Latest / The Buyer's Side with Fexingo: Procurement, Sourcing, and B2B Purchasing Decisions / How Should-Cost Modeling Cut Procurement Spend 15 Percent
Transcript
- Lucas: Last year, a purchasing manager at a Midwest electronics company was staring at a quote for a circuit board assembly. The supplier had quoted $47.50 per unit. Using a technique called should-cost modeling, she got that number down to $40.38 – a 15 percent reduction. Luna: So she basically rebuilt the supplier's cost structure from scratch? Lucas: Exactly. Should-cost modeling is exactly that – you estimate what a product should cost to produce, based on raw materials, labor, overhead, and a reasonable profit margin. Then you use that estimate as your negotiation anchor, not the supplier's list price. Luna: I've heard of it, but it sounds labor-intensive. How did she actually do it for a circuit board? Lucas: She started with the bill of materials. Circuit boards have a standard set of components – resistors, capacitors, chips – all with market prices. She pulled current pricing from distributors like Digi-Key and Mouser. Then she added the cost of the bare board itself, which is a commodity. Then assembly labor: the supplier's factory in Thailand runs at about $0.12 per solder joint. She knew there were 240 joints, so that's $28.80 in labor. Add overhead at 15 percent, freight, and a 10 percent profit margin – that gave her a should-cost of around $38.50. The supplier quoted $47.50. Luna: So the supplier was marking up by almost 25 percent above your estimate. Did she just walk in with that number? Lucas: Not quite. She invited the supplier's sales engineer and their production manager to a joint cost-modeling session. She presented her should-cost analysis transparently – here's my assumption on chip cost, here's your labor rate, here's your overhead percentage. She asked them to correct any errors. Luna: That's risky – showing your hand like that. Lucas: It can be, but in this case it paid off. The production manager confirmed the labor rate but said they had a newer wave soldering machine that reduced labor by 8 percent. So they adjusted the model downward by a couple of dollars. In the end they landed at $40.38, which was still 15 percent below the original quote. The supplier agreed because they could see the logic and still made their margin. Luna: So transparency actually helped the negotiation. It builds trust instead of adversarial bargaining. Lucas: Right. And the key is that should-cost modeling works best when you have access to public or industry-standard cost data. For electronics, it's relatively easy because component prices are transparent. In other industries it can be harder – complex machined parts, specialty chemicals – but the same principle applies. Luna: What about the long-term impact? Did she use the model for other parts? Lucas: She did. After that success, her team built should-cost models for their top 20 spend categories. Over the next year, they averaged 12 percent savings on those categories. The key is that once you build the model, you can reuse it for similar parts. She also used it to benchmark multiple suppliers – if one supplier could hit $40, why couldn't another? Luna: That makes sense. It gives you a fact-based floor for negotiations. But I imagine there's a learning curve. How did she get started? Lucas: She attended a two-day workshop on cost modeling, then practiced on a low-risk commodity – office supplies. She modeled the cost of a ream of paper: pulp, labor, energy, shipping. It helped her build confidence before tackling the circuit board. Luna: So start simple. That's a good takeaway. Lucas: Exactly. And there are software tools now that automate some of the data gathering. But the core skill is understanding the manufacturing process. You don't need to be an engineer, but you need to ask the right questions: What's the raw material? How is it made? What's the cycle time? How much waste? Luna: Those are exactly the questions a supplier might not want to answer. How do you get them to cooperate? Lucas: Frame it as a collaborative exercise. You're not trying to squeeze their margin; you're trying to remove unnecessary cost from the supply chain. Often the supplier learns something too – maybe their overhead allocation is off, or they have inefficiencies they didn't see. In that case, the supplier appreciated the fresh perspective. Luna: So it's a win-win if done right. Lucas: Yeah. And that brings up an important point: should-cost modeling isn't a one-time tactic. The buyer in our story updates her models quarterly because commodity prices change. Copper, for example, fluctuates and affects the board cost. She adjusts the model accordingly and renegotiates when justified. Luna: That's smart – it turns procurement into an ongoing cost management function rather than a firefighting role. Lucas: And that's the kind of concrete insight we try to bring every episode. If these conversations have sharpened your thinking in procurement, one way to support that is through our listener funding. Luna: Yeah, we don't run ads, so it's entirely fueled by people who value the ad-free approach. The link is buy me a coffee dot com slash fexingo. Lucas: Anyway, back to should-cost modeling – one final tip from that buyer: always include a sensitivity analysis. When you present the model, show how the cost changes if copper rises 10 percent or labor increases 5 percent. It shows you've thought through the risks and makes the supplier take the model more seriously. Luna: That's a great practical detail. So next time you get a quote that seems high, ask yourself: what should this really cost? Lucas: Exactly. We'll put a link to a free should-cost template in the show notes. Thanks for listening. Luna: Until next time, keep buying smarter.