Negotiate Against What the Part Costs, Not What the Supplier Quoted

How should-cost analysis replace a single quoted number with an independent, bottom-up cost model, and what that changes at the negotiation table.

For: purchasing and sourcing directors, category managers, VP supply chain, and design engineers at automotive, aerospace, and industrial manufacturers.

Quick answer
A supplier quote is a starting position, not the cost of the part. Should-cost analysis replaces that single number with an independent, bottom-up model of what the component should cost to make, built from its own geometry and manufacturing route: material, process, labor, overhead, and tooling. That model, not the quote, becomes the reference point for the negotiation, so the conversation runs on verifiable cost drivers instead of a percentage off a number the supplier chose. The method is not new. It is the same discipline the U.S. Federal Acquisition Regulation calls a "should-cost review," and in two recent Goken engagements it traced a should-cost that landed 53 percent below a client's target on one machining component, and cut production cost 14 percent on a consumer product rebuilt entirely from a physical teardown.

The quote is a starting position, not a cost

Most supplier negotiations start and end with one number: the supplier's quote. From there, procurement pushes down. A few percent here, a volume concession there. The problem is structural. If the quote is the only number on the table, then every round of the negotiation is measured against a figure the supplier chose, with the supplier's own margin, assumptions, and legacy pricing already inside it. The buyer is haggling, not costing.

Should-cost analysis changes what the negotiation is about. Instead of arguing over a discount, both sides work from an independent estimate of what the part should cost to produce, built from the part itself. Goken's cost engineering team describes should cost as the optimum cost a product should incur considering the best available manufacturing and regional processes, covering material, process, assembly, and overhead. (Goken, Cost Engineering) When that model exists, the quote stops being the anchor and becomes one data point to be explained.

What "should cost" actually means, and where it comes from

Should-cost analysis is a specialized form of cost analysis that does not assume the supplier's existing costs are efficient. This is worth stating plainly because it is the whole idea, and it is not a consulting invention. It is written into U.S. federal procurement law.

The Federal Acquisition Regulation defines a "should-cost review" as a method that, in its own words, does not assume that a contractor's historical costs reflect efficient and economical operation, and instead evaluates the economy and efficiency of the contractor's workforce, methods, materials, equipment, and management. Its stated objective is to reduce the cost of performance by finding those inefficiencies. (FAR 15.407-4, current version effective March 2026)

The U.S. Department of Defense formalized the same idea as a management practice under its Better Buying Power initiative in 2010, where a program's should-cost estimate is used as a tool to actively drive cost down rather than accept a historical estimate as fixed. (Defense Acquisition University, Should Cost) The takeaway for a commercial buyer is the same one the government reached decades ago: the number you are quoted carries the supplier's inefficiencies with it, and the only way to see them is to build the cost from the part up.

What goes into a real cost model

A defensible should-cost model is built from the part's geometry and its manufacturing route, not from a spreadsheet of last year's prices. Goken's process starts from a sourcing request with the CAD, drawings, and specifications, selects the best manufacturing method for the geometry, and builds the cost up from there. (Goken, Cost Engineering) Five inputs carry most of the number:

• Material. The raw material spec, grade, and form, priced against current market base rates rather than a legacy agreement. Material is often the single largest line, and the easiest to get wrong when an old rate is carried forward.
• Process and manufacturing. The actual sequence of operations the geometry requires, with real cycle and setup times. Whether a feature is laser cut, sheared, drilled, or pierced changes the per-part cost materially.
• Labor. Time on each operation at the regional labor rate, tied to measured machine time rather than a standard allowance.
• Overhead. Plant burden, utilities, and indirect cost applied to the operations, plus a reasonable margin. This is where a lot of a quote hides.
• Tooling. Dedicated tools, dies, and fixtures amortized over volume. Dedicated tooling can raise unit cost at low volume and lower it sharply at scale, which is exactly the kind of trade a model makes visible.

Logistics and packaging sit on top where they are significant, particularly on re-sourcing across regions. The point of the model is not precision to the cent. It is a line-item structure that lets a buyer and a supplier disagree about a specific number, with evidence, instead of trading round percentages.

Why supplier quotes sit above the should-cost number

Quotes drift high for reasons that are usually structural, not adversarial. Suppliers are not necessarily overcharging on purpose. The cost model simply surfaces things a quote is not built to reveal. In practice the same few causes come up repeatedly.

The FAR framing is the general case: historical cost does not reflect efficient operation. (FAR 15.407-4) Goken's machining and sheet metal case study is the specific case, and it is a clean illustration of two of the most common causes. On that engagement, the derived should-cost of a machining component came in 53 percent below the client's target cost, a gap large enough that the team had to defend it line by line before it was accepted. The gap traced to two root causes: a raw material rate locked into an outdated, low-volume supply agreement that no longer matched the market, and process costs based on standard cycle times rather than measured actual times. (Goken case study, 13 July 2026)

Read across engagements, the recurring causes are: legacy pricing that outlived its market conditions, standard or estimated times used where measured times would be lower, a process chosen early and never revisited after the geometry stopped needing its flexibility, region and currency assumptions baked into an old quote, and normal information asymmetry, since the supplier knows its own cost structure and the buyer does not. A should-cost model does not accuse the supplier of any of these. It just makes them checkable.

What independent cost modeling actually recovers

The honest headline is that savings vary by part, commodity, and how far the quote had drifted, so no single percentage is a promise. What follows are Goken's own reported results from named, recent engagements, each traceable to the published case study. Use them as evidence of what the method finds, not as a guaranteed number for any given part.

Two of these deserve a note. On the machining component, re-pricing the raw material to quarterly base rates with an index correction and opening the lathe operations to comparison closed the majority of the 53 percent gap without touching the part. (53% gap case study) On the teardown, a consumer product was rebuilt into a full bill of materials from a physical unit with no design files, and part and fastener consolidation cut cost and weight from the same design changes, because every fastener removed takes out both its material and the labor to install it. (Teardown case study, 10 July 2026)

Where this shows up in an engineering workflow

Should-cost analysis is not only a purchasing tool. It maps onto decisions engineering teams already own, and Goken's cost capabilities line up with each one.

• Sourcing a new part. Build the cost model at the design stage and compare should-cost across design options before a supplier is even selected. This is New Product Development in Goken's capability set: cost insight at the design stage.
• Re-sourcing during a supplier switch. Carry an independent baseline into the new supplier conversation so the switch is priced on cost, not on whatever the incumbent happened to charge. This is Purchasing and Supplier Development: should-cost of parts sourced across geographies, and cost models for specific suppliers.
• Costing a material or design change. Model the cost impact of a material substitution or a tolerance change before committing. This is where should cost meets value engineering, which Goken states can reduce cost 15 to 30 percent without compromising performance.
• Competitor teardown benchmarking. Tear down a competitor product, rebuild its BOM, and benchmark your own cost and design against it, even with zero starting documentation, as the teardown case study shows
  

Each of these is listed on Goken's cost engineering page under Spend Analysis, Value Engineering, Supplier Development, New Product Development, and Purchasing. (Goken, Cost Engineering)

The honest limits

A should-cost model is a rigorous estimate, not a guaranteed price, and it is worth being clear about what it does not do. It depends on the quality of the geometry and assumptions that go into it, so a model built from incomplete drawings or wrong volume assumptions will be wrong in the same direction. A should-cost number below the quote still has to be defended: in Goken's machining engagement, a 53 percent gap was not accepted on the strength of the number, it required a full line-item breakdown first. Savings vary widely by commodity and by how far the quote had drifted, which is why the results above are a range and not a rule. And a should-cost model does not replace the supplier relationship. It gives both sides a factual basis to work from. Used well, it makes the negotiation shorter and the outcome durable. Used as a blunt instrument, it just moves the argument.

How to run a should-cost check on one high-spend part

If you want to test the method without committing a program to it, run it on a single part where the spend is high and the quote has never been independently checked. A first pass looks like this:

1. Pick the part. Choose one high-annual-spend component with a stable design and a quote no one has ever built up from scratch.
2. Pull the inputs. Gather the CAD, drawings, specifications, current volume, and the current quote with its date.
3. Price the material against today's market. Re-price the raw material to current quarterly base rates, not the rate in any existing agreement. Note the delta.
4. Rebuild the process from the geometry. List the operations the part actually needs and ask whether a cheaper process meets the same geometry, for example shearing and piercing instead of laser cutting on simple features.
5. Use measured times, not standards. Where you can, base cycle and setup time on measured data rather than standard allowances.
6. Add labor, overhead, tooling, and a reasonable margin. Build the full line-item cost, then compare it to the quote and locate the gap by line, not in total.
7. Take the breakdown to the supplier. Discuss the specific lines where model and quote disagree. That conversation, not the discount ask, is where should cost earns its keep.

A low-commitment way to start

Pick one high-spend part and have it modeled bottom-up. Goken runs should-cost teardowns on individual components and full products, from CAD-based cost models to physical teardowns of undocumented parts, and hands back a line-item breakdown you can take straight into a supplier meeting. If a program has a cost gap that needs to be trusted before it can be used, that is the entry point. Talk to the Goken should-costing team.

Sources

Every factual claim above traces to one of these primary sources. Goken engagements are first-party and dated; the FAR and DAU references are U.S. government publications.

Goken Global (first party)

• Cost Engineering, VAVE & Should Costing Solutions.
• Should-Cost Analysis for Machining and Sheet Metal Components: How Engineering Closed a 53% Cost Gap, 13 July 2026
• Hair Straightener Teardown for Benchmarking: 14% Cost Reduction and 8.75% Weight Savings, 10 July 2026.

U.S. Government (primary authority)

• Federal Acquisition Regulation 15.407-4, Should-cost review (FAC 2026-01, effective 13 March 2026).
• Defense Acquisition University, Acquipedia: Should Cost.