For years, we've been challenging and exploring how the venture capital model, designed for software, does and doesn’t fit agri-food. Now we have $18 billion in receipts.

Eugen Kaprov just published an analysis of 113 agrifood-climate tech failures totaling $18.5B in venture capital losses. His autopsy is thorough and gets a lot right. Unit economics matter. Competing on a green premium alone doesn’t work. Capital intensity without a clear path to profitability is fatal. 

But the analysis also oversimplifies—or misses—a few key dynamics at play in the last cycle. And understanding these nuances is critical if we're going to take the right lessons and build something different.

The VC timeline explanation is off. The analysis suggests that venture fund timelines (hitting year 7 and needing to return capital) forced fire sales. This may have happened in some cases, but I believe the reality was messier than just the clock running out: it was a lack of acquirers for overvalued companies while capital dried up everywhere.

Macro factors get underplayed. Beyond just interest rates killing growth-at-all-costs models across every sector, other macro factors compounded to cause a funding desert. Climate policy uncertainty and sentiment under the Trump administration froze public and private sector investment in anything touching climate. And AI sucked oxygen and capital out of every other category. These weren't agtech-specific problems, but agtech felt them acutely.

Ultimately, investor psychology created a loop spinning the wrong way. Venture capital is particularly susceptible to the whims of human psychology. When ecosystem confidence collapsed, especially in subsectors like alternative protein and vertical farming, even good unit economics could not overcome the fear of no follow-on funding. 

So what do we do?

Innovating in agri-food systems - however critical - is not easy. We’ve felt the fear and pain of this collapse acutely through our portfolio companies' fundraising experiences and our own. Q3 2024 was the peak of these crashes; it’s also when we held our first close for Fund II. 

We’ve evolved our investment model based on our successes and learnings. We’re continuing our ecosystem building and advisory work in recognition that scaling innovation to impact at a systems level takes far more than capital. And we’re building "more than venture" approaches like our partnership with Wine Australia announced at evokeAG 2026.

But across the ecosystem, we all need to be willing to find solutions for harder questions.

How can we prove the non-concessionary business case? LPs don't have to deploy in agtech,  but putting on AI polish won’t work for long. Climate change is already impacting profitability and returns across agriculture - there’s an economic opportunity to focus on.

Are we willing to experiment with fit-for-purpose funding structures? When LPs think in strict capital allocation silos, how can we build partnerships that align incentives and work in agri-food?

Are we brave enough to acknowledge exit dynamics? If public market comps for agri-food companies will always be around ~4x EBIT (not 20x revenue), and strong venture exits will be $200-300 million, does that change what's fundable? Or how we fund it? Are we brave enough to innovate within these paradigms?

How do founders navigate the gap between "realistic exits" and "what VCs want to hear"? Saying you're targeting a $200-300 million exit will kill most fundraising conversations, but claiming unicorn potential without a credible path there destroys trust. Is there a way to hold both stories honestly?

Are we willing to let go of the ego but not lose the ambition? What if bigger funds aren’t better, higher valuations don’t mean more success, and growth-at-all-costs isn’t the only path? What if we measured success by customer value created and shareholder returns delivered, not dollars raised or valuations hit?

The $18 billion in failures tells us what doesn't work. Do we want to take those lessons and build what does?