BEIREK's intervention in this layer begins on the assumption side of the model: steady-state production and the commissioning curve are constructed as two separate lines, the duration and slope of that curve are calibrated reciprocally against the contractual performance test protocol, and what the test measures is negotiated to be sustainable output across a defined window rather than an instantaneous peak. Through the commissioning period, launch counts, rework volumes, and unplanned downtime causes accumulate in a single record — a record that serves simultaneously as the evidentiary chain in any delay or performance discussion with the contractor and as the basis for revising the first-year operating budget on realistic terms. On the operating side, the cadence established is weekly and turns on a single question: how many times was production started this week, and how long did each start take to reach acceptable output. Once those two figures are recorded consistently, decisions across planning, procurement, and sales come to rest on common ground — the batch size discussion moves from intuition to measurement, the shipment frequency negotiation from goodwill to cost, and the capacity investment case from nameplate rate to observed start-up curve. The value of the record lies less in the number it produces than in the fact that the decision-maker can no longer credibly disregard it. The genuine flexibility of a production system is measured not by how many variants it can produce but by whether it knows what restarting production actually costs; until that cost is measured, flexibility is not a capability but a liability concealed inside an average. ## Key Points - Start-up loss does not behave like a variable cost per unit; it is regenerated in full at every launch event, largely independent of how many units follow it. - When the loss is averaged into a scrap percentage or absorbed into overhead, the variable that disappears from reporting is not production volume but the number of launches. - An unmeasured start-up cost causes small-lot orders to be priced below their true cost, and the resulting drift in customer mix is rarely visible in the margin report. - In capital projects the same mechanic conceals the gap between demonstrated and sustainable capacity, and makes first-period debt service coverage appear stronger than it is. - Buy-side valuation discounts frequently arise not from the level of margin but from the seller's inability to demonstrate that margin is reproducible independently of the batch plan. ## Questions ### What is start-up loss, and how does it differ from ordinary scrap? Start-up loss is the time and material consumed at every initiation of production until the process reaches steady state; machine warm-up, tolerance calibration, first-article approval, and the re-establishment of line balance are its components. It differs from ordinary scrap in behaviour rather than substance: ordinary scrap rises with the quantity produced, whereas start-up loss is largely independent of quantity and is regenerated in full at every launch event. It should therefore be measured per launch, not per unit. ### How is start-up loss measured in practice? Measurement becomes possible once each launch event is treated as a discrete record. Four quantities are captured: elapsed time from initiation to first acceptable output, material consumed during that interval, quantity routed to rework, and the additional time required to reach nominal rate. Accumulated separately rather than blended into the scrap pool, these figures yield a reliable average cost per launch within a few reporting periods, which is sufficient to re-anchor batch and pricing decisions. ### Does increasing batch size solve start-up loss? Enlarging the batch does not eliminate start-up loss; it merely divides it across more units, so the reported scrap ratio improves while the absolute loss remains unchanged. Larger batches simultaneously raise inventory levels, lengthen the working capital cycle, increase obsolescence exposure when product mix shifts, and reduce lead-time flexibility. The defensible decision comes from comparing measured launch cost against inventory carrying cost on a single sheet rather than optimising a ratio in isolation. ### How should commissioning-period productivity loss be reflected in the financial model? Steady-state production and the commissioning period belong in two separate assumption lines, with the duration and slope of the ramp calibrated consistently against the contractual performance test protocol. The critical distinction is whether the test measures an instantaneous peak or sustainable output across a defined window. Absent that distinction, first-period debt service coverage appears stronger than it is; lender requirements for a reserve account and initial covenant flexibility represent the pricing of exactly this exposure. --- Source: https://www.beirek.com/en/blog/production-startup-loss-cost Publisher: BEIREK LLC — https://www.beirek.com