Should-Cost Modeling

4.1 When and Where to Apply Should-Cost Analysis

Should-cost analysis is most useful when the organization needs to understand why a component costs what it does and whether that cost is economically justified. It is particularly valuable for custom or semi-custom parts whose prices cannot be validated through a simple market comparison. A standard bearing or commodity fastener may have transparent competitive pricing; a machined housing, molded assembly, fabricated enclosure, electronic module, or proprietary packaging format often does not. In these cases, process economics provides a stronger reference point than historical price.

The method can answer several different questions. Procurement may want to determine whether a supplier quotation reflects reasonable manufacturing economics. Engineering may want to compare two material or geometry options before releasing a design. A design-to-value team may need to understand how much cost could be removed by changing a manufacturing process. Management may want to establish a component target before requesting quotations. The same modeling framework can serve each purpose, but the required precision and assumptions differ.

Should-cost analysis is usually justified when at least one of the following conditions exists:

• Material spend is significant. The component or assembly represents enough annual spend that a modest percentage improvement can create meaningful economic value.
• Price transparency is low. The item is customized, supplier quotations vary materially, or few comparable market prices are available.
• Design choices drive cost. Material grade, geometry, tolerances, finishes, component count, or processing requirements materially influence the economics.
• Supplier negotiations need evidence. The organization needs a fact base stronger than a general request for a price reduction.
• Alternative processes are plausible. A component might be machined, cast, forged, molded, stamped, fabricated, or otherwise produced through different routes.
• New-product economics are uncertain. Early supplier quotations or engineering estimates need to be tested before they become embedded in a business case.

Not every component warrants a detailed model. Should-cost work requires technical data, process knowledge, and analytical effort. Modeling hundreds of low-value parts can consume a team while producing little incremental savings. A useful prioritization considers annual spend, absolute price gap, design flexibility, supplier leverage, process complexity, and confidence in the current price. A $2 part purchased in enormous quantities may warrant detailed analysis; a $2 part used several hundred times annually usually does not.

The team should also decide whether the purpose is price challenge or design challenge. A price-challenge model holds the design essentially constant and asks what an efficient supplier should require to manufacture it. A design-challenge model asks how the economics would change if material, tolerances, geometry, cycle time, part count, or manufacturing route changed. Confusing the two can create unrealistic negotiation targets. A supplier cannot reasonably deliver the economics of a redesigned part while continuing to manufacture the existing specification.

Should-cost models are especially powerful when used jointly by engineering and sourcing. Engineers understand which specifications and process assumptions are technically required; sourcing understands supplier markets, quotation structures, capacity conditions, and commercial behavior. Effective procurement support connects these perspectives so that the model becomes both technically defensible and commercially useful.

Finally, should-cost is a decision tool, not an assertion of exact truth. A model represents what a reasonably efficient production system should cost under specified assumptions. The actual supplier may have different equipment, labor rates, utilization, scrap, purchasing contracts, overhead, capital structure, or strategic considerations. The objective is to make those differences explicit. When the modeled cost and quoted price diverge, the gap becomes a question to investigate rather than automatic proof that the supplier is overcharging.

4.2 Building a Component Should-Cost Model Step by Step

A useful should-cost model follows the physical flow by which the component is created. The analyst starts with the finished part, understands its design requirements, determines the likely manufacturing route, calculates the resources consumed at each stage, and then adds the commercial economics required to convert manufacturing cost into a sustainable supplier price. This process creates traceability between technical decisions and dollars.

The first step is to define the object being modeled. Record the part number, revision, annual volume, drawing, specification, dimensions, weight, material, manufacturing location, current supplier, quoted or actual price, and any tooling or packaging arrangements. Where multiple variants share the same process, identify whether a family model can be built rather than independently modeling every part.

The second step is to establish functional and technical requirements. Review the drawing and specification for tolerances, surface finish, heat treatment, coatings, testing, cleanliness, certification, traceability, cosmetic requirements, and special characteristics. These requirements often explain processing steps that are invisible when looking only at the final component. The analyst should not assume that an expensive operation is unnecessary simply because its purpose is not immediately obvious.

The third step is to determine the starting material and material quantity. For a machined metal component, the relevant input may be bar stock, plate, billet, forging, or casting rather than finished-part weight. For an injection-molded component, the model needs resin consumption plus runner or process loss. For a stamped component, coil width, nesting efficiency, and scrap become important. The model should distinguish finished material retained in the product from material purchased but lost during production.

The fourth step is to define the manufacturing routing. Break the process into meaningful operations such as cutting, forming, molding, casting, machining, welding, heat treatment, coating, cleaning, inspection, testing, assembly, and packaging. For each operation, identify the equipment or labor resource required. The routing should describe an efficient, technically feasible production process rather than mechanically replicate the incumbent supplier’s current method.

The fifth step is to estimate cycle time. Cycle time may come from process equations, equipment benchmarks, historical analogues, supplier observations, engineering judgment, or time studies. Automated machine time and manual labor time should be distinguished where appropriate because their economics differ. For batch processes, setup time should be allocated across the expected batch quantity rather than charged in full to each component.

The sixth step is to assign appropriate rates. Machine rates should reflect equipment economics, utilization, maintenance, energy, and other costs included in the chosen convention. Labor rates should distinguish direct wages from burdens such as benefits, payroll taxes, and shift premiums if those items are not captured elsewhere. Rates should be appropriate to the assumed manufacturing geography rather than automatically borrowed from the client’s own plant.

The seventh step is to model yield, scrap, and rework. If 100 input units produce only 97 acceptable units, the cost of the three failed units must ultimately be borne by the good production. Material loss may occur at one stage while quality yield loss occurs at another. Analysts should avoid using a single generic scrap percentage when the major economic loss can be modeled more directly.

The eighth step is to incorporate secondary and externally purchased content. A fabricated assembly may contain bought-out fasteners, seals, electronics, labels, or coatings performed by another supplier. These items should be estimated separately rather than buried inside an unexplained conversion rate. Purchased subcomponents with large value may require their own should-cost models.

The ninth step is to add tooling, capital, overhead, and supplier margin according to transparent conventions. These elements should not be used as balancing figures to force the model toward the quoted price. Each should have an independent rationale based on useful life, expected volume, capital requirements, business complexity, and reasonable economics.

The tenth step is to calculate the modeled piece price and reconcile it with actual or quoted economics. A clear cost bridge should show material, conversion, purchased content, tooling amortization, overhead, logistics where included, and supplier profit. If the modeled result is $14.20 against a supplier quote of $18.50, management should be able to see which assumptions create the $4.30 difference and which questions need further investigation.

The model should then be subjected to sensitivity analysis. Volume, raw-material price, labor rate, cycle time, utilization, yield, tooling life, and supplier margin can each influence the result. Testing realistic ranges identifies which assumptions matter most. A conclusion that remains directionally unchanged across plausible ranges is much stronger than one that depends on a single uncertain cycle-time estimate.

Throughout construction, the analyst should preserve sources and assumptions. Every major input should indicate whether it comes from a drawing, market benchmark, supplier quotation, plant observation, technical calculation, database, or expert estimate. A should-cost model becomes reusable institutional knowledge only when another practitioner can understand and update it without reconstructing the original analyst’s reasoning.

4.3 Modeling Materials, Labor, Manufacturing Processes, Tooling, Overhead, and Margin

The credibility of a should-cost model depends less on spreadsheet complexity than on disciplined treatment of its economic building blocks. Materials, labor, processing, tooling, overhead, and margin behave differently and should be modeled separately. Combining them into broad percentages may produce a plausible total while obscuring the drivers that engineering or sourcing can actually challenge.

Materials: Start with the technically appropriate raw-material form, not simply finished-part weight. Calculate gross input quantity based on geometry, nesting, machining allowance, runner systems, trim, kerf, or other process losses. Apply a material price consistent with grade, form, geography, purchase scale, and time period. Where scrap has recoverable value, subtract realistic scrap revenue rather than assuming all unused material is a total loss.

Material assumptions should also reflect purchasing economics. A large processor may buy resin, steel, aluminum, fiber, or chemicals under contracts unavailable to a small supplier. Commodity indices can provide useful references, but conversion premiums, alloy additions, minimum quantities, and regional delivery costs may matter. The analyst’s task is to estimate reasonable material economics for the supplier scale being modeled, not simply apply the lowest market price that can be found.

Labor: Separate labor content from labor rate. Labor content is the number of minutes or hours required to operate, handle, inspect, assemble, test, package, or support the component. Labor rate is the cost per hour for that work. Improvement opportunities may come from either dimension: simplifying assembly reduces content, whereas relocating production changes the rate. Keeping them separate prevents teams from confusing product-design opportunities with footprint economics.

Automation requires careful treatment. An automated process may have little direct labor but significant equipment cost, maintenance, and capital recovery. Conversely, low-volume production may economically favor manual work. The modeled process should reflect expected annual volume and practical utilization rather than automatically assuming maximum automation. A theoretically fast automated line is not lower cost if demand cannot support the required capital.

Manufacturing processes: For each operation, estimate the economic resource consumed. Machining may depend on machine-hour rate, cutting time, tool consumption, and setup. Injection molding depends on machine tonnage, cycle time, cavity count, resin, tooling, and yield. Stamping depends on press rate, strokes, die configuration, material utilization, and run speed. Welding may depend on joint length, process type, fixtures, consumables, robot or operator time, and inspection requirements.

The model should reflect process interactions. A tighter casting tolerance may increase casting cost but reduce machining. A more expensive near-net-shape forging may reduce material waste and cycle time. A multi-cavity mold increases tooling investment but reduces piece conversion cost. Should-cost modeling is most valuable when it evaluates the total economics of the process chain rather than optimizing each operation independently.

Tooling: Tooling should be treated according to ownership and commercial arrangement. A customer may pay separately for a mold, die, fixture, gauge, or cutting tool, in which case it should not also be fully embedded in the piece price. Where the supplier funds tooling, the model should amortize the investment across an appropriate expected volume or contract period and consider financing or return requirements if material.

Tool life matters. A die expected to produce one million parts has different unit economics from one requiring refurbishment every 100,000 cycles. Likewise, prototype or bridge tooling should not be treated as equivalent to long-life production tooling. Analysts should record tooling cost, useful life, maintenance assumptions, residual value where relevant, and the volume over which recovery occurs.

Overhead: Overhead should cover legitimate indirect resources required to operate the manufacturing business but should not become an unexplained markup. Relevant items can include supervision, plant support, maintenance, quality, indirect labor, utilities not included in machine rates, facility costs, information systems, engineering support, and corporate expenses. The appropriate allocation basis may be labor, conversion cost, machine hours, floor space, or another driver depending on the business.

Double counting is a common error. If the machine rate already includes depreciation, maintenance, energy, and factory occupancy, those costs should not be added again through a broad plant-overhead percentage. Similarly, labor burden should not appear in both the hourly labor rate and an overhead markup unless the distinction is explicit. The model should document what each rate includes before adding further layers.

Supplier margin: Profit should be modeled separately from manufacturing cost. A supplier needs an economically reasonable return, and a should-cost model that assumes zero margin is not a credible commercial benchmark. The appropriate margin depends on industry structure, intellectual property, capital intensity, capacity utilization, switching costs, risk, service requirements, and competitive conditions.

Margin should also be applied to an appropriate base. Some models use an operating-margin assumption on sales; others use a markup on cost. These are not mathematically equivalent. The convention should be clear. For example, a 10 percent margin on sales requires a higher price than a 10 percent markup on cost. Misunderstanding this distinction can create material errors when large purchased content is involved.

Finally, the model should show both current-design economics and credible alternatives. If changing material, relaxing a tolerance, reducing machining, consolidating components, or altering process selection reduces cost, management should be able to see exactly which cost blocks change. This is what converts should-cost from a procurement calculator into a design-to-value tool.

4.4 Testing Assumptions and Reconciling Should-Cost to Quoted or Actual Cost

A should-cost result should never be accepted simply because the arithmetic works. The most important analytical work often begins after the first model is complete. Teams must test whether the process is feasible, whether the assumptions reflect the relevant manufacturing environment, and why the result differs from supplier quotations or actual prices. Reconciliation creates credibility and often exposes the most valuable commercial or technical questions.

The first test is internal consistency. Verify units, currencies, material quantities, annual volumes, cavity counts, setup allocations, machine utilization, labor minutes, scrap treatment, tooling amortization, and margin calculations. Simple errors can create large apparent gaps. A model should include automatic checks where practical, such as confirming that process yields reconcile, annual tooling recovery does not exceed the intended amount, and calculated manufacturing capacity can support forecast volume.

The second test is technical plausibility. Engineering or manufacturing specialists should review the proposed process route and key cycle assumptions. Could the geometry actually be molded in the assumed cycle? Can the selected machine achieve the tolerance? Does the casting require heat treatment? Is manual assembly possible in the modeled time? Does inspection require specialized equipment? Expert review is especially important when the model crosses technologies outside the analyst’s direct experience.

The third test is external triangulation. Compare the model with analogous client parts, previous quotations, alternate supplier bids, market benchmarks, teardown estimates, commodity references, or independent expert input. No single comparator needs to match exactly. The objective is to determine whether the order of magnitude and major cost drivers are consistent with observable evidence.

When a quoted or actual price exceeds should-cost, decompose the difference rather than labeling it supplier margin. Potential explanations include lower supplier volume, unfavorable raw-material purchasing, inefficient equipment, excess setup, low utilization, quality requirements, expedited logistics, tooling recovery, currency, capacity constraints, risk premiums, minimum-order economics, engineering support, or genuinely higher profit. Each hypothesis suggests a different action.

The same discipline applies when the quoted price is below the model. The supplier may have better material purchasing, a more efficient process, depreciated equipment, higher capacity utilization, lower labor cost, common tooling, cross-customer volume, or a deliberate commercial strategy. The analyst should first learn from the difference rather than assume the model is correct and the supplier is behaving irrationally.

Sensitivity analysis should focus on variables that could change the conclusion. A model with a $3 gap may not support a strong challenge if a plausible cycle-time range changes calculated cost by $4. In contrast, a $6 gap that remains at least $5 across reasonable labor, material, and utilization scenarios provides a robust basis for discussion. Tornado charts or simple low/base/high tables can show management which variables matter.

Supplier engagement can then be used to refine the fact base. The objective is not necessarily to disclose every model assumption. Buyers can ask targeted questions about material yield, process route, cycle time, tooling recovery, labor content, capacity, overhead treatment, or unusual cost drivers. In effective strategic sourcing discussions, the model changes the conversation from “reduce your price by 8 percent” to “help us understand why this operation requires the cost implied in your quote.”

Reconciliation should ultimately classify the gap. Some difference may be supported by legitimate economics and accepted. Some may become a negotiation opportunity. Some may indicate that the existing specification forces unnecessary cost. Some may reveal that another supplier or manufacturing route is structurally advantaged. The model has succeeded when the organization understands which category applies and can take the appropriate action.

4.5 Should-Cost Model Template and Supplier Challenge Fact Base

A should-cost model must be usable by people beyond its original author. A standardized template improves review, allows models to be compared across parts, and creates a reusable repository of manufacturing economics. The template should be detailed enough to expose major assumptions while remaining simple enough that sourcing, engineering, finance, and manufacturing leaders can understand the result.

A practical should-cost model should contain the following sections:

• Part definition: part number, revision, description, annual volume, dimensions, weight, material, supplier, manufacturing region, current price, and baseline date.
• Technical requirements: key tolerances, finishes, treatments, testing requirements, certifications, and other specifications that materially influence manufacturing cost.
• Material calculation: raw-material form, gross input quantity, net finished quantity, material price, process loss, scrap recovery, and net material cost.
• Process routing: each manufacturing operation, equipment type, setup time, cycle time, labor content, yield, machine rate, labor rate, and calculated conversion cost.
• Purchased content: bought-out components, outsourced operations, packaging, and other externally sourced inputs.
• Tooling and capital: tooling investment, ownership, useful life, expected volume, maintenance, and unit amortization.
• Commercial economics: applicable overhead, logistics if in scope, supplier margin, and resulting modeled piece price.
• Reconciliation: quoted or actual price, should-cost result, absolute and percentage gap, principal explanations, and estimate confidence.
• Sensitivity cases: low, base, and high assumptions for the variables that materially influence the conclusion.

Each major assumption should carry a source. Useful source labels include drawing, supplier quotation, client purchasing data, commodity index, plant observation, process benchmark, expert estimate, and calculation. The model should also record the date of time-sensitive inputs such as commodity prices, exchange rates, energy costs, and labor rates. An unsupported number can still be used as an explicit hypothesis, but it should not be mistaken for verified data.

The supplier challenge fact base should be shorter than the underlying model. Its purpose is to prepare the cross-functional team for a commercial and technical discussion. A useful challenge pack identifies the current price, modeled range, largest gap drivers, evidence supporting those drivers, questions to ask the supplier, alternative process or design options, and the value at stake. Detailed calculations should remain available as backup rather than dominate the discussion.

Questions should be specific. Instead of asking why a component is expensive, ask what raw-material yield is assumed, why two machining operations are required, how tooling is being recovered, what annual volume drives the quotation, or what prevents use of a lower-cost process. The strongest questions expose assumptions that can be discussed and changed. They also make it easier to distinguish a commercial disagreement from a technical constraint.

The challenge pack should define negotiation boundaries before supplier discussions begin. The team should know its target, which model assumptions are strong, where uncertainty remains, which design alternatives are credible, and which concessions may have value to the supplier. Volume commitments, longer contracts, specification changes, batch-size flexibility, forecast visibility, or tooling arrangements may reduce supplier economics without simply transferring margin from one party to another.

Any price reduction agreed with a supplier should be separated from savings requiring a design change. If the supplier accepts a lower price for the unchanged part, that is principally a commercial outcome. If engineering changes material, tolerance, geometry, or process requirements and the supplier price falls as a result, the economics belong to the design change. This distinction is important for governance and for broader product cost reduction programs, where multiple workstreams may otherwise claim the same benefit.

Finally, completed models should become a reference library rather than disappear after a negotiation. Process rates, material assumptions, tooling economics, cycle-time benchmarks, and supplier insights can accelerate future product development and target costing. Models should be updated when major input assumptions change and retired when technology or manufacturing conditions make them obsolete. Used consistently, the should-cost library becomes a cumulative capability: each analysis improves not only the immediate sourcing or redesign decision but also the organization’s ability to understand and challenge product economics before cost becomes embedded in the next generation of products.