Crops

The vegetable that pays in hydroponics does not come down to the "crop"

Two factories grow the same lettuce, and one runs in the black while the other runs in the red. Across all the factories I have watched on the floor, this was the thing I found most puzzling. If you are about to build a business plan on the premise that “you can turn a profit as long as you pick a vegetable that pays,” it is worth stopping for a moment. Because the thing that decides, at the very end, whether profit is left over is not the name of the vegetable.

The vegetable that pays is not a fixed list — you size it up as a product

Search for “hydroponics vegetables that pay” and you get endless rankings: lettuce is good, no, strawberries, no, melons. As you read, you start to feel that the right answer is lying around somewhere. Yet within the range I have seen on the floor, that right answer does not necessarily hold at your own place. With the same lettuce, if where you sell it or how you run it differs, the profit appears or vanishes.

I have seen this kind of scene with lettuce many times myself. Sold to a farm stand, it turned over solidly, but the moment I tried to wholesale the same thing for commercial use, the unit price dropped all at once, and I was busy with nothing to show for it. The reverse, too: an herb I could only grow in small amounts, carried straight to a restaurant, unexpectedly turned a profit. Lining these up, you start to feel that what pays is not the crop itself but the crop paired with “where you sell it.”

This feeling connects to a story that reorders the very idea of crop rankings one level up. The way I size it up, the vegetable that pays is something you hit more often by viewing it as the product of your own channel price, turnover days, and growing difficulty, rather than as a fixed list. Lettuce was exactly that. At the farm stand I could take a high unit price, so the speed of turnover translated straight into profit. Pushed into commercial channels, the unit price dropped, only the turnover remained, and no profit was left over. Herbs were the opposite: the volume was small, but with direct sales to restaurants the unit price was high, and turnover on the shelf worked too. With the same equipment, the buyer decides the price and the turnover, and onto that the labor of growing is stacked. So the question is not “which vegetable” but “in my channel, which one has a unit price x turnover worth its growing difficulty.” There are people for whom lettuce is the right answer and people for whom herbs are, and that is simply because their channels differ.

That said, one caveat is needed here. The high unit price at the farm stand holds only on the premise that you grow at small scale, in a hybrid setup. A closed-type lettuce factory built at large scale is on a different ground entirely. Bagged product for mass retailers who can move volume, and commercial-use cuts for restaurants and prepared foods, become the main battleground, and small-lot, high-price channels like farm stands are too small to absorb that volume relative to the scale. So read the “reason lettuce dominates” that comes later on the premise of this mass retail and commercial use.

In fact, the categories where vertical farms can make the numbers work are narrowed down almost entirely to leafy greens, herbs, and fruit-bearing crops. Studies report that staples like rice, wheat, and corn have, at current costs and technology, no prospect of clearing the profit line in the first place — even though those three crops are foundational ones accounting for roughly 60 percent of the world’s food energy (see 1). What pays is not a property fixed to the crop; it is dragged along by whether that crop can hold a buyer where it can take both price and turnover.

The labor of defending the channel’s price also goes into the product

If direct sales let you take a high unit price, you would think everyone would rush to direct sales to restaurants. Yet in reality, their number is limited. Why? The labor of going out to win a channel, and the effort to maintain the relationship, ought to be a “hidden cost” separate from growing difficulty. Should this, too, be folded into the product?

To put the conclusion first: it should. But the way it goes in is different. If growing difficulty is “a division that recurs every time you make one unit,” then opening a channel is closer to “a fixed cost that hits heavily at first and works thinly after.” A relationship with a restaurant costs sales labor to start up, but once trust is built, the unit price stays stable and keeps turning. So if you look only at the first year’s profit and loss it does not pay, yet from the second year on it suddenly starts to work. As far as I have seen, direct sales had exactly that character.

The reason no one rushes in lies there too. This hidden cost cannot be dissolved by capital investment. Anyone can copy lettuce’s capital investment, but a relationship with a particular chef cannot be replicated. So the high price stays high. Sales labor is not outside the crop selection; it is the very factor that explains “why that channel’s price is defended.” Incidentally, where there is room to raise the unit price itself without changing the crop is a question that should be set up separately. In practice, for each channel, on top of unit price and turnover, you add one column for “the labor per month of maintaining this relationship.” Then a reversal sometimes appears in the numbers: the commercial wholesale that looked easy actually sinks under management cost, while the direct sales that looked troublesome survive. Still, idealizing direct sales too much is dangerous too: orders can suddenly drop at the buyer’s convenience, off-spec product may not be taken off your hands so waste piles up, the relationship itself may depend on a single person in charge — you carry these weak points on a separate tab.

Looking at the research side, you run into the same place. When you search for why vertical farms and urban agriculture have a hard time spreading, things like profitability, initial investment, and energy cost are raised as the main barriers to adoption (see 2, 3). It is less that technology is lacking, and more that how to put together the numbers is the bottleneck. How to secure a channel that yields a price, and how to maintain it, touches the very core of that profit side.

Lettuce-centered is not the answer that pays but the answer that is hard to break

Look around at actual vertical farms and you see that most are lettuce-centered. Even among leafy greens it is lettuce above all, and they do not reach much into fruit-bearing crops (see 7). Is this because “lettuce pays best,” or is it only that it looks that way for another reason? It is worth asking again.

The reason many factories are lettuce-centered is not that “lettuce pays best” but that lettuce fits the factory format best. What works here is not the channel but another term of the earlier formula, the growing-difficulty side. Lettuce has a short growth period and turns over fast, is low in height and easy to stack in tiers, and the management of light and nutrient solution is established. In other words, in the factory’s closed environment it can be run with high reproducibility, as planned. Fruit-bearing crops do not go like that. Tomatoes and strawberries need control of pollination and fruit set, the growing period is long, and they are tall too. With the same floor area and utility cost, the labor of growing becomes far heavier.

The difficulty story has one more side I have felt strongly on the floor. Lettuce, even when the management of temperature or nutrient solution concentration slips a little, is unlikely to turn into a catastrophic crop failure. Growth unevenness is relatively low too, and it costs less than for other crops to learn enough to stabilize shipment quality in the early going. Where this comes to matter is when you picture the lineup of companies entering vertical farming. Even on the floors I have helped start up, there were very many operators newly entering from other industries. People who do not know plant physiology in their bones come in carrying a large capital investment. From that position, “we cannot afford to fail” stands first. Choosing lettuce first — which does not get wiped out even if you slip up a little, and has a low learning cost — was, rather than profit maximization, a perfectly natural call to avoid going under.

So the landscape of lettuce-centered is the result not of “the answer that pays” being chosen, but of “the answer that is hard to break in the factory format.” Put the other way around, for a grower who can manage the difficulty of fruit-bearing crops, or a person who holds a channel that stands precisely because it is a fruit-bearing crop, the room for a higher unit price sits just outside that row of look-alikes.

This view is backed up by the structure of energy and cost as well. Comparing an open-type greenhouse that can ventilate against a fully artificial-light vertical farm, in almost every location the greenhouse has higher energy efficiency, and there is a comparison showing that this difference in energy efficiency reaches 45 to 94 percent depending on location (see 4). Since a closed factory shoulders the whole utility cost, leaning toward crops that turn over fast and can be run as planned is a natural dynamic. The range of crops you can grow also changes depending on whether it is fully closed or sunlight-utilizing type, but the more closed the environment, the more strongly the constraints of turnover and difficulty come out. And the place where hydroponics shows a large yield advantage over soil cultivation is, for the most part, leafy greens (see 6). How large that rate of increase turns out to be also varies considerably depending on the crop and growing method, it is reported (see 5). That the crops well-suited to factories lean toward leafy greens including lettuce works from both sides, difficulty and yield.

High-price fruit-bearing crops do not turn a profit on price alone

If there is a channel that sells strawberries or tomatoes at a high price, you might as well swing all the way over to that. Even a factory in the red on lettuce might turn into the black. There is probably also the naive expectation that, even knowing the labor of growing is heavy, if the unit price is higher than that, the books should balance. On the other hand, the doubt also surfaces: if it went that well, everyone would already be doing it.

Honestly, it is true that the books do balance at a high price sometimes. But the unit price is only one term of the formula, and turnover days and marketable yield come into play at the same time. Strawberries take several months to harvest, and the several-turns-a-month of lettuce cannot be hoped for. So even at several times the unit price, looked at as annual sales per unit of area, the gap does not open as much as you would think. Moreover, fruit-bearing crops have large variation in fruit-set rate and spec rejections, and as a floor-level feeling the marketable yield is hard to read. When this is not stable, the shelf that was supposed to be high-price drops out of the calculation entirely.

There is also the equipment problem. You cannot divert the equipment now in the red on lettuce straight over to fruit-bearing crops. Height, pollination, and the design of light and nutrient solution are all different things, and conversion carries a separate investment. Once investment is on, how many years you reckon to recover after deciding the crop also changes. So rather than “swing over and it transforms,” it becomes a conditional story: only when two things come together — the technology to stably exceed a certain yield level even after paying the equipment-conversion cost, and a channel that catches that fruit-bearing crop while keeping the high price — does it finally clear the profit line. The reason everyone does not do it is that satisfying both at once is hard. Turned around, for those who can satisfy them, room is still left. Strawberry and melon factories are in fact emerging (a grower-first perspective on the “vertical farm x fruit” that is hot now, strawberries and melons), and there are crops like wasabi where a good fit has come into view (Can wasabi be grown in the hydroponics of a vertical farm).

Here is a trial calculation where the substance of the difficulty shows clearly. In one model calculation, the smallest scale at which strawberries are commercially profitable is, at the current level of technology, about 115,000 square meters. A number nearly out of reach. Yet with just a 20 percent rise in unit yield, that break-even point is estimated to drop all at once to about 1,200 square meters — a change of order of magnitude (see 7). Fruit-bearing crops do not become “profitable because the unit price is high”; whether you can exceed a certain yield level is the key to the numbers. In the same calculation, lettuce too is said to be so price-sensitive that a slight drop in price collapses the whole basis for profitability at once. In other words, both unit price (= channel) and yield are variables that, depending on conditions, reverse profitability, and it is not the crop itself but the way this product works that decides the numbers. That said, this is one point in one model calculation, and depending on your own variety, equipment, and electricity rates, the result moves greatly. What you should read off is not the absolute value of the numbers but the way the effect works — that a slight move in yield or price changes the basis for profitability by an order of magnitude. It overlaps with the floor-level feeling that fruit-bearing crops’ yield rate is hard to read, and in facility cultivation the optimal nutrient solution concentration differs by variety, so even when you optimize one thing, studies report the result changes depending on another condition (see 9, 8).

Even among leafy greens, the gradations of each term are fine-grained

Even with the same leafy greens, lettuce, komatsuna, and baby leaf are different. That the difficulty term works strongly for fruit-bearing crops is as we have seen, but choosing between leafy greens too comes down, in the end, to that same product of “channel price x turnover x difficulty.” It is not that, because they are leafy greens, the conditions are similar and it makes little difference which you pick.

Leafy greens against one another, too, are viewed by the same channel price x turnover x difficulty. Only, unlike fruit-bearing crops, the axes where differences appear are fine. In terms of turnover days, baby leaf is fast at a few weeks when picked young, komatsuna takes a bit more, and leaf lettuce harvested as a head takes longer still. This difference in turnover days works in slowly, through the annual number of turns per unit of area.

The channel side is finer still: the preference for spec and pack form differs by crop. Within the range I have seen, product for mass retailers was centered on bagged leaf lettuce and baby leaf, komatsuna lined up in bunches, and commercial use put spec uniformity first — there was such a tendency. By channel, the crop and pack form that get through change. Furthermore, there is variation in environmental response by variety, and even with the same leaf lettuce some varieties are weak to heat and prone to bitterness and bolting. With a closed type, the season itself can be controlled, but if you miss on variety selection, the weakness to environment comes out as-is.

So going in thinking “any leafy green is the same” is dangerous. A crop chosen for ease of growing does not mesh with the pack form or spec your channel demands, and though you can grow it, you cannot sell it out. You end up carrying inventory that way. The product is the same, but within leafy greens the gradations of each term are fine-grained. Viewing it that way is closer to the truth.

Do not start from the ranking — line up from your own channels

Up to here, the axis of viewing crops as the product of channel price, turnover, and difficulty has been fairly well established. What you finally wonder is how to handle the world’s “vegetables that pay” rankings and the cases of successful factories, and what someone about to decide a crop from here should start with.

Rankings and success cases can be a reference, but you cannot make them a blueprint as-is. The biggest reason is that the information flowing around is skewed toward “stories that went well.” Factories that pulled out, or cases that folded in the red, leave almost no data. So even if you copy the top crops, you will not get the same result. Still, even a case of a factory that turned a profit centered on leafy greens, read together with its conditions — which channel and how it was run for it to hold — is plenty usable as material to fit into your own formula.

While watching out for survivorship bias, the one thing you can still say is in the direction that niches where demand exceeds supply do actually exist. For example, the branded baby romaine product (Andy Boy Baby Romaine Hearts) that D’Arrigo California commercialized in the U.S. in early 2026 was a product released after two years of trial cultivation, and it is reported that right after commercialization, demand exceeded supply. It is a niche, a superior alternative to romaine lettuce, appealing the expense advantages of low waste and many shipment units per carton. Read this not as a guarantee of achievement that “therefore it pays,” but as one example in the direction that, even within leafy greens, there really are vacant lots where demand exceeds supply. Just as there are pullouts not spoken of behind survivorship bias, there is, just as surely, demand not yet filled — both sides.

The order of starting is not from the crop. First, line up your buyers and prices in a single row. Onto that, write in turnover days, growing difficulty, marketable yield, and even the labor of maintaining the channel, then compare crops by annual profit per unit of area. This is the first move. As for how to lay it out, set leafy greens that turn over fast and are easy to read at the center, and consider fruit-bearing crops secondarily. Fruit-bearing crops have room in unit price, but they only stand when the conditions come together, so it is realistic to first solidify the base numbers with leafy greens and then layer them on.

In fact, it has been pointed out that there are quite a few facilities in the red among domestic vertical farms. Even peer-reviewed research shows that roughly 80 percent of Japan’s PFALs turned over within about ten years, and that the margins are so thin that a mere 30 percent drop in lettuce price brings them close to bankruptcy (see 7), and industry commentary too has repeatedly spoken of the abundance of facilities in the red (see 10). It is a structure where, behind the success cases that surface, there is a fair number of pullouts not spoken of. That is exactly why, rather than starting from the crop with “it is top of the ranking so it will work at our place too,” you line up your own channel, turnover, and difficulty in a single row and compare by annual profit per unit of area. This order is the way least likely to crumble from under your feet.