Vertical Farm Basics and Overview
The future of vertical farming cannot be measured by exit news — not by a millimeter
Does vertical farming have a future? The moment you put that question into a grow-or-die binary, the answer stops coming. Reports of exits and reports of new entrants arrive with equal weight, so if you believe one, the other always snags on something. So let’s stop scoring for a moment. Instead of measuring the future by the light and dark of the news, ask: “If this business keeps running, what is being satisfied when it does?” Move the axis of judgment there, and you start to see that exit news is neither evidence for nor evidence against a future. That axis is what we’ll build together from here.
A note: the “future” I’m talking about splits in two. One is whether the industry itself grows — demand, how fast the technology advances, how it holds up against the climate compared with open-field cultivation. The other is how many years the one operation you’re looking at right now keeps going. On the growth side, indoor cultivation has plenty of tailwinds of its own, and that side is worth arguing separately. This article deals with the latter. The industry growing and the single vertical farm in front of you surviving look like they overlap, but they don’t. So here, I narrow the focus to “how to size up, from the outside, whether your one operation keeps running.”
What divides the future is not technology but whether you can keep going
What’s going to happen to vertical farming from here? Lately, news that a big company has pulled out keeps coming, and at about the same volume you hear about new companies entering. Watching from the outside, you can’t settle down. Is this a growing industry, or one that’s already finished — which is it?
When you watch exit news from the outside, “it failed technically” rarely takes center stage. It’s not that the vegetables wouldn’t grow, or the quality was bad. What’s front and center looks more like operational matters — electricity costs rose, the sales channels wouldn’t widen. When that repeats a few times, something comes into view. This isn’t a problem of the technology for growing; it’s a problem of whether you can keep the thing running. Once you see it that way, the future starts to feel like something you can’t measure by today’s mood either.
Look a little more closely at the substance of these exits from the outside, and what surfaces still tends to be: electricity costs that crept up and got heavier, sales channels that didn’t widen as hoped, people who didn’t stay. What stops is not the technology for growing; it’s whether the operation keeps running for years. So measuring the future as up or down by the mood on the day exit news breaks is off the mark. When you look at one operation from the outside, the first footing is how that vertical farm supports the four things — power, capital, sales channels, and people. This doesn’t mean “the four decide it.” Location, policy, local demand, climate, and what you grow all matter separately too. It’s just that these four are easy to count from the outside and they tend to bite hard. A vertical farm where those four mesh tends to keep going quietly, regardless of today’s news. Where they don’t mesh, even with a good public mood it tends to stop sooner or later. Whether the industry as a whole grew or finished, and whether the one operation in front of you keeps running, are separate stories. So rather than lumping the industry together, go one operation at a time, starting from these four. That, I believe, is the reading that lowers your odds of getting it wrong.
From the research side, the direction lines up too. Reviews that reexamine urban and indoor agriculture on sustainability grounds say, over and over, that the equations “local, therefore sustainable” and “grown in the city, therefore kind to the environment” don’t hold as they stand. What decides whether it lasts is the way it’s managed, the energy source, what you grow, and the climate (see: 1, 2). The read is that operating conditions decide it, more than the technology itself. One nationwide US survey also lists the biggest walls as profitability, financing, and production cost. When that fails to come together, that’s often what an exit comes down to (see: 3).
Power, capital, sales channels, and people work by supporting each other
Even if electricity costs run heavy, you can bear it if you have a sales channel that pays well. Even if the sales channel is thin, you can hold out if your people stay and waste doesn’t pile up. The four — power, capital, sales channels, and people — bite not separately but linked together. Is it a combination where, if one slackens, another can carry it? Have you ever looked at a vertical farm with that eye?
What’s scary, conversely, is a vertical farm where all four are at the edge. A single price rise sets off a chain that collapses it. That’s exactly why, the better a vertical farm is doing, the more worth asking it, “Where is your weakest spot?” A vertical farm that has someone who can say the weak point out loud while things are going well probably tends to last longer — this is my feel for it, not a correlation I’ve verified. But an operation that knows its own weakness is far more reliable than a distant industry forecast. The future, surprisingly often, shows up at the feet of an operation like that.
The point that the heaviness of electricity costs can “be carried elsewhere” lands straight away when you look at the cost-structure numbers. In a PFAL vertical farm, electricity makes up 20 to 40 percent of total production cost, and within that electricity, lighting eats 60 to over 80 percent (see: 4). Power is not a small line item you can trim with a clever fix. It’s a cost that bears down heavily from the start. What makes it worse is that both the initial construction cost of getting started and the operating cost of keeping it running (power, utilities, and so on) bite as walls. Lighten only one side, and you jam on the rest (see: 5, 6). Capital, too, bites in both the weight of building and the weight of running. That is exactly why the view “is this a combination where, if one slackens, another can carry it?” does the work.
See the whole industry and the one operation in front of you as separate questions
Judge the future by the conditions for staying open — it makes sense, yet one snag remains. Power, capital, sales channels, and people are all invisible while you watch from the outside. Only at exit news do you find out, after the fact, “ah, so this was the thin spot.” Which makes it seem as if whether something keeps running is known only to the insiders, and all the bystander can do is wait for the exit report. You start to think that way too.
But this snag is just two questions overlapping. “Whether the industry as a whole grows or finishes” and “whether the one operation you’re looking at right now keeps running” are separate questions. The former you can barely predict from the outside. The latter — you can’t predict the outcome from outward form either, and let’s be honest about that. But for the latter, even without peering inside, you can count candidate weaknesses ahead of time. Even if you can’t see the numbers themselves: what the vertical farm leans on for electricity, where it draws its capital from and whether it has a picture that extends past the point its subsidies run out, whether its sales are skewed toward a single buyer, whether its people keep turning over. This kind of outward form is, surprisingly, visible.
There’s one thing to be careful about here, though. Look back at a vertical farm that exited, and from the outside you can read it: the lopsidedness you could have spotted even from outside often showed up first — contracted power on a single leg, a reliance on subsidies with no picture beyond them, sales dependent on a single buyer. But you can’t take this alone and bet “therefore it will exit.” The same lopsidedness is perfectly common in vertical farms quietly running in the black. There are vertical farms skewed to a single buyer that have run for years, and places that launched on subsidies and rode straight onto the rails. Line up only the ones that exited and extract the common factors, and it sure looks like “this is the cause” — but as long as you haven’t looked at the surviving side, that isn’t proof of cause. The most outward form sorts for you is candidate weaknesses.
So what the bystander can do is neither wait for the report nor go chasing a prediction, but count the operation’s lopsidedness ahead of time. You can’t peer inside, but what it leans on shows up in the outline. Watch that, and when the exit report does come you’ll have a handle for reading it — “ah, that thinness.” Sometimes you’ll be wrong. Even so, it beats staring at the reports having counted nothing, by exactly the margin of having seen the fragility ahead of time.
The way the outward form “is the buyer skewed to a single company?” bites shows up sharply in the numbers, in the form of price sensitivity. In one model calculation, the minimum scale at which lettuce just barely pencils out is as small as 17 to 38 square meters. But drop the sale price by 20 percent and the break-even point jumps to 1,700 square meters; drop it by 30 percent and you enter the zone where most vertical farms don’t pencil out at all (the region where the minimum viable scale begins to diverge); drop it by 35 percent and the minimum viable scale leaps past 100 hectares (see: 5). A small move in the sale price changes the required scale by an order of magnitude. How much sales channels and price bite is right here. Power is similar. Environmental impact changes greatly with the makeup of the electricity the vertical farm runs on, and the assessment flips depending on whether you run it on wind or on coal — that’s the read from the research side (see: 7, 8). Viability, too, is pulled hard by electricity costs, so it likewise turns on the power source (see: 4, 8). The “what it leans on” you count from the outside is not just an outline — it’s a matter that swings the order of magnitude of viability.
A yes-or-no binary throws the judgment off
Even after counting one operation from the outside, in the end you’ll want to sum it up in a phrase — “so, does this deal have a future or not?” But that “yes or no” has a trap. Yes-or-no is a way of speaking that crushes the lopsidedness you went to the trouble of breaking into four back into a single pass-or-fail mark. The moment you declare it, what it leans on goes out of view.
There’s a trap in the opposite direction too. See a precedent that’s doing fantastically — payback in so many months, a return of tens of percent — and you’re tempted to climb aboard with “well then, it has a future.” But that strong showing is probably a single photograph of the four happening to mesh right now. Your own place has different power and different sales channels, so borrowing the result alone won’t carry the substance over. What you may borrow is not the numbers but the way of seeing.
As a source for the “losses keep going” story, the figure that about 70 percent of Japan’s vertical farms run in the red is often cited (see: 9). But this is a press-based number rather than a statistic that properly counts all operators, and it’s not even clear what range it uses as the denominator. If anything, the recent Ministry of Agriculture, Forestry and Fisheries “Survey on the Actual State of Large-Scale Protected Cultivation and Vertical Farms” (FY2025 edition) shows operators in the black or at break-even making up 64 percent overall — above half — with losses at roughly a third. That points nearly the opposite way from the press-derived “about 70 percent in the red” (in the same survey, even narrowing to PFAL, about half are in the black or at break-even). So if you take one thing home, it’s closer to reality to frame it as “it swings widely by type and by year” than as the old impression that “many vertical farms run at a loss.” Just as we don’t lift one operation’s strong showing straight onto the industry’s future, we don’t use a single “70 percent” to make a flat claim about the whole industry. That fits better with the view we’ve held throughout — looking at the conditions for staying open, one operation at a time.
Review the future not once but through fixed-point observation
The future isn’t something you can decide is a “yes” once and be done with. It’s a subject you keep watching, the conclusion left open. Whenever any of power, capital, sales channels, or people wobbles, review it then. It doesn’t suit being decided once and shelved.
The same goes for the span of time. The way viability looks during the startup phase differs from after the operation settles. Deciding it’s finished from the first loss alone, and deciding you’re safe because year one went well, are both too quick. Don’t make a conclusion out of the numbers at any single point in time.
What you grow changes how it looks too. Growing staples like rice and wheat indoors does not pencil out at today’s technology and current prices. A Swedish calculation for wheat puts the electricity cost alone at around 40,000 dollars per ton — reaching roughly a hundred times the world wheat price (see: 8). On top of that, LED efficiency gains are already near their ceiling, so it’s hard to assume technology will slash the cost all at once (see: 8). So “deciding it’s finished from today’s numbers alone” is too quick, but “climbing aboard the idea that technology will eventually solve everything” is just as quick. The stance of holding judgment in reserve and observing at a fixed point is also a way of standing equally clear of both sides’ jumping to conclusions. This holds strictly “at today’s technology and current prices” — it does not mean it’s impossible in principle forever.
So when you size up a business’s future from the outside, what should you count starting today? What it leans on for power — where, at what price, on a contract of how many years. What the capital is covered by, and whether there’s a picture that keeps running after the subsidies run out. How many companies the sales are split across, whether it’s dependent on a single one. Whether the people stay, or keep turning over. These four outlines you can check without peering inside. Just because at least these four can be counted from the outside, don’t forget that location, policy, demand, climate, and the choice of crop all matter separately too. The better things are going, the more you ask “which is weakest” and re-count it now and then. Not to predict the outcome, but to see the candidate fragilities ahead of time. That alone is enough to keep the way of seeing going.