We started Transition on a simple belief: human prosperity is capped by physical, planetary constraints. The best founders see this and see the opportunity to build a company that matters.
Few industries are more physical, or more foundational, than agriculture. It is also one of the oldest. Sub segments are large, you have a ~$90B seed market and a ~$70B in pest protection market. Yet it’s owned by 20th century incumbents and almost no one has succeeded in building a company worth billions since. Will that change?
The last few years have shown how exposed farming is to geopolitics. Russia’s 2022 invasion of Ukraine sent grain prices spiking. This year it is fertiliser: since shipping through the Strait of Hormuz nearly halted, prices have jumped, and the World Bank expects a 31% rise across 2026.
The deeper, more continuous pressure on agricultural outputs is the impact that climate change has on heat, drought, and erratic weather. Farmers struggle to deliver the same yield in these increasingly difficult conditions. One farmer told the BBC this week that the heatwaves have cost them half their output.
Why the past has failed
Ag tech solutions usually fall in one of the following three categories: inputs, seeds or hardware/robotics. Across these categories, there are structural forces why startups face major challenges on 1) reaching scale velocity and 2) distribution - both of which are key to building a venture scale company.
Startups have to move from working product to real adoption fast.
Farms are low-margin businesses operating in heterogeneous environments, with seasonal needs and volatile outputs. Selling new technologies in this context is particularly difficult, start-ups struggle to prove consistent ROI (many natural elements outside your control also impact the result), it’s slow to do so (need to wait a full growing season, which for many crops is once a year), the willingness to pay up front is minimal (limited cashflow), and the risk aversion is high (margins are tight so there is no room for error). In addition to that, in certain categories like pest-protection inputs and gene-edited seeds, start-ups need to pass regulatory approval, which is slow and expensive. In other segments like hardware, the challenge is financing when the equipment must be bought up front.
The companies that move fastest usually deliver a meaningful uplift in farm economics and they remove all possible friction to adoption. Upside Robotics is a good example. Its lightweight, solar-powered robots have the potential to cut over half of the farmer’s fertiliser use whilst managed remotely. They don’t require farmers to buy heavy equipment or change how the farm runs to prove the economics.
Startups have to earn trust across a fragmented pool of farmers, varying in size and crop types. This is particularly difficult in Europe compared to the US, given the different languages and national industry associations. Incumbents have grown slowly over generations to own the majority of the market.
For inputs and seeds, the “Big Four” - Bayer, Syngenta, Corteva, and BASF - control 70% of the global pesticides market and 60% of the seed market. Their agricultural divisions return EBITDA margins of 15 to 20% (FY 2025). Corteva, the only pure-play agritech company, trades at roughly $50B and ~14-19x EV/EBITDA . However, the Big Four M&A activity has historically capped out around ~$200M to $400M.
The picture is similar in hardware, with six machinery giants owning roughly half the legacy equipment industry. John Deere, the biggest of these, is worth ~$160B with EBITDA margins of ~20% and EV/EBITDA ~25x. Acquisitions also seem to be capped. Blue River sold to John Deere for $305M in 2017, considered as one of ag-tech’s large exits, and still not enough to return a classic seed-stage fund.
Many startups respond by partnering with or licensing to these incumbents. That might get the product to the market, but rarely yields a venture-scale outcome. To succeed, companies need to find a way to reach farmers directly. This is something that Farmer’s Business Network (raised ~$918M since their founding in 2014, including backing from Google Ventures and Temasek) has been known for doing well, including offline methods like using flyers distributed locally. This is also something that Dominic Roth did incredibly well at Stenon (raised a total of $41.6M since 2018 with investors like Founders Fund) with targeted facebook campaigns.
Why now might be different
Ag tech saw an investment boom in 2021/22, with $18.9B invested, followed by a dip due to a couple of years of missed exits that left investors wary. Many failures came from false assumptions about how the sector works and ignoring or underestimating the constraints (see this substack for a deeper dive). Now, armed with those lessons, there might be a case for looking at ag tech again, with a more realistic lens.
As software becomes cheap and easy to build, capital is moving back into the physical world, to atoms rather than bits. Y Combinator’s CEO Garry Tan recently made waves by highlighting that the cracks in chemical pesticides are now “obvious”, and calling for builders to address them.
So what changed? And what hasn’t changed?
As mentioned in the introduction, farmers are experiencing unprecedented pressure from geopolitical crises and climate change. Although the ability to pay is lower than ever, the willingness to try and adopt new solutions is high. The industry is at a breaking point.
Another factor that is rarely talked about is biosecurity. Unlike what we might initially think, agriculture has a much higher likelihood of large scale bioterrorism attacks than human-centred attempts. This is because the pathogens studied are in lower-safety labs, modern farming packs thousands of near-identical plants or animals densely together so a single introduction can take a whole farm or region out, and sites sit remote and unguarded, with symptoms usually showing weeks after spread has begun (read more in this substack).
The good news is that, unlike human biodefence which is government-dependent with lumpy revenue, farmers already pay for pest protection, whether or not an attack ever comes. That means founders building in biosecurity can develop technology that addresses everyday threats (e.g., for rapid pest detection) and catches a novel or deliberate one, creating a defensible biosecurity venture.
Conclusion
So, will we see multi-billion dollar exits in ag tech in the coming decade?
The structural problems have not gone anywhere, but for the first time in over a decade, the pieces seem to be moving: regulatory doors are opening for biologicals and edited traits, and AI and robotics are coming down the cost curve. We see the biggest potential in alternative crop inputs (biological or physical) and AI/robotics that add clear ROI with low upfront costs.
When we invest, we don’t back a sector. We back the right team taking a big swing at a gigantic, hard, physical problem. On that basis, we’re open. If you’re a founder, scientist, operator or investor building toward multi-billion dollar outcomes in ag tech, we’d love to exchange ideas. Please reach out.