Latest / The Buyer's Side with Fexingo: Procurement, Sourcing, and B2B Purchasing Decisions / How One Buyer Used Catastrophe Modeling to Reshore a Critical Supplier
Transcript
- Lucas: In 2011, a massive monsoon season in Thailand flooded hundreds of industrial estates, and one of the worst-hit was a hard disk drive factory. That single event caused a global shortage that cost the electronics industry roughly $30 billion. But what happens when a flood like that takes out a supplier that's not making hard drives, but something far more custom — like a precision casting for automotive brake systems, and there is no second source? Luna: That's a nightmare scenario for any buyer. Single source, long lead times, high switching costs. I'm guessing this is the case you're going to walk us through? Lucas: Exactly. So the buyer — let's call him Mark — was a procurement manager at a tier-one automotive supplier based in the Midwest. His company made brake calipers for several major automakers. One critical component, a cast iron housing, came from a single foundry in central Thailand. That foundry had been reliable for over a decade. Great quality, competitive pricing, strong relationship. Luna: And then the 2011 floods hit? Lucas: Not exactly. The floods hit a different region, but they came close enough to the foundry's supply chain — they relied on Thai sand for molds, and local trucking routes — that production slowed for six weeks. Mark's plant nearly shut down. They had to airfreight emergency units from a backup foundry in China at five times the normal cost. The CFO was furious. Mark was told: fix this. So he did, but not the way most buyers would. Luna: What was his approach? Lucas: Instead of just finding a second source in a nearby low-cost country, Mark proposed reshoring the backup supply to Mexico. The CFO immediately pushed back because the unit cost from a Mexican foundry would be 18 percent higher than the Thai incumbent. But Mark had been reading about catastrophe modeling — the kind insurers use to price hurricane risk — and he realized he could apply similar methodology to supply chain disruptions. Luna: So he built a quantitative risk model, not just a cost comparison. Lucas: Right. He gathered three types of data. First, historical climate data for the Thai region — flood plain maps, monsoon intensity trends, river basin levels. Second, geopolitical risk scores from a third-party provider — things like labor strikes, port congestion probabilities, and regulatory instability. Third, he modeled the cost consequences of various disruption scenarios: a two-week outage, a six-week outage like they had, and a worst-case three-month shutdown. Luna: And he plugged in the cost of expedited freight, lost sales penalties, maybe even customer goodwill damages? Lucas: Exactly. For each scenario, he calculated the total cost of disruption: premium freight, overtime at his own plant, potential contractual penalties for late delivery to the automaker, and a percentage of lost future business if they gained a reputation for unreliable supply. Then he assigned probabilities to each scenario based on the climate and political data. Luna: So he had an expected annual loss from the Thai single source. Lucas: Yes. The model showed that over a three-year horizon, the expected disruption cost from staying single-sourced in Thailand was roughly $1.8 million. Adding a Mexican second source would cost an additional $2.3 million in upfront tooling, qualification, and higher per-unit cost. But the model also showed that the Mexican site had dramatically lower climate and political risk — less than a 2 percent chance of any major disruption per year. Luna: So the expected disruption cost with Mexico was near zero. The net present value favored the reshoring. Lucas: Mark calculated that over three years, the total cost of ownership for the dual-source strategy — including the higher unit cost — was actually $4.1 million less than sticking with single-source Thailand, once you factored in the expected disruptions. The CFO approved the investment. They qualified a foundry in Monterrey within 14 months. Luna: And how did it play out in practice? Lucas: Three years later, Mark's company had used the Mexican source for roughly 40 percent of volume. There was one minor port strike in Thailand that caused a three-week delay, but because they had the Mexican capacity, they simply shifted production. No premium freight, no plant shutdown. The automakers never even knew there was an issue. Luna: That's a great outcome. But what about the smaller buyer who doesn't have access to catastrophe modeling software or a data team? Lucas: That's a fair point. Mark didn't use expensive software. He built the model in Excel using publicly available flood risk data from the Thai government and the World Bank's political stability index. He even used historical shipping delays from a freight forwarder's blog. The key wasn't the tool — it was the mindset of quantifying risk instead of just gut-feeling it. Luna: So any buyer could do a simplified version: list your top three risks per supplier, assign a rough probability, estimate the cost impact, and compare to the premium of a second source. Lucas: Exactly. And you don't need to build a perfect model. Even a rough estimate can be powerful in a conversation with finance. The CFO at Mark's company later told him that the model was the only reason he approved the investment — because it translated a vague fear into a dollar figure. Luna: And that's really the takeaway: procurement often talks about risk in qualitative terms, but finance speaks in numbers. You have to meet them there. Lucas: Right. And speaking of numbers — and this is something we don't talk about much on the show — but we deliberately keep this podcast ad-free. No sponsors, no commercials. If you find value in these conversations and want to support that approach, you can find us at buy me a coffee dot com slash fexingo. No pressure, just an option. Luna: Yeah, it's a small way to keep the show independent. And we appreciate anyone who's ever contributed. Lucas: So back to Mark's story — one thing I found really clever was how he presented the model to the CFO. He didn't lead with the climate data. He led with the $4.1 million savings figure, and then walked backward through the assumptions. Luna: That's classic executive communication. Start with the bottom line, then defend it. Lucas: Yeah. And he also built in a sensitivity analysis — showing that even if disruption probabilities were half of what he estimated, the reshoring still had a positive net present value. That gave the CFO confidence that the decision was robust. Luna: So the lesson here isn't just about catastrophe modeling. It's about procurement using financial logic to make a strategic case that goes beyond unit price. Lucas: Exactly. Mark didn't just buy components — he bought resilience. And he proved that resilience has a measurable return. That's the kind of thinking that elevates procurement from a cost center to a value driver.