Ramp Cost

Costing a Production Ramp: Pilot Spend, Scrap Exposure, and Launch-Delay Money

The money side of a ramp: pilot premium, scrap exposure, launch labor, and delay cost, plus how to quote the launch phase without burning margin.

Ramp cost is not steady-state part cost with a rounding error added. During launch, every unit carries a premium of 110% to 300% over the mature cost, driven by slow cycles, manual operations, extra inspection, low first-pass yield, and soft tooling. A part quoted at a $180 steady-state cost can run $592 loaded during a 250-unit pilot, roughly 3.3 times higher. The estimator's first job is to separate that temporary launch spend from the production part price. Quote them as distinct lines, because a customer who sees pilot cost folded into piece price will assume your production number is uncompetitive.

Build the pilot quote from four buckets, not one blended figure. Variable premium cost scales with volume: pilot units times steady-state cost times the premium factor. Fixed setup covers soft tooling, fixtures, programming, and line setup that does not move with quantity, often $50,000 to $150,000 on a new line. In the worked example, $63,000 of variable premium sits on top of $85,000 of fixed setup, so setup is more than half the $148,000 total. That split matters for negotiation and accounting: the fixed adder is frequently the capitalizable portion while the variable premium is expensed launch cost. The Pilot Build Cost calculator keeps the two apart.

Scrap is the quietest ramp cost and often the largest. At 3.2% first-pass yield early in a build, you scrap or rework roughly 97 of every 100 units started, and each carries material plus the labor and machine time already sunk into it. Even at a healthier 70% ramp yield, a $180 part means every 10 good units cost you the material and processing of about 14 starts. Price the ramp on units started, not units shipped, and carry a yield assumption per gate. The Ramp Scrap Exposure calculator converts a yield figure into expected scrap and rework dollars so the number lands in the quote, not the variance report.

Launch labor is a separate line from production labor and estimators routinely miss it. Beyond the operators sized by output, a ramp needs engineering and quality hours for containment, sorting, first-article inspection, and daily launch stand-ups, plus the hours to qualify equipment and transfer the process into the plant. On a typical new line these support hours can add 15% to 40% on top of direct labor for the first several weeks. Size them explicitly with the Launch Support Labor, Line Qualification Workload, and Process Transfer Workload calculators rather than hiding them in an overhead rate that was set for a mature line.

Overhead and machine-time absorption distort ramp economics because the line runs slow and half-empty. If a station is planned for 85% utilization at full rate but runs 25% during ramp, the fixed burden per good unit is roughly three to four times the steady-state absorption. Estimators who apply the standard machine-hour rate to ramp volume will either understate cost, if the rate assumes full loading, or scare the customer, if they load the whole burden onto a handful of units. State the utilization assumption explicitly and let the Equipment Ramp Utilization calculator show where absorption normalizes as volume climbs.

The largest ramp cost never appears on a router: launch delay. Every week the date slips, you lose the margin on the volume that week would have shipped, plus carrying cost on inventory and idle readiness spend. On a program shipping 2,000 units a week at $40 contribution margin, a one-week slip is $80,000 of lost margin before any penalty clauses. The Launch Delay Cost calculator quantifies margin lost per week, and Ramp Revenue Risk weighs the revenue exposed if the ramp simply runs slower than committed. A buffer that costs $15,000 to hold can be cheap against an $80,000-per-week slip.

A defensible ramp quote states its assumptions as numbers a customer can challenge. Name the premium factor and where it came from (prior launches of similar parts), the per-gate yield assumption feeding scrap, the utilization assumption feeding overhead, and the support-labor weeks. Separate one-time setup from per-unit cost, and show the loaded-to-steady-state ratio so the buyer sees pilot cost as temporary. A quote that reads 'pilot at 140% premium, converging to steady state by build four' is far harder to argue with than a single inflated piece price with no explanation behind it.

Ramp estimates go wrong in predictable ways. Quoting pilot units at steady-state cost ignores the premium entirely and guarantees a loss on the launch phase. Forgetting the fixed setup adder, which can exceed half the total, understates the bid badly. Assuming pilot per-unit cost predicts production cost bakes launch inefficiency into the production quote and loses the award. And leaving launch-delay and scrap exposure out of the risk case means the program has no financial argument for buffer stock or an earlier readiness spend, so it slips anyway and the cost lands as a variance nobody forecast.

Published 2026-07-01.