Blog  /  Guide

Guide

Should-Cost Analysis in Procurement: Methodology, Model and Worked Example

Cost It Right

Apr 28, 2026  •  Updated

Listen to this article

AI narration • 18:26 listen

0:00 / 18:26

Should-cost of ₹304.68 versus a supplier quote of ₹393.00, with the gap split by raw material, cycle time and other lines

Should-cost analysis is the bottom-up estimate of what a part ought to cost if a capable supplier makes it efficiently: raw material, process, overheads, rejection, inventory carrying cost, packing, freight, tooling and a fair profit. Buyers build it before or alongside a supplier's quote, then compare the two line by line. The gap between them is where the negotiation happens.

Most purchase teams compare five quotes and pick the lowest. That tells you who is cheapest. It does not tell you whether the cheapest price is fair, or which line in it is inflated.

In the worked example below, a supplier quotes ₹393 for a machined steel flange that should cost about ₹305. Two-thirds of that ₹88 gap comes from just two inputs: raw material weight and cycle time. Neither would show up if you only compared totals.

This guide covers what should-cost analysis is, what goes into the model, the eight steps to build one, that full example in rupees, and how to use the result without damaging the supplier relationship.

What is should-cost analysis?

Should-cost analysis estimates what a product should cost based on how it is actually made, instead of accepting what a supplier says it costs. It answers one question: if a competent supplier used sensible material, a realistic process and normal overheads, what would this part cost?

The idea comes from US government procurement. The US Federal Acquisition Regulation describes should-cost review as "a specialized form of cost analysis" that does not assume a contractor's historical costs reflect efficient operation, and examines "the economy and efficiency of the contractor's existing work force, methods, materials, equipment" (FAR 15.407-4, Should-cost review). Manufacturing buyers apply the same thinking to bought-out parts.

In an Indian auto component context, the same idea goes by several names:

  • Should-costing or a should-cost model: the cost estimate itself.

  • Zero-based costing (ZBC): building that estimate from zero, from the drawing and process, rather than from last year's price.

  • Cost breakup analysis: checking a supplier's submitted breakup against your own estimate.

Is should-cost the same as sourcing cost analysis?

No. Should-cost asks what the part should cost to make. Sourcing cost analysis asks what it will cost you to buy and receive it, including freight, duties, payment terms and supplier risk. Use should-cost to judge whether a price is fair, and sourcing cost analysis to compare suppliers in different locations.

What does a should-cost model include?

A should-cost model splits the price into ten elements and calculates each one from its own drivers. The first three are the direct cost of making the part; the rest are allowances on agreed bases.

Element

What drives it

How it is calculated

Net raw material

Grade, gross weight, RM rate, scrap rate

Gross weight × RM rate − recoverable scrap weight × scrap rate

Conversion (process) cost

Machine, cycle time, efficiency, MHR

Machine hour rate (MHR) × cycle time in hours ÷ efficiency, per operation; or a job-work rate for plating and heat treatment

Bought-out parts (BOP)

Fasteners, inserts, sub-parts

Quantity × BOP rate

Overheads

Plant costs not inside the MHR

% on conversion cost (base varies by OEM)

Rejection

Process scrap and inspection rejects

% on raw material plus conversion

Inventory carrying cost (ICC)

Money tied up in material

Raw material cost × interest rate × days of inventory ÷ 365

Profit

Supplier margin

% on conversion cost, or on total cost (agreed with the OEM)

Packing

Bins, trays, cartons

Per part, from the packing standard

Freight

Plant to customer

Trip cost ÷ parts per trip

Tool amortisation

Dies, moulds, fixtures

Tool cost ÷ committed volume, if not paid separately

Two details decide whether a model is useful or just a number:

  1. Every percentage needs its base written next to it. "10% profit" on conversion cost and "10% profit" on total cost can differ by several rupees per part.

  2. Every input needs its source and date. A should-cost built on last quarter's steel rate is not a should-cost. Record the RM base date, the index used and the volume assumed.

For the full formula and how each manufacturing process changes it, see our guide to auto component costing.

How do you do should-cost analysis, step by step?

Build the model from the drawing outwards: fix the part and the route first, then cost material, then each operation, then the allowances, and only then open the supplier's quote. Opening the quote first anchors you to the supplier's numbers.

  1. Collect the part data. Drawing and revision, material grade, finished weight, tolerances, surface treatment, annual volume, packing standard and delivery location. A wrong drawing revision makes everything after it wrong.

  2. Fix the process route. List every operation in order (for example: bar cut, CNC turning, VMC drilling, deburr and wash, blackening, inspection) and the machine class for each. This is where a cost engineer's process knowledge matters most.

  3. Work out gross weight and net raw material. Gross weight comes from bar length, blank size, forging stock or shot weight, depending on the process. Subtract only the scrap that can actually be sold, at the current scrap rate.

  4. Cost each operation. For each machine: MHR × cycle time ÷ efficiency. Use realistic cycle times from your own plant, a time study or a comparable part, not the supplier's. Add job-work rates for outsourced operations.

  5. Add the allowances on their agreed bases: overheads, rejection, inventory carrying cost and profit. Use your OEM's norms if you are a supplier, or your purchase policy if you are a buyer.

  6. Add packing, freight and tool amortisation. Amortise tools over the committed volume, and note that volume in the model.

  7. Compare with the supplier's breakup, line by line. Put both in the same format. Where they differ, write down which input differs (weight, rate, cycle time, percentage) and by how much.

  8. Record assumptions and refresh the model. Note the RM base date, index, volume and utilisation behind every number. When steel moves or volumes change, update the model before the next price discussion; our guide to raw material price volatility covers how indexed contracts handle this.

Steps 3 and 4 usually explain most of the gap between a should-cost and a quote, which the worked example below shows.

Worked example: should-cost vs a supplier quote in rupees

For a machined EN8 steel flange, the should-cost comes to ₹304.68 per part, and the supplier quotes ₹393.00. The ₹88.32 gap is 29% above should-cost, and the line-by-line comparison shows exactly where it comes from. All inputs are illustrative; replace them with your own rates and norms.

The part: EN8 steel flange, turned and drilled from bar, blackened, 1,50,000 parts committed over the programme.

Should-cost inputs

Input

Value used

Bar input weight per part

2.1 kg

Finished weight

1.5 kg (chips 0.6 kg)

EN8 bar rate

₹72 per kg

Scrap (chips) rate

₹22 per kg

CNC turning

MHR ₹750, 4.5 min cycle, 85% efficiency

VMC drilling

MHR ₹900, 2.5 min cycle, 85% efficiency

Deburr and wash

₹6 per part

Blackening (job work)

₹4 per part

Overheads

12% on conversion cost

Rejection

1.5% on raw material plus conversion

Inventory carrying cost

12% a year on raw material, 45 days

Profit

10% on conversion cost

Packing and freight

₹5 and ₹7 per part

Turning fixture

₹3,00,000 over 1,50,000 parts

Should-cost vs supplier quote, line by line

Element

Should-cost calculation

Should-cost (₹)

Supplier quote (₹)

Gap (₹)

What drives the gap

Net raw material

2.1 kg × 72 − 0.6 kg × 22

138.00

166.20

28.20

Supplier assumes 2.4 kg bar at ₹76, scrap credit at ₹18

CNC turning

750 × 4.5 ÷ 60 ÷ 0.85

66.18

88.24

22.06

6.0 min cycle instead of 4.5

VMC drilling

900 × 2.5 ÷ 60 ÷ 0.85

44.12

52.94

8.82

3.0 min cycle instead of 2.5

Deburr and wash

6.00

6.00

0.00

Blackening

4.00

4.00

0.00

Conversion cost

120.30

151.18

30.88

Overheads

12% × 120.30

14.44

22.68

8.24

15% on a higher conversion cost

Rejection

1.5% × (138.00 + 120.30)

3.87

9.52

5.65

3% on a higher base

Inventory carrying cost

138.00 × 12% × 45 ÷ 365

2.04

3.28

1.24

60 days of inventory instead of 45

Profit

10% × 120.30

12.03

18.14

6.11

12% on a higher conversion cost

Packing and freight

12.00

18.00

6.00

₹8 packing, ₹10 freight

Fixture amortisation

3,00,000 ÷ 1,50,000

2.00

4.00

2.00

Amortised over 75,000 parts

Total per part

304.68

393.00

88.32

(Each line is rounded to two decimals; totals are the sum of the rounded lines.)

What the comparison tells the buyer

Bar chart of the ₹88.32 gap: cycle time ₹30.88, raw material ₹28.20, overheads ₹8.24, profit ₹6.11, packing and freight ₹6.00, rejection ₹5.65, fixture ₹2.00, inventory carrying cost ₹1.24
  • Two inputs explain two-thirds of the gap. Raw material (₹28.20) and cycle time (₹30.88) add up to ₹59.08 of the ₹88.32. Challenge these first.

  • Percentages compound. The supplier's overhead, rejection and profit lines are higher partly because their rates are higher, and partly because they are applied to an inflated conversion cost. Fix the cycle time and these shrink on their own.

  • Some gaps may be legitimate. A 2.4 kg bar could be right if the supplier's bar size or cut-off allowance differs from your assumption. Ask before you assume padding.

  • Volume is a commercial question. The fixture line doubles because the supplier amortises over 75,000 parts. Confirm the committed volume in writing and the line goes back to ₹2.

On 1,50,000 parts, the full gap is worth about ₹1.32 crore over the programme. Even closing only the raw material and cycle time gaps saves about ₹88.6 lakh.

How do you use a should-cost gap in negotiation?

Use the gap to ask about specific inputs, not to demand a lower total. "Your price is high" invites a token discount. "Your turning cycle time is 6 minutes against our 4.5; can you walk us through the route?" invites a real conversation, and often a real correction.

A practical order for the discussion:

  1. Share the format, not your numbers, first. Ask the supplier to fill your standard cost breakup so both sides compare the same lines.

  2. Start with the biggest physical input. Weight and cycle time are facts that can be checked on the drawing or on the shop floor. Percentages are policy and come later.

  3. Ask, then show. Ask how they arrived at a number before showing yours. Suppliers often correct an assumption themselves.

  4. Agree the commercial terms separately. Committed volume, payment terms and RM index base date change the price legitimately. Settle them in writing, then rework the affected lines.

  5. Record what changed and why. Six months later, when steel moves, both sides need to know which base the price was agreed on.

Some example wording buyers use:

  • "Our model uses a 2.1 kg input from a 65 mm bar. What bar size and cut-off allowance have you assumed?"

  • "Your overhead is 15% on conversion. Our norm for this commodity is 12%. Which costs are in your 15% that are not in your MHR?"

  • "Can you confirm the fixture is amortised over 1,50,000 parts, in line with the nomination letter?"

When is a quote higher than should-cost justified?

A supplier can be fairly priced above your should-cost when the difference has a real cause: a tighter tolerance than your model assumed, a lower volume, a special packing requirement, a quality or PPAP requirement your model left out, or a genuinely higher cost location. The test is whether the supplier can point to the input that causes it. If they can, update your model. If they cannot, it is margin.

A quote far below should-cost deserves the same attention. It usually means a missing cost head, an unrealistic cycle time or a supplier buying the business, all of which tend to come back later as amendments.

For a line-by-line method to read a supplier's breakup, see our guide to supplier cost breakdown analysis. For more on why quotes and should-cost differ, see supplier quote vs should-cost.

How is should-cost different from ZBC, TCO and price analysis?

Should-cost tells you what a part should cost to make. The other methods answer different questions, and mature purchase teams use several of them on the same decision.

Method

Question it answers

Built from

Best used for

Should-cost analysis

What should this part cost if made efficiently?

Drawing, process route, rates, norms

Judging whether a quote is fair; negotiation

Zero-based costing (ZBC)

What does it cost if we build the estimate from zero, ignoring past prices?

Same as should-cost; the two terms are often used interchangeably

New part nomination; cost reduction on running parts

Vendor-based costing (VBC)

What does the supplier say it costs?

The supplier's submitted breakup

Line-by-line comparison against ZBC

Price analysis

Is this price in line with the market?

Competing quotes, past prices, catalogues

Commodity-like parts with many suppliers

Total cost of ownership (TCO)

What will this part really cost us over its life?

Price plus freight, duties, payment terms, quality, inventory, risk

Choosing between suppliers or locations

Cost-benefit analysis

Is this project or investment worth it?

Expected benefits against total costs

Capex, make-vs-buy, programme decisions

In practice: build the should-cost (or ZBC), compare it with the supplier's VBC to negotiate, then use TCO to choose between suppliers whose prices are already fair. See our guide on zero-based costing in procurement for how ZBC is set up.

What are the most common should-cost mistakes?

Most weak should-cost models fail on inputs, not on the formula. The mistakes we see most often:

  • Old rates. RM and scrap rates from a past quarter, or an MHR nobody has revised since the machine was bought.

  • Finished weight used as gross weight. Raw material is bought on input weight, not on what the part weighs at dispatch.

  • Selling all waste as scrap. Scale loss, melting loss and chips that cannot be recovered are not scrap credit.

  • Supplier cycle times copied into your own model. Then the model can only ever agree with the quote.

  • Percentages without bases. "12% overhead" means nothing until you know what it is 12% of.

  • Missing heads. Packing, freight, ICC, tool amortisation and surface treatment are the lines most often left out, on both sides.

  • No record of assumptions. Without the RM base date, volume and index written down, the next amendment turns into an argument.

  • Treating should-cost as a target to impose. It is a benchmark for a conversation. A supplier who cannot make money at your number will find it back in quality or delivery.

How do manufacturers run should-cost analysis at scale?

One should-cost in a spreadsheet is manageable. Hundreds of parts, several plants and quarterly steel revisions are not. Teams that do this well keep the rates and norms as shared masters, build every model from them, and get suppliers to quote in the same format so the comparison is automatic.

In Cost It Right, that looks like this:

  • ZBC, VBC and client-based costing on the same part, with cost version history and side-by-side comparison.

  • Standard costing formats for 24+ technologies, including machining, forging, casting, sheet metal, plastic and rubber, so every plant and supplier fills the same lines.

  • A vendor RFQ portal (how buyer-side RFQ automation works) where every supplier submits the same breakup through a secure link, and an RFQ comparison report that shows each line, total cost of ownership and the ZBC benchmark together.

  • Raw material indexation, linking RM rates to external commodity indices such as LME for aluminium and copper, with the impact simulated down to finished-good level before an amendment is approved.

  • Approval workflows and a full audit trail, so every agreed price shows who approved it, when and on what basis.

One multi-plant auto component group cut its end-to-end costing time by 60% after moving costing into the system.

See how leading OEMs and Tier 1 suppliers run should-cost analysis in Cost It Right. Book a demo.

FAQs

What is should-cost analysis in procurement?

It is a bottom-up estimate of what a part should cost if a capable supplier makes it efficiently, built from material, process, overheads and a fair profit. Buyers compare it with the supplier's quote, line by line, to find inflated inputs and negotiate on facts.

How is should-cost calculated?

Should-cost = net raw material + bought-out parts + conversion cost + overheads + rejection + inventory carrying cost + profit + packing + freight + tool amortisation. Each line is calculated from its own drivers, such as gross weight, RM rate, machine hour rate and cycle time.

What is a should-cost model example?

For a machined EN8 steel flange, a should-cost model built from a 2.1 kg bar input, realistic cycle times and standard allowances comes to ₹304.68. Against a supplier quote of ₹393.00, the ₹88.32 gap is mostly explained by raw material weight and cycle time.

What is should costing methodology?

The methodology is to build the cost from the drawing and the process route, not from past prices: collect part data, fix the route, calculate gross weight and net material, cost each operation, add allowances on agreed bases, add packing, freight and tooling, compare with the supplier's breakup, and record every assumption.

Is should-cost the same as zero-based costing?

They are closely related and often used interchangeably. Both build the cost from first principles. Zero-based costing stresses ignoring historical prices; should-cost stresses the efficient-supplier benchmark used in negotiation.

What is the difference between should-cost and total cost of ownership?

Should-cost estimates what the part should cost to make. Total cost of ownership adds what it costs you to own it: freight, duties, payment terms, inventory, quality and supply risk. Use should-cost to judge a price, and TCO to choose between suppliers.

How often should a should-cost model be updated?

Update it whenever a major input changes, at minimum each time raw material prices are revised under your contract (often quarterly), and whenever volume, process or design changes.