Latest / The Buyer's Side with Fexingo: Procurement, Sourcing, and B2B Purchasing Decisions / How One Buyer Used Should-Cost Modeling to Negotiate a 12 Percent Price Cut
Transcript
- Lucas: So there is a moment in almost every procurement career where you look at a supplier's quote and you just know — you can feel it — that the price is too high. But you can't prove it. You don't have a competing bid, the supplier knows they're your only option, and you're stuck. Luna: Right, you're negotiating blind. And that's where should-cost modeling comes in. Lucas: Exactly. And today I want to talk about a buyer who used that exact technique to cut prices by 12 percent on a category where he had zero competitive leverage. This was at a mid-sized medical device manufacturer — let's call the buyer Mark. He was sourcing precision-machined components, things like surgical instrument housings and implantable device casings. Luna: And those aren't commodities — you can't just throw a spec out to five suppliers and pick the lowest bidder. Lucas: No, exactly. The parts require tight tolerances, specialized equipment, and the supplier had been making them for years. Mark had tried to get alternative quotes but the switching costs were high — retooling, validation, FDA paperwork. So his incumbent supplier knew they had pricing power. Luna: And this is where if these conversations have moved your work forward in some small way, listener support is what keeps the show ad-free and independent. If that's meaningful to you, you can support us at buy me a coffee dot com slash fexingo. But back to Mark. Lucas: Thanks for that. So Mark decided he wasn't going to accept the blind spot anymore. He built a should-cost model from the ground up. He started with the raw material — 316L stainless steel — and calculated the exact weight of each part, then added scrap rates and current metal prices. Luna: Did he get the metal prices from public indices or from his supplier? Lucas: He used London Metal Exchange prices for the base, then applied a premium for the specific alloy. That alone showed him the material cost was about 30 percent lower than what the supplier had embedded in their unit price. Next he looked at machine time. He estimated cycle times based on the part geometry and the type of CNC machine required — a five-axis mill, in this case. He found the average machine-hour rate for that equipment in the supplier's region by talking to equipment dealers and checking industry surveys. Luna: So he was building a bottom-up cost picture. That requires detailed specifications, which not every buyer gets from engineering. Lucas: That's a key point. Mark had to go back to his internal engineering team and say, 'I need the exact tolerances, the surface finish requirements, the inspection specs.' And engineering was initially reluctant — they thought procurement was going to mess with their design. But he convinced them that understanding the cost drivers would actually protect the design from being value-engineered later. Luna: So he turned them into allies. Smart. Lucas: Right. With the specs in hand, he then modeled labor costs. He knew the supplier's location, so he used Bureau of Labor Statistics data for machinist wages in that area, then added a factor for benefits and burden. And he found something interesting: the supplier's implied labor cost in their quote was about 15 percent above the regional average. That suggested either inefficiency or padding. Luna: Or maybe the supplier was using more skilled labor than the average? Could there be a legitimate reason? Lucas: Good question. Mark actually considered that and decided to give the supplier the benefit of the doubt on skill premium — he adjusted his model upward by 5 percent. But even then, the gap remained. Then the final piece was overhead. He allocated a percentage for SG&A, quality control, shipping, and profit margin based on what's typical for precision machine shops — around 25 to 30 percent overhead plus 10 percent profit. His model produced a total cost that was 15 percent below the supplier's quote. Luna: Fifteen percent. So he had a target. But how did he present it without destroying the relationship? Lucas: That's where Mark was really skilled. He didn't walk in and say, 'Your quote is wrong.' He scheduled a joint cost-modeling session. He shared his assumptions openly and asked the supplier to critique them. He said, 'Here's what I think this part should cost to make. Tell me where I'm off.' And the supplier initially pushed back on a few points — they argued their machine-hour rate was higher because they used newer equipment, and they had a point. So Mark adjusted his model upward by 3 percent on that line item. Luna: So the negotiation became collaborative. They were discussing facts, not positions. Lucas: Exactly. After two sessions, they landed on a should-cost that was 12 percent below the original quote. The supplier agreed to a phased reduction — 6 percent immediately, another 6 percent over the next year as they realized process improvements. And Mark committed to a longer contract term to give them stability. Luna: So the buyer didn't just take the full 15 percent. He compromised to keep the relationship healthy. Lucas: Right. And the supplier actually appreciated the process. They told Mark that no other customer had ever taken the time to understand their cost structure that deeply. It strengthened the partnership. And Mark got a 12 percent price cut without switching suppliers or burning bridges. Luna: I want to go back to something you mentioned earlier — the internal alignment piece. How did Mark get engineering to share those spec details? Lucas: He framed it differently. Instead of saying 'I need this for a cost model,' he said 'I want to protect our design from cost-cutting later by showing that it's already optimized.' That resonated. Engineers hate when someone later changes their design to save a nickel. So they helped him understand which tolerances were critical and which had some flexibility. That actually made his model more accurate. Luna: So the model became a tool for cross-functional collaboration, not just a negotiation tactic. Lucas: Yes. And that's really the lesson from this case. Should-cost modeling works best when it's transparent and collaborative — internally with engineering, externally with the supplier. If you try to use it as a weapon, suppliers will resist. But if you use it as a tool for joint understanding, you can often find savings that benefit both sides. Luna: What about categories where you don't have as much visibility into cost drivers? Like services or software? Lucas: That's harder. Should-cost modeling works best for manufactured goods where you can break down inputs — materials, labor, machine time, overhead. For services, you'd need to model labor hours and skill levels, which can be more variable. But the same principle applies: build a bottom-up estimate based on publicly available data and reasonable assumptions. Even an imperfect model is better than no model. Luna: And it forces you to understand your own requirements better. Mark had to clarify specs he'd never looked at before. Lucas: That's a hidden benefit. The process of building a should-cost model often reveals that you don't actually need all the tolerances you've been paying for. In Mark's case, they found one dimension that was spec'd to plus or minus one thousandth of an inch, but after talking to engineering, it turned out the functional requirement was only three thousandths. Loosening that tolerance alone reduced machining time by 8 percent. Luna: So the model paid for itself even before the negotiation started. Lucas: Exactly. And that's something any buyer can try next week. Pick a high-spend, single-source part. Spend a week gathering data — raw material prices, labor rates, machine hour costs. Build a basic spreadsheet. You might not get to 12 percent, but you'll almost certainly find something. At minimum, you'll walk into your next supplier meeting with facts instead of hunches. Luna: And if the supplier pushes back on your assumptions, you can learn something about their process. That's valuable too. Lucas: Right. Either way, you're better off. That's the power of should-cost modeling.