Latest / Business Strategy Talks with Fexingo: Competitive Advantage, Market Positioning, and Growth Planning / How John Deere Built a Moat in Precision Agriculture
Transcript
- Lucas: You know, when most people think about a moat, they picture something like ASML's lithography machines or Intel's fabs — hard tech that takes billions and decades to replicate. Luna: Right. And today we're talking about green tractors, which feels almost quaint by comparison. Lucas: But John Deere's moat today is about something far more durable than a diesel engine. It's about data — and specifically, the network effect that happens when every tractor in the field is collecting information and sending it back to the same platform. Luna: So walk me through it. How did an 183-year-old plow company become a data company? Lucas: The shift really accelerated around 2017, when Deere acquired Blue River Technology for about $305 million. Blue River had developed 'see and spray' — computer vision that lets a tractor identify individual weeds and apply herbicide only where needed. That's not a hardware play. That's a software and sensors play. Luna: And that's where the moat starts — because the algorithm gets better the more fields it sees. Lucas: Exactly. Deere now has over 500,000 connected machines globally, and they're generating terabytes of data every day — soil conditions, planting density, yield maps, weather patterns. Every new piece of data improves the model for every other farmer using the system. That's a classic data network effect. Luna: So the competitive advantage isn't that Deere makes a better tractor — it's that the tractor becomes smarter over time, and the competitors' tractors don't. Lucas: That's the core insight. CNH Industrial's New Holland tractors are fine machines. AGCO's Fendt brand is excellent. But when a farmer buys a Deere, they're not just buying iron — they're buying into an ecosystem that's constantly learning. Luna: And Deere has locked that in with a subscription model. The 'Operations Center' is the digital hub — you pay a recurring fee to access your field data, get analytics, and run autonomous features. Lucas: Right. As of this year, 2026, Deere's recurring revenue from software and services is over 25 percent of total revenue. For a company that sold about $55 billion in equipment last year, that's north of $13 billion in high-margin, recurring streams. Luna: And that's sticky. If a farmer has five years of field data in Deere's cloud, switching to a different tractor brand means losing that history and the algorithms that optimize their planting. Lucas: It's a textbook switching-cost moat. But there's a second layer: Deere has also made it difficult for third-party software to access its data. If you want to use a competing analytics platform, you have to go through Deere's API, and they control the terms. Luna: That's where the controversy comes in. Farmers and regulators have been pushing back — right-to-repair laws, data portability requirements. In 2023, Deere signed a memorandum of understanding with the American Farm Bureau to allow farmers to repair their own equipment, but there's still friction. Lucas: And that's the real threat to the moat. If regulators force Deere to open up its data the way Europe forced banks to open up payment data under PSD2, then the network effect could weaken. Farmers could take their data to a startup that offers better analytics. Luna: But so far, that hasn't happened at scale. And Deere keeps pushing the frontier — they're testing fully autonomous tractors, and they've got a partnership with SpaceX for satellite-based connectivity in rural areas. Lucas: Which widens the moat further. Because now you need not just a tractor and software, but also satellite bandwidth and machine-learning models trained on millions of acres. That's a combination that's brutally hard to replicate. Luna: And it's not just about farming. Deere is using the same playbook in construction — their 'Smart Construction' platform does similar data collection for bulldozers and excavators. Lucas: Yeah, they're essentially building an industrial IoT moat. And the interesting thing is, the hardware margins are actually compressing — steel and components costs are up — but the software margins are expanding. So the overall profitability is rising. Luna: Let's talk about the competition. How are CNH and AGCO responding? Lucas: They're trying. CNH launched a digital platform called 'CNH Industrial Services' and acquired a precision-ag startup called Raven Industries. AGCO has its 'Fuse' platform. But they're starting years behind, with smaller datasets. Deere has something like a ten-year head start in field data. Luna: And the sheer installed base gives Deere a cost advantage too — they can amortize software development over millions of acres, while a competitor might have thousands. Lucas: Right. The unit economics are brutal. Deere's R&D spend on software is massive — about $2 billion a year — but spread over a huge base, the per-acre cost is tiny. A smaller player would have to charge farmers much more for similar software, or lose money. Luna: So the moat is really a combination of three things: the data network effect, the switching costs from locked-in data, and the scale advantage in software R&D. Lucas: Plus a fourth: the brand. Deere has been a trusted name in farming for over a century. Farmers trust that the equipment will work in a blizzard or a drought. A startup might have better AI, but if the tractor breaks down and the nearest service center is 200 miles away, that AI doesn't matter. Luna: Which brings us to the dealer network. Deere has about 1,500 independent dealers in North America alone. That's a physical moat that a software company can't replicate overnight. Lucas: Exactly. And Deere is now using that network to push its software subscriptions. Dealers are trained to sell the Operations Center along with the tractor. It's a classic bundling strategy — you're already buying the machine, so why not add the analytics for a few hundred dollars a year? Luna: I want to push back a little, though. Some analysts argue that Deere's moat is overrated because farmers are increasingly tech-savvy and willing to use multiple platforms. A farmer might use a Deere tractor but run a third-party analytics tool like Climate FieldView from Bayer. Lucas: That's a fair point. And Deere has had to open up somewhat — they now allow some data export to platforms like Granular. But the core analytics — the predictive models for planting and harvesting — those are still exclusive to Deere's system. Luna: So the moat is not absolute. It's more like a wide river that can be crossed if you have a good boat. But most competitors are still on the shore trying to build a raft. Lucas: That's a great metaphor. And the key question is: how fast can competitors build that boat? Deere is investing about 7 percent of revenue in R&D, and a big chunk of that is software. Meanwhile, CNH and AGCO are more capital-constrained because their hardware margins are thinner. Luna: Okay, so let's look at the financials. Deere's operating margin in 2025 was around 20 percent. That's higher than most industrial companies. And the software subscriptions carry a gross margin north of 70 percent. Lucas: Right. So as the revenue mix shifts toward software, the overall margins should keep expanding. That's the bull case for Deere's stock — it's not a cyclical equipment maker anymore, it's a tech-enabled industrial with recurring revenue. Luna: But there's a risk: if commodity prices fall, farmers cut spending. And Deere's new equipment sales are still cyclical. In a downturn, farmers might defer buying a new tractor and just keep using the old one — even if it means losing the software benefits. Lucas: That's true. But Deere's software subscriptions are relatively cheap — a few thousand dollars a year — compared to a $200,000 tractor. So even in a downturn, farmers might keep the subscription to get the efficiency gains. The software is counter-cyclical in a way. Luna: And the autonomous tractor could be a game-changer. If Deere can sell a fully autonomous system that lets a farmer run a fleet from a tablet, the value proposition becomes even stronger. Lucas: They've already started limited commercial deployment in 2025. The autonomous tractor costs about $500,000 — more than a standard one — but it can run 24 hours a day, which effectively doubles the farmer's labor capacity. The ROI is compelling. Luna: That also raises the switching cost. If you've integrated autonomous Deere tractors into your workflow, you're not going to switch to a competitor that doesn't have autonomy. Lucas: Exactly. And that's the scary part for everyone else. Deere is pulling ahead in autonomy, data, and services, while the hardware is already best-in-class. The gap is widening, not narrowing. Luna: So in summary, John Deere's moat is a combination of data network effects, switching costs, scale in software R&D, a trusted brand, and a vast dealer network. It's a modern moat built on an industrial foundation. Lucas: And the lesson for other industrial companies is clear: don't just sell the machine. Sell the intelligence that makes the machine smarter over time. That's where the defensible advantage lives. Luna: Honestly, if this episode moved your thinking about industrial moats even a little, I'll say what we always say: if it was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Lucas: And it keeps the show ad-free, which we love. So thank you to everyone who's chipped in. Luna: Yeah, it makes a real difference. Anyway, back to the moat: one thing we didn't touch is the role of regulation. Do you see antitrust action as a real threat? Lucas: It's possible. The Biden administration took a tougher stance on data monopolies, but under the current administration, it's unclear. Deere has enough political clout in farm states that I think they can fend off the most aggressive moves. But the right-to-repair issue is real — if farmers can repair their own tractors and access the data, some of that lock-in weakens. Luna: So the moat is durable but not impervious. And that's probably how Deere wants it — they keep investing to stay ahead. Lucas: Right. And as long as they keep the data loop spinning, and keep adding capabilities like autonomy, I think the moat will only widen. It's one of the most interesting strategic stories in industrial America right now.