Should-cost analysis is a way of working out, from a part's design, what the part ought to cost to make. It also suggests which supplier should make it. A buyer does this before ordering, so that a supplier's price can be judged against a fair price.
A buyer who only has a supplier's quote does not know whether it is fair. Should-cost analysis builds the price from its parts: the material, the machine time, the finishing, the supplier's overheads and a normal profit. The software reads these from the part's CAD model and from stored rates for materials and machines in each region. A machine hour in one industrial town costs a different amount from a machine hour in another, and the software keeps both rates.
Here is a typical case. A purchase executive at a tractor-parts company in Faridabad gets a quote of ₹290 for a steel mounting plate. The software reads the model. The plate weighs 0.8 kg, and steel costs about ₹70 a kg, so the material is ₹56. Cutting takes 6 minutes on a machine that costs ₹900 an hour, which is ₹90. Painting adds ₹20. The overheads are about 15% and the profit about 10%. The total is about ₹210. The executive now has a reason to ask the supplier why the quote is ₹80 higher, or to ask other suppliers.
Should-cost analysis does not prove that the supplier is overcharging. The supplier may have a small order, an old machine or a high cost of transport that the software does not know about. The number gives the buyer a question to ask, not an accusation.
The design in PLM often decides most of the cost. So engineers also use should-cost analysis during design, to see what a thicker wall or a tighter tolerance would add to the price. The results go to procure-to-pay software, where the purchase is made, and can be compared with the instant quotes of an on-demand manufacturing platform.