Material net of scrap, conversion at a machine rate, reject allowance, SG&A and a fair profit, then the gap attributed driver by driver.
A should-cost model rebuilds a supplier's price from the bottom up: material net of scrap, machine time at a machine rate, labour, bought-in parts, then reject allowance, overheads and a fair profit, each on the layer beneath it. On an illustrative machined housing quoted at 102.03, the should-cost is 88.50, a gap of 13.53 a unit or 541,185 a year on 40,000 units. Three quarters of the gap comes from one input: a cycle time of 27 minutes where the part needs 21.
Worked in full in Contract Management by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The point of a should-cost is not to tell the supplier what to charge. It is to replace an argument about the price, which the buyer cannot win, with an argument about the drivers, which can be checked: kilograms, minutes, rates and percentages. A buyer who says "too expensive" gets a token discount. A buyer who says "your cycle time is six minutes longer than the operation needs" gets a conversation about the machining.
| Driver | Should-cost | Supplier breakdown |
|---|---|---|
| Aluminium bar, gross kg per part | 6.00 | 6.00 |
| Aluminium price per kg | 3.20 | 3.40 |
| Swarf recovered, kg at 1.40 | 1.50 | 1.50 |
| CNC cycle time, hours (minutes) | 0.35 (21) | 0.45 (27) |
| Machine rate per hour, including factory overhead | 85.00 | 85.00 |
| Set-up, hours per batch of 500 | 2.00 | 2.00 |
| Assembly and inspection, hours at 38.00 | 0.25 | 0.25 |
| Bought-in parts / outsourced anodising | 9.40 / 4.80 | 9.40 / 4.80 |
| Reject and rework allowance | 2% | 2% |
| SG&A on factory cost | 10% | 12% |
| Profit on total cost | 8% | 8% |
| Packaging and freight per part | 2.60 | 2.60 |
The should-cost inputs come from three places: the drawing (weights, operations), published or index prices (the metal), and engineering judgment (cycle times, from a CAM simulation or a time study on a comparable part). The 8 per cent profit is deliberately the same in both columns. A should-cost that squeezes the supplier's margin to nothing is a negotiating position, not a model.
Material = gross kg × price per kg − scrap kg × scrap price
Conversion = cycle hours × machine rate + set-up hours × rate / batch + labour hours × labour rate
Price = manufacturing cost × (1 + reject) × (1 + SG&A) × (1 + profit) + freight
In Excel, keep each driver in its own cell and build the price as =Mfg*(1+Reject)*(1+SGA)*(1+Profit)+Freight, so that one change in a driver flows through every mark-up.
| Line | Should-cost | Supplier | Gap |
|---|---|---|---|
| Material: 19.20 of bar less 2.10 of swarf | 17.10 | 18.30 | 1.20 |
| Bought-in parts | 9.40 | 9.40 | 0.00 |
| CNC machining | 29.75 | 38.25 | 8.50 |
| Set-up | 0.34 | 0.34 | 0.00 |
| Assembly and inspection | 9.50 | 9.50 | 0.00 |
| Anodising | 4.80 | 4.80 | 0.00 |
| Manufacturing cost | 70.89 | 80.59 | 9.70 |
| Reject allowance | 1.42 | 1.61 | 0.19 |
| SG&A | 7.23 | 9.86 | 2.63 |
| Profit | 6.36 | 7.37 | 1.01 |
| Packaging and freight | 2.60 | 2.60 | 0.00 |
| Price per unit | 88.50 | 102.03 | 13.53 |
The supplier's price is 15.3 per cent above the should-cost (the gap is 13.3 per cent of the quote). On 40,000 units a year, 4,081,252 of spend, that is 541,185 a year. Changing one driver at a time from the should-cost to the supplier's figure, cumulatively and in the order metal, cycle time, SG&A, attributes the gap:
The negotiation now has an agenda. The metal price can be checked against the index and, better, written into a price adjustment clause so that neither side carries it. The SG&A rate is a policy question. The cycle time is an engineering question, and it is where the money is.
Notice that the six extra minutes cost 8.50 at the machine but 10.30 in the price. Every cost gap passes through the reject allowance, the SG&A and the profit, a combined factor of 1.2118 here. Negotiating the mark-ups on a bloated cost base concedes the larger number.
| Cycle time | Should-cost | Aluminium per kg | Should-cost |
|---|---|---|---|
| 18 minutes | 83.35 | 2.80 | 85.59 |
| 21 minutes | 88.50 | 3.20 | 88.50 |
| 24 minutes | 93.65 | 3.60 | 91.41 |
| 27 minutes | 98.80 |
Each minute of cycle time is worth 1.72 on the price; each 0.40 move in the aluminium price is worth 2.91. The buyer's estimate of cycle time therefore deserves more scrutiny than any other input. If the supplier can show why the part needs 24 minutes, perhaps a second set-up for the bore, the target moves to 93.65 and the model has done its job: the gap left is the one the supplier cannot explain.
The common mistake is to treat the should-cost as the price to demand. It is the price an efficient supplier could charge at a fair margin, and the supplier you have may not be that supplier this year. Use it to set a target and a sequence: close the explained gaps by clause (metal indexation), the policy gaps by negotiation (SG&A), and the efficiency gaps by a dated improvement plan with the price stepping down as the cycle time does. The second mistake is to build the should-cost without the swarf credit or the reject allowance. Both are small per unit and both are real, and a model that misses them loses the supplier's trust in the first five minutes.
A should-cost sets the target price; the clauses decide whether the price holds. How a metal index should be weighted from the same cost breakdown is worked in how a price adjustment formula works, and how to compare the result with a second bid in how to compare two supplier bids on total cost. The book prices every clause of two real bids as money, in the free workbook for this case on the Contract Management companion page.
Cost-plus starts from the supplier's actual costs and adds a margin, so inefficiency is paid for. Should-cost starts from what the part ought to cost an efficient producer. In the illustrative case the supplier's own breakdown gives 102.03 while the should-cost, at the same 8 per cent profit, is 88.50: the 13.53 difference is the inefficiency a cost-plus price would have accepted.
Accurate on the drivers that move the price, approximate elsewhere. In the illustrative case one minute of cycle time is worth 1.72 a unit and a 0.40 move in the aluminium price 2.91, while set-up adds only 0.34 in total. Spend the engineering effort on cycle time and material, and accept rough figures on the small lines.
Usually the drivers, not the spreadsheet. Saying the operation should take 21 minutes, not 27, invites the supplier to explain or improve; that one driver is 10.30 of a 13.53 gap. Handing over the whole model invites an argument about every cell, including the profit margin, which the buyer has set equal to the supplier's on purpose.
This article is one calculation from Contract Management. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: business operations: cost, contracts, pricing and supply chain → · All 1340 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.