What is should-cost modeling?

If your negotiation strategy starts with a supplier’s quote, you are always negotiating from their number, not yours. That is the position most procurement teams are stuck in — reacting to whatever price lands in the RFQ response, with no independent basis to push back. Should-cost modeling flips that: it gives your team a defensible, data-built estimate of what a product or service should actually cost, so the negotiation starts from evidence instead of leverage alone. This guide covers what a should-cost model actually contains, how to build one, and where it pays off most in energy and industrial procurement.

What is should-cost modeling?

Should-cost modeling is the process of building an independent cost estimate for a product or service from its underlying cost drivers, rather than accepting a supplier’s quoted price at face value. The model typically breaks a category down into raw material cost, labor and processing time, overhead allocation, logistics, and a reasonable supplier margin. Once assembled, that should-cost figure becomes the baseline your negotiation team works from — not the number the supplier proposed. It is a core technique in strategic sourcing, and increasingly common in should-cost analysis programmes across capital-intensive industries where a single contract can carry millions in avoidable premium.

How does should-cost modeling differ from price benchmarking?

Price benchmarking compares what others have paid; should-cost modeling calculates what the item should cost regardless of what anyone has paid before. The distinction matters because benchmarks can be inflated across an entire market — if every supplier is overpricing a category, benchmarking just confirms the inflated number as normal.

• Basis. Benchmarking rests on historical prices paid and market data. Should-cost rests on underlying cost drivers — materials, labor, overhead.
• Best for. Benchmarking suits commodities with transparent, liquid markets. Should-cost suits complex, custom, or opaque-margin categories.
• Risk. Benchmarking inherits market-wide overpricing. Should-cost requires accurate cost-driver data.
• Negotiation posture. Benchmarking says “others pay less”. Should-cost says “here is what this should cost, and why”.

Most mature procurement functions use both — benchmarking for fast categories, should-cost modeling where the stakes and complexity justify the extra rigour.

What cost elements go into a should-cost model?

A complete should-cost model separates a price into five components: materials, labor, overhead, logistics, and margin. Each is estimated independently, then summed to produce the target cost.

• Materials: raw input costs, indexed to current commodity prices where relevant.
• Labor: hours required multiplied by a realistic regional or role-based rate.
• Overhead: a supplier’s facility, equipment depreciation, and administrative allocation.
• Logistics: freight, customs, and handling from origin to point of use.
• Margin: a reasonable profit allowance — not zero, but not whatever the market will bear either.

How do you build a should-cost model step by step?

Building a should-cost model follows five steps: select the category, decompose the cost structure, source input data, validate the model, and use it in negotiation.

1. Select high-impact categories first. Prioritise by spend and complexity — the Pareto principle applies here as it does in most procurement diagnostics.
2. Decompose the cost structure into the five elements above for the specific product or service.
3. Source reliable input data — commodity indices, labor rate surveys, freight benchmarks, or internal engineering estimates.
4. Validate against known data points, such as past invoices or a category where you already have cost transparency.
5. Bring the model into the negotiation as the opening position, not a fallback if talks stall.

Where does should-cost modeling deliver the most value in energy procurement?

Should-cost modeling pays off most in categories where specifications are complex and supplier margins are opaque — exactly the profile of most drilling services, EPC packages, and long-cycle equipment contracts in oil, gas, and energy infrastructure. Unlike a retail commodity with a public spot price, these categories rarely have transparent market benchmarks, which is precisely why a from-first-principles cost model carries so much negotiating weight.

It also complements a broader cost-and-value programme: should-cost modeling strengthens the rates lever, while specification and efficiency work capture the remaining share of the value that most cost programmes leave on the table — the pattern examined in why rate cuts miss 70 percent of savings potential.

What are the common pitfalls when implementing should-cost models?

• Treating the model as a one-time exercise rather than updating it as input costs shift.
• Using stale or generic labor and material rates instead of region- and category-specific data.
• Presenting the model as an ultimatum rather than a starting point for a fact-based conversation.
• Skipping validation against any real data point, which lets small assumption errors compound into a model nobody trusts.
• Building models only for direct materials and ignoring services categories where the same logic applies.

Frequently asked questions

Is should-cost modeling the same as target costing?

No. Target costing starts from a price the market will bear and works backward to a design. Should-cost modeling starts from actual cost drivers — materials, labor, overhead — and builds a fair price forward, independent of what a supplier chooses to quote.

How accurate does a should-cost model need to be to be useful?

A model within 5-10% of actual cost is usually accurate enough to change a negotiation. The goal is not a perfect number — it is a defensible one that shifts the conversation from a quoted price to a cost structure the supplier has to respond to.

Can should-cost modeling work for services, not just materials?

Yes. Service categories substitute labor hours, skill-level rates, and overhead ratios for materials and freight. It is harder to benchmark than physical goods, but the same should-cost logic applies to contract labor, maintenance, and technical services.

How often should a should-cost model be updated?

Revisit it whenever a major cost input moves — steel, freight rates, energy prices — or at minimum once a year for recurring categories. Contracts with price-adjustment clauses should have models refreshed on the same cycle as the clause triggers.

Do you need special software to build a should-cost model?

No — a well-structured spreadsheet with reliable cost-driver data is enough to start. Dedicated should-cost or spend-analytics software helps at scale, mainly by automating data collection and market-index updates, not by changing the underlying method.

Summary

• Should-cost modeling builds an independent cost estimate from materials, labor, overhead, logistics, and margin — not from what a supplier quotes.
• It delivers the most value in complex, opaque-margin categories like drilling services and EPC packages, where public benchmarks do not exist.
• A model does not need to be perfect to be useful — it needs to be defensible enough to change the negotiation.

See it working — book a demonstration

Continue reading: Cost & Value Intelligence — the philosophy behind cost and value decisions.

Related service: should-cost based tendering. Overview: oil and gas procurement.

References

1. Should-Cost Modeling: Because You Must Get the Cost Right. GEP. Accessed 26 August 2026.
2. Procurement 101: Unlocking Savings and Efficiency with Should-Cost Analysis. CADDi. Accessed 26 August 2026.