Vertical Farm Basics and Overview

Why Listing the Pros and Cons of a Vertical Farm Never Gives You the Answer

Show one manager the drawback “the upfront investment is large” and he balks; show it to another and he leans in, thinking “that is actually cheap.” The first came from open-field farming, the second has run large greenhouses. The same single sentence cuts in opposite directions depending on where you stand. A list of vertical farm pros and cons is usually written as an average that belongs to no one’s bottom line in particular. So you can read the whole thing through and still not see which items matter for your own operation and which do not.

I myself have been involved with more than ten sites over more than ten years. I have been on the team for some of the largest startups in Japan. And what I saw again and again was people who walked in trusting a tidy list, then froze the moment they met the reality of the floor. In this article, I lay out how to “run that list through your own conditions.”

A Pros-and-Cons List Is Nothing More Than an Industry Average

Start looking into vertical farms and you usually run into a tidy list. Unaffected by weather, can be grown pesticide-free, allows planned production. Then the downsides follow: it costs a lot in electricity, the upfront investment is heavy. But no matter how long you stare at that list, you cannot see which of those items will really matter for your own operation. Have you never felt that strange lack of traction?

Even within the same list, the items hit in completely different ways. “Unaffected by weather” is a big deal for someone working open fields. But if you are thinking about it from an already stable greenhouse, or from another line of business entirely, it does not resonate as much. Depending on your starting point, half the merits can simply vanish. “It costs a lot in electricity” is the same. The unit price of power swings hard with fuel prices and world events at any given moment, and it also shifts with how you structure your contract, your scale, and how you procure. The very same “electricity cost” can be fatal or a rounding error depending on when, and under what conditions, you draw it. So before you take the pros and cons at face value, you have to put them through your own conditions first. Otherwise you will never know what they truly weigh.

The pros-and-cons lists out in the world are, in effect, an average for the whole industry. A free-floating generality that belongs to no one’s bottom line. Just looking at it, you cannot in principle tell whether it applies to your operation. The one bit of effort that is missing is “running it through your own conditions.” Let me give you the conclusion up front. The list by itself means nothing. Only the items that remain after you run them through your own electricity unit cost, sales channels, and frontline capability actually move your bottom line. What remains is the real set of pros and cons for that particular person.

The research backs this up too. In a comparison that grew lettuce hydroponically, yield per unit area came out roughly 11 times that of open-field growing, while the energy it took ballooned to as much as 82 times that of open-field growing (see 1). The same single item, “high productivity,” swings to a powerful merit or a powerful burden just by changing the axis you view it from. Even if the list says “high yield” in one line, behind it is energy use several times over. So you cannot get an answer by tallying the items up at face value.

Stress-Test Drawbacks Harshly, Merits Modestly

When you run the pros and cons through your own conditions, the two are not symmetrical. In what I have seen, missing a merit usually amounts to a missed opportunity. You could have made more of it but did not. It stings, but the business keeps turning. Miss a drawback, though, and it cuts into the bottom line. Set out with the electricity cost penciled in too kindly, and it bites every month as a loss. With the same single item, getting it wrong is a deduction on one side and a lost point on the other. This asymmetry is something I have watched play out again and again on the floor.

So you change how you run them through, too. On the drawback side, plug the worse numbers into your own electricity unit cost and sales channels. Especially for things like the electricity cost, which swing upward depending on fuel prices and world events at any given moment, do not set them at an optimistic value. If a drawback still remains even after you substitute the worst-case value, then it is real, and you build it head-on into your design. On the merit side, by contrast, it is enough to count as your strength only those that still matter even when you estimate them under the best conditions. Drawbacks harshly, merits modestly — run them through the sieve asymmetrically. That, I believe, is how you do it without getting burned on the floor. Stop reading the list as if every line weighed the same, and tip the sieve toward the side where you stand to lose. The drawbacks that survive the worst case and the merits that survive even good conditions. Only when you have both in hand does the outline of your own bottom line start to come into view.

Take a drawback too lightly and it eats into your bottom line — and the reality on the floor bears this out, too. Vertical farms have been demonstrated again and again to “work as technology,” yet there is no end to cases that stall short of that because the numbers do not add up. Even an older survey that followed case studies kept citing siting and cost barriers as factors blocking adoption (see 2, 3). That said, the impression of “deficits everywhere” misreads reality if you lump the types together. Split by cultivation form and the picture sharpens. According to the Ministry of Agriculture, Forestry and Fisheries’ FY2025 survey, among operating facilities as a whole, 64% were in the black or breaking even. Of those, greenhouses and the combined sunlight-plus-artificial-light type both had over 70% in the black or breaking even, while fully closed PFALs that run entirely on artificial lighting had only about 50% in the black or breaking even (see 12). In other words, what is struggling is not so much the industry itself as the PFAL that supplies all its light with electricity. And the PFAL is what this article mainly has in mind. Putting the drawback side through the wringer is not over-caution; it is bracing head-on for the standard way the most loss-prone type tends to hit a wall.

What Moves the Bottom Line Is Electricity Unit Cost, Sales Channels, and Frontline Capability

A pros-and-cons list has many other items lined up, starting with pesticide-free and space-saving. And yet the levers that truly move the bottom line are not many. By my reckoning, what matters comes down to three: electricity unit cost, sales channels, and frontline capability.

Why these three? Because a vertical farm’s bottom line, pushed to its core, comes down to “lighting things up with electricity, selling them for some price, and harvesting as much of them as you can on the floor.” What holds the large part of the cost is power, and what holds the ceiling on revenue is the sales channel. The fat part of the money going out and the fat part of the money coming in. And what finally swings both of them is how much of the crop you can bring in at a good packout rate on the floor: frontline capability. Pesticide-free and space-saving are certainly merits, but they are not the leads in the bottom line. They are no more than fine adjustments that operate inside these three. Seasoning, once the headliners are set.

These three differ in nature. Electricity unit cost and sales channels can be locked down on paper before you break ground. What contract you assume for electricity and at what price, what you grow and who you sell it to. These can be decided in advance, at the stage of drawings and contracts. So the reach of this article is, first, to draw the outline of the bottom line with these two. Frontline capability, on the other hand, is a multiplier that kicks in once you are up and running. With the same equipment, the same electricity unit cost, and the same sales channel, how much you harvest and the quality both change completely depending on how the floor is managed. So frontline capability has to be held on a separate layer — the variable that tips the bottom line last.

Here let me separate out one point that is easy to confuse. “Labor cost” and “human resources” are talk of the same people, yet they work entirely differently within the bottom line.

Seen as an accounting line item, labor is one cost line. And it is not a light line, either. Even in the lettuce cost estimate I cite later in this article, labor is the single largest line item, above the electricity cost (see 6). In the FY2025 survey too, the highest ratio overall was labor, accounting for roughly 32 to 36% across greenhouses, combined types, and PFALs alike. Even in a PFAL, electricity cost is 24%, with labor larger (see 12). In other words, the lead on the cost side is not just the electricity cost everyone talks about, but labor as well.

Yet view human resources as a “driver that moves revenue” and the story no longer fits inside an accounting line item. Floor skill decides yield, quality, and packout rate across the board. Look at output per labor hour by financial result, and for PFAL lettuce, businesses in the black get 5.2 kg per labor hour and those in the red get 3.0 kg, nearly a twofold gap (see 6, 12). With the same equipment, how much you harvest swings this much depending on whether the floor is running well or not. This is exactly what I have watched many times on the floor. Bring in the most advanced system, and if there is no one who can use it well, yield does not rise. “A vertical farm’s productivity is decided by people, not by the most advanced system.” This is the single biggest realization I have come to over the years of doing this work.

So I view people on two layers. Labor as an accounting line item is, like power and sales channels, something you estimate on paper first. And human resources as a revenue driver is held separately — a multiplier that tips the bottom line once you are up and running. Both hold true. Not estimating labor too small on the cost side, and not putting off investment in people on the revenue side, do not contradict each other.

The matter of scale and crop, at root, rides on these three axes too. Scale works on the power side, that is, on cost efficiency. It is a matter of how much yield you can divide the fixed costs of equipment and lighting over, so if it is too small, the electricity cost per plant gets heavy. Crop is on the sales-channel side. Do you push lettuce out to a market anyone can enter, or deliver a high-unit-price item to a fixed buyer? With the same factory, what you grow and who you sell it to changes the sales-channel numbers across the board.

The view that “power is the large pillar of production cost” is backed by cost-structure estimates too. In a PFAL vertical farm, power accounts for roughly 20 to 40% of production cost, and on top of that, lighting eats more than 60 to 80% of that power. This structure has been consistently confirmed across multiple estimates and life-cycle assessments (see 4). For vertical farming lettuce, there is also a report that it takes about three times the energy of greenhouse cultivation, of which around 60% is LED lighting (see 5). The fat part of the money going out is held by lighting, that is, by electricity. You can treat this as a fairly stable fact.

The idea that scale works on the power side has also taken concrete shape in model estimates. Economies of scale operate on construction cost, and there is an estimate that when scale grows 100-fold, the per-unit construction cost drops by an average of 55%. On top of that, estimate the break-even for lettuce and the minimum viable scale moves greatly depending on premises like selling price and wages. Even a slight drop in the selling price makes the required scale balloon at once. With the same lettuce, just nudging the conditions of scale and selling price changes the premise for profitability across the board. The wage side is the same: a slight rise in labor cost moves the required scale. Labor cost, too, works on the bottom line independently (see 6). An item only takes on weight once you put it through power and sales channels — this is exactly the kind of swing I mean.

Sort Drawbacks Into Three Layers: Structure, Design, and Operation

There are drawbacks like power, whose weight is decided by your own conditions. Once you see that, you next want one more level of sorting. For even within a single word, “drawback,” things of different natures are mixed together. Things you can eliminate in the first design, things you can absorb in daily operation, and things that structurally remain no matter what you do. These three live together inside the one word “drawback.” The weight of the upfront investment, for instance, can be reduced through your choice of scale and equipment, but it never goes to zero. A slightly poor packout rate, on the other hand, you can tighten up in operation. “Eliminated by design, absorbed in operation, structurally remaining”: this distinction can be drawn if you set out to draw it. If anything, it is because the three get lumped together unsorted that countermeasures spin their wheels.

First, the “structurally remaining” things. These are the fate you take on the moment you choose the vertical farm method. The flagship is the weight of the upfront investment. Unlike open fields or greenhouses, unless you stack the building, lighting, HVAC, and nutrient solution equipment first, you cannot harvest a single plant. You can make it lighter through scale or used equipment, but it does not go to zero. Using electricity as the energy source itself belongs here too. Think of this not as something to eliminate, but as a precondition you fold into your profit and loss from the start, for how much it bites every month.

Next, the “eliminated by design” things. These are decided by a single judgment before you set out: what contract you draw electricity on and how far you assume the unit price, where you set the scale, where you site it, which crop you sell to whom, and so on. Once you have built, they are hard to move afterward, but flip that around and, before you break ground, you can choose again. This is where the earlier “run it through the worst case” thinking kicks in. You lay out, in advance and on paper, a design that still turns even when you plug in bad numbers.

Last, the “absorbed in operation” things. Packout rate, how you run the HVAC, the hands-on labor of your people. These are a realm you tighten up through daily accumulation, and the final numbers do not come out until you have run it. This is the layer where the earlier “frontline capability” kicks in.

What matters is the order of the distinction. Swallow the structural ones. Stamp out the design ones before breaking ground. On top of that, leave to operation only “the part there was no way to move by design.” In what I have seen on the floor, much of the weight foisted onto operation traces back, at the root, to deferral at the design stage. So before you say “you cannot know until you run it,” question it one level: is this really an operational problem, or was it something you could have decided by design? Not getting this backward is what matters most.

The view that “using electricity as the energy source itself structurally remains” comes out clearly in research comparing resource use too. A vertical farm can cut land, water, and pesticides compared with conventional agriculture, while energy consumption rises instead. This trade-off among resources has been confirmed repeatedly across multiple life-cycle assessments (see 7). That you can hold down land and water is itself a real merit, but it comes only as one set with the precondition of “carrying energy structurally.” And what is more, depending on where you get that energy, the very conclusion changes. There is also an analysis that if you produce with electricity centered on fossil fuels, carbon dioxide emissions actually come out larger than conventional agriculture, and only when you premise the use of renewable energy or waste heat can it flip (see 8). Resource saving is a real merit. But it comes, from the start, as one set with the structural drawback of energy. This is why you cannot pull just one side out of the list and count it.

A Flagship Merit Becomes a Price Only When It Rides a Sales Channel

Sort the drawback side into three layers, and this time something catches your eye on the merit side. What snags most is the flagship merits, like “pesticide-free” and “stable supply.” When these are talked about, they are always lined up as something valuable. But ask whether they really get onto the selling price, and you cannot quite feel sure. Pesticide-free, so you can sell it high; planned production, so you can deliver steadily. As reasoning, I get it. But unless there is actually a buyer who pays that extra, as a merit it is all signboard and no shop.

Pesticide-free and stable supply are the “seed” of a merit, but they are not yet the merit itself. They become a merit only once they reach the selling price; until they do, they stay pie in the sky. So discount every flagship merit on this one point: is there a buyer who pays the premium? This is where the earlier asymmetry kicks in. Count pesticide-free and stable supply by whether they still remain even when you estimate under the best conditions. “Looked on favorably” is not enough. Whether you have a buyer who commits on paper, by name: “I will take this much of this item, at this price, every week.” If it goes that far, you add it to the bottom line as a real merit. If it does not, you design it as zero.

Goodwill and money are different things. Even if vertical farm vegetables are looked on favorably in consumer surveys, that favorability does not become a sales channel as is. Unless the wholesaler or retailer standing in between passes the premium through, that goodwill does not reach your own selling price. So what to look at is not the consumer’s feelings, but whether the buyer standing between you and the money actually loads that premium on and passes it through. Thought of that way, pesticide-free and stable supply are less a wellspring that generates a merit, and more a “fitting that bolts onto a sales channel,” becoming a price only once it is tied to a specific channel. First lock down your sales channel, then confirm how much that buyer pays for pesticide-free or planned production. Until that is decided, do not put the flagship merits into the bottom-line numbers. Looking at it that harshly is just about right.

That said, this strength of “stable supply” has one more precondition. You harvest steadily only because the floor is managed properly. A vertical farm is tough against the weather but fragile against the human factor. Production dropping from a management mistake or a shortage of floor skill really does happen. I myself, on the day of our first harvest, had problems pile up and ended up doing the packing work alone for close to 60 hours straight. The flagship of stable supply only turns into a real price at the negotiating table when there is frontline capability behind it.

The conditions under which that stable supply becomes a “price” are, if anything, strengthening, given the way the climate has been swinging in recent years. In 2026, a heat wave of 42 to 47°C hit Maharashtra, India, and open-field vegetable yields dropped by 50 to 100% depending on the region. Some on the ground even said, “at 46 to 47°C, crops cannot survive.” Even when open-field crops are wiped out on this scale, a facility that can manage temperature can keep shipping. The strength of “weatherproof” becomes real value to buyers and procurers alike, the more extreme weather turns from forecast into fact. Even so, the precondition does not change: that, too, is only the case if the floor is running.

And what is more, even if you find a buyer who pays the premium, that is not the end of it. To actually go and capture that premium, there is often an additional outlay required: certification to claim pesticide-free, or the inventory and logistics to turn planned production into “stable supply.” A flagship merit only nets out when you look not just at whether there is a channel to ride, but at whether it survives after you subtract the investment needed to get it there. A merit that costs money to enjoy, you count after deducting that part. Here too, viewing it modestly is just about right.

The same goes for the flagship of “high productivity.” That “being productive” does not become the bottom line as is comes out sharply in research that lines up the productivity numbers against the economic numbers. There is an estimate that a 10-tier vertical farm has roughly 100 to 200 times the productivity per unit area of open fields (see 9). Yet that overwhelming productivity advantage has not, at present, been converted into an economic advantage large enough to offset the high capital and operating costs. There is also an analysis that, if anything, the enormous capital investment blocks getting major crops onto the bottom line at all (see 10). Harvesting a lot physically and turning that into money are two different problems.

On the difficulty of the sales-channel side, there are points that touch on the mechanics of distribution too. In research that estimates demand on the premise of pushing to wholesale markets, how to reconcile the uniform, planned vegetables a vertical farm produces with the trading rhythm of the existing market is treated as a challenge (see 11). That said, the research described here focuses mainly on demand-forecasting methods and does not pronounce that it is “structurally incompatible with the market.” In fact, many vertical farms take the form of delivering directly under contract to supermarkets and food service, rather than to ordinary wholesale markets (see 6). So this is something to keep in the back of your mind as one point that comes up when you push to wholesale. Either way, even the technical merit of planned production will not be the strength you expected unless it matches the sales-channel side that receives it.

A Net Pros-and-Cons Table Is the Doorway to Judging Profitability

By now, the net pros and cons that remain after running them through your own conditions should have taken on a fairly clear shape. At the end, let me set down just two boundary lines.

One is the handling of subsidies and government support. The systems that push entry along are certainly generous, but they only ease the early burden for a while; they belong on a separate layer from the bottom line of a business that has to keep turning every month. Pile subsidies into the merit column and the clean outline you worked to draw goes blurry. Whether the bottom line stands, look at first on the bare numbers with support stripped out.

The other is that what you build here is, in the end, only the “doorway.” Run things through electricity unit cost and sales channels, and hold the drawbacks that remain even at the worst case and the merits that remain even under good conditions. That far is the reach of this article. But that does not decide the bottom line entirely. What you can lock down on paper goes as far as power and sales channels. What finally decides the bottom line is frontline capability after you are up and running: that is, who runs that design and how. So once this net table is made, you can move into a deeper level of examination: how to set the scale, which method to choose, how to design the sales channel, and whether it still holds when seen over a multi-year payback. Above all, how you recruit, train, and retain floor talent is the final doorway deciding whether you can actually hold the “frontline capability” I am talking about here.

To sum up, it comes to this. The pros-and-cons lists going around out there are nothing but an average that belongs to no one. Run it through your own electricity unit cost and sales channels, and sieve it asymmetrically: drawbacks at the worst case, merits under good conditions. Only what remains becomes the net table that moves your own bottom line. Do not pile subsidies into the bottom-line column. Look first at whether it stands on the bare numbers with support stripped out. And the net table made here is only the doorway; the talk of scale, method, and sales-channel design, and the frontline capability that runs them, is the deeper talk that lies beyond.

The payoff of this work is that you can finish it while it is still just lines on paper. Electricity unit cost, sales channels, and crop are all hard to move once you have broken ground. But before breaking ground, you can plug in as many bad numbers as you like and redraw it as many times as you want. The merits that remained even after you plugged in the worst case, and the drawbacks you have resolved to swallow as structural. Hold both, and only then can you move on to the next examination with peace of mind. Put it the other way: setting out with this left vague is what costs the most.