The unit economics of venture-backed vertical farming

Imagine growing a salad inside a luxury penthouse using only designer lamps. That’s vertical farming: a world where VCs poured billions into warehouses, convinced that kale would scale like software.

But plants are stubborn. Unlike code, you can't copy-paste a strawberry. You’re paying tech-hub rents and massive electricity bills to mimic the sun, just to sell a product that shoppers expect to be cheap.

It’s a brutal math problem where you often spend two dollars to grow one dollar of lettuce. Without VC subsidies, these high-tech greenhouses are usually just very expensive landlords for very picky greens.

Wait, if the math is that bad, why did anyone invest?

VCs were drunk on the "software is eating the world" Kool-Aid. They saw a warehouse full of sensors and LEDs and didn't see a farm; they saw a data center that happened to output arugula.

The pitch was "Moore’s Law for Spinach." Founders promised that as LED efficiency doubled and robotics got cheaper, the cost of indoor farming would plummet while traditional farming would get pricier due to climate change.

They bet on a future where tech would eventually outrun biology. Turns out, a head of lettuce doesn't care about your fancy API; it still needs the same amount of photons to grow, no matter how many disruption slides you show.

But couldn't they just use solar panels to get that light for free?

It sounds like a clever loophole, but it’s actually a physics nightmare. You’re taking sunlight, converting it to electricity, and then converting that electricity back into light. You lose energy at every single step.

It’s like using a flashlight to grow a plant while standing in a sunny field. You’re buying millions in hardware just to recreate a fraction of the energy the sun provides for free.

Traditional farmers get their 'fuel' delivered every morning at zero cost. No matter how cheap solar gets, it can't beat the price of 'free.'

So why not just use a greenhouse and get the best of both worlds?

You’ve hit on the industry’s biggest 'oops' moment. Greenhouses are the sensible middle ground—they use free sunlight but still protect plants from weather. They’ve been profitable for decades because they don't try to fight physics.

But VCs didn't want 'sensible.' You can’t stack greenhouses because the top floor would shade the bottom. To get that 'infinite scaling' software vibe, you have to go vertical, which means ditching the sun entirely.

They traded free energy for the ability to stack lettuce like server racks. It’s a case of over-engineering a solution until the original goal—growing food cheaply—becomes secondary to the tech.

Seriously, if it's basically a data center, why do they still need farmers?

That was the dream: a 'lights-out' facility where robots did the dirty work. They wanted to turn biology into a 'set it and forget it' operation, just like a server farm.

But plants are messy. They leak and rot. While a server just sits there, a strawberry needs a delicate touch that a million-dollar robotic arm still struggles to replicate without turning it into jam.

They didn't eliminate labor; they just made it pricier. You traded a seasonal worker for a six-figure software engineer to fix the robots.