Most cost reviews begin with a number the supplier chose. A quotation arrives: ₹740 per component. The obvious moves are to compare it with other quotations, or to ask for less. Both are useful. Neither explains why the number is ₹740.
Benchmarking tells you whether a price is unusual. Negotiation tests how much of it the supplier is willing to give up. But a supplier who has quoted on a conservative scrap assumption, an oversized press or a rejection allowance from a bad year can concede a few per cent and still leave the real gap untouched. The concession may not even last, because nobody on either side has agreed what the part should cost.
Start from the part
A should-cost model starts where the supplier’s own estimator starts: with the drawing. For a pressed-steel bracket, that means the blank weight and material grade, the nesting on the coil and what happens to the scrap. It means the forming stages, the press tonnage, the cycle, the machine rate and how much of the press’s time the part really uses. Then the tooling architecture and how it is amortised, the secondary operations, the labour content, the rejection assumption, overhead, logistics and margin.
Each of those is a question with an answer that can be checked. Together they turn a single number into a set of explained differences.
The question is never only whether ₹740 is high. It is what would have to be true for ₹740 to be right.
Why it changes the conversation
A negotiation built on explained differences is a different conversation. The supplier is no longer defending a price; they are discussing assumptions, some of which will prove right. That is healthy. A model that finds an under-recovered overhead line is as valuable as one that finds a padded material line, because a supplier losing money on a part is a supply risk, not a saving.
It also shows which differences belong to whom. Some cost is priced in, and procurement can address it. Some is made in, and the supplier’s process has to change. Some is designed in, which means the conversation has to include engineering.
What makes a model credible
Should-cost models fail when they are built by people who have never run the process. A model that assumes a forging with no flash, a casting with perfect yield or a machine that never changes over will be dismissed in the first meeting, and rightly. Credibility comes from specialists who have stood next to the press, the furnace and the machining cell, and who build the model the way the supplier would.
That is the standard we hold ourselves to. The quote is where our analysis starts, not where it ends.