The Sector Everyone Wrote Off Is Working Again
Vertical farming spent 2023 through 2025 as a cautionary tale. Bowery Farming, once valued near $2.3 billion on more than $700 million of venture capital, shut down. Plenty, backed by roughly $1.19 billion, filed Chapter 11. AeroFarms went through bankruptcy in 2023. Industry trackers counted fourteen controlled-environment agriculture bankruptcies in a single year, and of the twenty-three companies that signed a joint vertical farming manifesto in 2022, fewer than half are still operating.
And yet the survivors are posting profitable quarters. AeroFarms restructured and now supplies national grocery chains with microgreens. Plenty came out of bankruptcy with a single crop in a single facility. The lesson of the shakeout was not that indoor farming does not work — it was that the version funded like a software company does not work. What replaced it is smaller, duller, and considerably more likely to survive.
Why the Warehouse Became the Default Building
A vertical farm needs a big clear-span box with a tall ceiling, a concrete slab, truck access, and industrial zoning. That is the definition of a warehouse. It also needs to be near the customer, because the entire value proposition is shelf life and freight — you are selling lettuce that was cut yesterday twenty miles away, not lettuce that rode a refrigerated truck from Salinas for four days.
Urban and inner-ring industrial stock delivers both at once, and there is a great deal of it sitting empty. The same vacancy that makes warehouses attractive for indoor pickleball facilities and event venues makes them attractive here, and the rent is usually the smallest line in the budget rather than the largest.
The catch is that a warehouse is optimized for storing dry goods, not for holding a rainforest at a stable temperature. Everything expensive about the conversion follows from that mismatch.
What Actually Killed the First Generation
The failure pattern is remarkably consistent, and it is not agronomic. Companies scaled before unit economics worked. They behaved as technology companies first and farms second — custom robotics, engineering salaries, flagship facilities — and built capacity before they had buyers for the output.
Underneath that sat a structural problem: a vertical farm replaces free sunlight and free rain with purchased electricity and engineered systems. Produce has to command a price premium large enough to cover that substitution. For commodity lettuce sold into a national grocery contract, it frequently does not. The farms that died were selling a commodity at a premium cost structure and hoping volume would close the gap.
The Three Numbers That Decide Everything
Before agronomy, before equipment vendors, before a single rack drawing, a conversion lives or dies on three figures.
- Energy per kilogram. Lighting consumes 65 to 85 percent of a vertical farm's energy and HVAC most of the rest, so electricity can reach 40 percent of total operating cost. Legacy facilities burned 40 to 120 kWh per kilogram of lettuce. Well-designed 2026 builds land closer to 11 to 15. That spread is the difference between a business and a science project.
- Revenue per square foot. High-value leafy greens and herbs sold into premium local channels can generate upwards of $200 per square foot of growing area annually. Anything sold into commodity channels will not.
- Contracted offtake before commissioning. Not a letter of intent. A signed volume commitment from a grocer, distributor, or restaurant group that covers a meaningful share of week-one capacity.
If you cannot produce credible figures for all three, the correct move is to keep modeling rather than to sign a lease.
Sell It Before You Build It
The single behavioral difference between the survivors and the casualties is sequencing. The casualties built capacity and then went looking for buyers. The survivors secured buyers and then built exactly enough capacity to serve them.
This inverts how most conversion projects run, and it is uncomfortable, because it means spending months on sales conversations for a product you cannot yet grow. Do it anyway. A grocery buyer's answer to "would you take 400 pounds of basil a week at this price" is the most valuable piece of information in the entire project, and it costs nothing to obtain.
What to Verify Before You Sign a Lease
Four building conditions disqualify more candidate warehouses than anything else, and all four are checkable in an afternoon:
- Electrical service. A grow facility's connected load is enormous relative to warehouse use. Utility service upgrades routinely run six figures and take longer than construction — sometimes a year or more. Ask the utility what capacity exists at the transformer before anything else.
- Slab capacity. Multi-tier racks loaded with water concentrate serious weight per square foot. A slab poured for pallet racking may need reinforcement.
- Zoning and use classification. Farming is not a warehousing use. Many industrial zones do not list agriculture at all, which can mean a use variance rather than a permit.
- Drainage and sanitary connection. You will periodically dump spent nutrient solution. Where it legally goes is a question with a real answer, and sometimes an expensive one.
The Realistic Entry Path
The version of this project that works for someone who is not deploying institutional capital looks like this: lease a modest bay rather than a campus, run one crop family rather than a catalog, commission a single pilot rack and grow it to spec for a full season before ordering the rest, and sign your first buyer before your first light fixture.
That path is slower and far less impressive in a pitch deck. It is also the one still standing in 2026. Work through the full build in our warehouse to indoor vertical farm guide, and price your own version with the conversion cost calculator.