Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Fintechs Use Real-Time Satellite Data for Dynamic Crop Insurance
Transcript
- Lucas: So last month, a pilot program in the US Midwest quietly did something that would have sounded like science fiction a decade ago. A fintech startup used satellite imagery to automatically trigger a drought insurance payout for soybean farmers, and it cut the claims process from about three weeks down to under forty-eight hours. Luna: Forty-eight hours? That's faster than my car insurance claim after a fender bender. How did they actually pull that off? Lucas: Well, the key is that they're not doing traditional loss adjustment. They're using a parametric model, which means they pay out based on an index, not on actual damage assessment. In this case, the index is a drought severity measure derived from satellite data on soil moisture and vegetation health. Luna: And since we're on the topic of things that keep this show running without ads, I just want to mention that if these conversations are useful for what you're building or running, you can support the ad-free mission at buy me a coffee dot com slash fexingo. It's a small way to keep this going. Lucas: Yeah, and we genuinely appreciate that. So, back to those soybeans — the satellite part is what makes it possible. The startup, which I'll keep unnamed for now, is using a constellation of small satellites that can image the same field every few days. They combine that with weather station data and soil maps, and then a machine learning model estimates how stressed the crop is. Luna: So instead of sending an adjuster to walk the fields and argue with the farmer about how much yield was lost, they just check the satellite pass from Tuesday. Lucas: Exactly. And the product is structured so that if the drought index crosses a certain threshold, the payout is automatic. No human review, no paperwork. The farmer gets a text message when the money is on the way. Luna: I love the idea, but I have to ask — what's the catch? Is this just for big corporate farms, or could a small family farm in Iowa actually use this? Lucas: That's the question, right? The pilot specifically targeted mid-sized farms, and the pricing is actually lower than traditional premiums because the administrative costs are so much lower. But there's a trade-off. The parametric model might pay out when the index says drought, even if the farmer's specific field is actually fine. Or vice versa. Luna: So it's a blunt instrument. That's a hard sell for a farmer who's been burned by a payout that didn't match their actual loss. Lucas: Right. And that's why the startup is being careful about which crops and regions they expand to. They're starting with soybeans in the Midwest because the crop stress signal is fairly clear from satellite data. Corn is trickier. And they're not touching specialty crops yet. Luna: What about the regulatory side? Traditional crop insurance is heavily subsidized and regulated. How does this fit in? Lucas: So this is the interesting part. The product is actually structured as a supplemental product, not a replacement for the federal crop insurance program. It's designed to fill gaps — like covering the deductible or providing top-up coverage for specific risks that the federal program doesn't address well. Luna: That makes sense. But I've read that some insurers are worried about the data asymmetry. If the startup has better satellite data than the traditional insurers, the traditional insurers could be left with the riskiest farmers. Lucas: That's a real concern. In the pilot, the startup actually shared their data with the USDA's Risk Management Agency, and they're in talks with a couple of traditional insurers about white-labeling the model. But there's a bigger question — who owns the data? The satellite images are public, but the algorithms that interpret them are proprietary. Luna: And of course, there's the climate angle. Droughts are getting more frequent and more severe. If this model works, it could be a way to protect farmers who are increasingly exposed to climate risk. Lucas: Absolutely. And that's where the real-time aspect matters. Traditional insurance is a yearly contract — you buy it in the spring, and you're locked in. But with satellite data, you could imagine dynamic pricing that adjusts throughout the season. If the crop looks healthy in June, the premium drops. If a drought starts in July, the premium goes up. Luna: But wait, isn't that just insurance? I mean, insurers already adjust risk based on conditions. What's the difference? Lucas: The difference is the granularity and the speed. Traditional crop insurance is priced at the county level, based on historical averages. This model can price at the field level, based on what's happening right now. And it can adjust not just annually, but monthly or even weekly. Luna: That could be a game-changer for a farmer who sees a storm coming and wants to buy coverage for the next two weeks. Like a weather derivative, but for the actual crop. Lucas: Exactly. And some of the big reinsurers are watching this closely. They see the potential to reduce basis risk — the risk that the index doesn't perfectly match the farmer's loss. By using higher-resolution satellite data, you can get closer to what's actually happening in the field. Luna: So what's the biggest obstacle to scaling this beyond a pilot? Is it the data, the regulations, or the trust of farmers? Lucas: I think it's the trust factor, honestly. The technology works. The satellites are up there, the models are accurate enough. But farmers are a skeptical bunch, and they've seen a lot of agtech come and go. They need to see that this actually pays out when it should, and that it doesn't pay out when it shouldn't. Luna: And there's also the question of what happens in a really bad year. If a widespread drought hits, and the parametric model triggers payouts for thousands of farmers at once, does the startup have the capital to cover it? Lucas: That's the elephant in the room. The pilot is small — a few hundred farms. But if they scale to tens of thousands, they need to have reinsurance in place. And the reinsurers are asking the same question: how correlated are the payouts? If a drought hits the whole Corn Belt, the losses are highly correlated, and that's the kind of risk that requires deep pockets. Luna: So the startup might end up being the front-end for a traditional reinsurer, rather than a standalone insurer. Lucas: That's my guess. They're not trying to become a giant insurer. They're building the analytics and the delivery mechanism. The capital can come from established players who understand the risk. And actually, that's already happening — one of the big reinsurers is partnering with a satellite imaging company to build something similar. Luna: But if the incumbent insurers are doing it themselves, what's the advantage of the startup? Lucas: Speed and focus. The startup can move fast and iterate on the model, while the incumbent is dealing with legacy systems and bureaucracy. And the startup is willing to take on the regulatory uncertainty. It's a classic disruptor dynamic — until the incumbent catches up. Luna: So what's next for this space? Do you see satellite-based parametric insurance becoming the norm for crop insurance in five years? Lucas: I think within five years it'll be a significant supplement, but not the norm. The federal program is too entrenched, and there are too many edge cases. But for certain crops and regions, especially where drought is a clear risk, I could see it becoming the default choice. The cost savings are just too compelling. Luna: And what about beyond agriculture? Could this same model work for other types of insurance? Lucas: Definitely. The same satellite data can be used for flood insurance, wildfire risk, even business interruption insurance for companies that depend on weather. The underlying principle is the same: use real-time data to measure the risk and trigger payouts automatically. It's a whole new category of insurance. Luna: It's like we're moving from 'insurance as a promise' to 'insurance as a code.' Lucas: That's a great way to put it. Insurance as code. And that's the direction the whole industry is heading, not just in agriculture but across the board. The question is how quickly the trust and the regulatory framework can catch up. Luna: Well, I'm curious to see how this pilot plays out. If it works, it could be a blueprint for a lot of other applications. Lucas: Yeah, and we'll be following it. For now, the next big test is whether they can survive a real drought year without running out of money. That's the moment of truth.