Supplier Should-Cost Breakdown for China Sourcing: Build an Evidence-Graded Model

A supplier should-cost model is a structured hypothesis about the resources and commercial assumptions behind a product—not a claim that you know the factory's books. Start with one controlled product revision and quotation scope. Model material, conversion, tooling, quality and packaging drivers as evidence-graded ranges. Then use the gaps to ask precise questions and test design, volume or process alternatives.

The result should improve the discussion even when the supplier will not disclose a full breakdown. It should not become a spreadsheet accusation that a supplier's overhead or margin is “wrong”.

Define the product and commercial boundary first

A should-cost built against an unclear scope will explain the wrong product. Lock these fields before calculation:

• drawing, specification, bill-of-material or approved-sample revision;
• order quantity and annual-volume scenario;
• material grade, finish, tolerances and test requirements;
• packaging configuration and labelling scope;
• Incoterm, named place, currency and payment assumptions;
• tooling ownership and whether tooling is separate or amortised;
• included inspection, certification or documentation; and
• quote date and validity.

Use the product specification sheet as the technical baseline and the supplier RFQ to keep commercial assumptions comparable. If the supplier quote covers a different revision, MOQ or Incoterm, fix that mismatch before interpreting a cost gap.

Build the model from physical drivers

Avoid beginning with arbitrary percentages such as “materials are always 50%”. Build from measurable product and process inputs where possible.

Materials and yield

For each material, distinguish finished-product net content from gross input. A blank, moulding shot, cut length or sheet nest can require more input than remains in the saleable unit. Use:

material cost per unit = gross input quantity × evidenced material price − documented recoverable scrap credit

If you only know net weight, create a yield range rather than a confident scrap percentage. Use a scrap credit only when evidence shows who receives the recovery value and how it enters the quoted scope. Do not assume all scrap is avoidable or recoverable.

When a supplier attributes a price change to a commodity, use the separate raw-material price-claim process to test the index, grade, content and effective period. An index move is not automatically the same as the product's cost move.

Conversion time and process

Map the main operations: cutting, moulding, machining, forming, finishing, assembly, testing and packing. For each operation, record the machine or line, cycle time, labour touch time, batch size and setup treatment.

Do not multiply every cycle minute by a fully loaded labour rate if the process runs automatically. Do not treat machine time and labour time as interchangeable. If the process uses several cavities or produces several units per cycle, calculate output per cycle explicitly.

Use ranges where the evidence is weak:

• observed or supplier-stated cycle time;
• plausible low and high utilisation;
• direct labour touch time; and
• machine or conversion rate source.

Tooling and one-off costs

Keep one-off tooling separate unless the quote explicitly amortises it. If the business chooses an amortisation scenario for commercial comparison, show the assumption:

tooling amount per unit = tooling cost allocated to scenario ÷ scenario units

That arithmetic does not change tooling ownership or guarantee tool life. Link it to the tooling record and state what happens when actual volume is lower or higher than the scenario.

Quality, testing and packaging

Quality is not a free residual. Include the agreed inspection, test, fixture, certification and record-keeping scope. Separate routine unit or batch controls from one-time product validation. If a required laboratory or certification cost is unknown, keep an open line rather than assigning zero.

Build packaging from the approved pack specification: inner pack, retail pack, inserts, carton, palletisation and print or setup. Packaging MOQ can make the effective unit cost different at trial volume, so keep unused packaging and write-off risk visible where it sits within the commercial scope.

Grade every input by evidence confidence

Attach a confidence label to each input. A model with visible uncertainty is more useful than a precise-looking total assembled from guesses.

An ISO-hosted Auditing Practices Group paper discusses defined criteria, current provider information, performance monitoring, verification and risk-based controls over external providers. The paper expressly says it has not been endorsed by ISO, ISO/TC 176 or IAF. It offers procurement-control context; it does not reveal a supplier's actual cost or prescribe this should-cost model.

Use ranges instead of invented precision

For uncertain inputs, calculate low, working and high scenarios. Preserve the driver behind each range.

Examples:

• material input varies with yield;
• conversion varies with cycle time or cavities;
• packaging varies with print run and MOQ;
• tooling amount varies with scenario volume; and
• currency-sensitive inputs vary with the selected rate date.

Do not compensate for an unknown overhead or margin by hiding it inside an inflated machine rate. Keep unresolved commercial layers visible.

Worked example: a fictional cost stack

The following figures are fictional and are not Chinese market rates. They show how to preserve a range rather than claim a supplier's true cost.

Assume the supplier quote is AUD8.10 per unit on the same fictional scope. The model does not prove a AUD0.90–1.90 overcharge. The difference may contain factory overhead, margin, financing, risk, minimum-order inefficiency, export handling, omitted buyer assumptions or errors in the model.

The next step is to test the largest uncertain drivers, not demand the low scenario as a target.

Test the model with sensitivity analysis

Change one driver at a time and observe the unit-cost movement.

Sensitivity analysis identifies leverage. It does not tell you which change is safe. Product engineering, quality and regulatory requirements remain hard gates.

Turn gaps into supplier questions

Ask neutral, evidence-led questions:

• Which product revision and annual-volume assumption does the quote use?
• What gross material input and yield are assumed for the key component?
• Which operation controls the cycle time?
• Is labour touch time separate from unattended machine time?
• Which testing, inspection and records are included?
• Is tooling billed separately, amortised, or both?
• Which packaging quantity and print setup underpin the unit cost?
• Which cost changes at the next MOQ, and why?
• What changed since the prior quote: input price, process, yield, scope, currency or commercial risk?

Treat a supplier refusal to disclose confidential data as a commercial constraint, not proof of misconduct. You can still ask for scope confirmation, change drivers and scenario pricing.

Negotiate the driver, not just the total

Use the model to structure options. A lower MOQ may raise packaging or setup cost. A design change may reduce material but require new tooling. A longer commitment may help purchasing efficiency but increases the buyer's demand and inventory exposure.

Compare the resulting offer using the supplier quote comparison process. If volume is the proposed lever, test it with the lower-MOQ negotiation framework rather than assuming the largest order is cheapest in total.

Preserve quality, compliance and supplier viability

Do not use should-cost analysis to pressure removal of required testing, inspection, compliant materials, traceability or safe process controls. A model that ignores those requirements is not lean; it is incomplete.

Also avoid treating supplier margin as waste. A supplier's price may need to cover people, equipment, maintenance, quality systems, working capital and risk; the mix is supplier-specific and often unobservable. The buyer's job is to understand the offer and improve the total commercial design, not declare a universal acceptable margin.

Escalate safety, regulatory, technical and accounting questions to the appropriate qualified specialist. Keep unresolved lines open.

Keep one controlled should-cost record

The final record should include:

• product and quote revision;
• quantity, annual volume, Incoterm and currency;
• cost blocks and formulas;
• low, working and high inputs;
• evidence reference, owner, grade and date;
• supplier questions and responses;
• approved model changes;
• scenario decision and rationale; and
• next refresh trigger.

Refresh the model when the specification, process, quantity, material basis, currency or commercial scope changes. Do not overwrite the prior version. A should-cost history is valuable because it shows whether the product changed, the evidence improved or the negotiation merely moved the total.

Sources

• ISO 9001 Auditing Practices Group — Auditing the Procurement and Supply Chain Processes