The first, most critical fact to recognize about our food system is that it’s solar powered.
Take a drive along US-85 in Wyoming, or US-26 through Nebraska, or US-18 through South Dakota and it is impossible to miss dozens of clusters of cows standing by the roadside, slowly, contemplatively chewing. That’s what cows do all day, every day — they eat grass. They use that plant energy to locate more grass, and eat that too. A cow’s ceaseless job is to convert the planet’s most fibrous, lowest quality feed into milk and meat. A cow is an engine that turns sunlight to dollars.
It is also an engine that pollutes: Cows alone contribute almost as much to climate change as all of the cars in the world — over 10% of all human-produced greenhouse gases. Their nearly indigestible diet nourishes a microbial community which respires methane into belches and nitrous oxide into manure. Tropical rainforest, and all the carbon it keeps, is leveled to create pasture. Gasoline is burned to keep the butchered meat cold and fresh. Every pound of steak we produce adds almost thirty pounds of carbon to the atmosphere. Every pound of steak protein requires 25 pounds of vegetable protein to grow.
It used to be that the only folks to complain about the global environmental impact of livestock were the sort of people who would uncomfortably lock gazes as they explained the healing power of crystals. But the rise of solar energy has taught Silicon Valley that one person’s fringe is another’s early adopter, and venture capitalists are lining up against the rancher, betting that modern technology can outcompete the world’s livestock. Just, a vegan mayonnaise company, has raised $220 million from private investors and is looking to raise $200 million more. Impossible Foods, a veggie burger company, raised nearly $700 million. Beyond Meats went public in May at a valuation of over $1 billion and is, at the time of this writing, worth over $9 billion.
All of this is intended to position Silicon Valley in advance of an unavoidable trend: For the climate to sustain, global dairy and meat consumption must be cut in half from today’s levels. In developed nations like the US, beef consumption must drop 80% or more by 2050 to make room for consumption by billions of increasingly well-off, meat-hungry Asians and Africans in the coming decades.
So this is what disruption looks like: Four out of every five of those cows on the side of the road must vanish for the planet to survive.
But the cow is a new sort of target for Silicon Valley. A cow is not a hunk of capital like an internal combustion engine or a coal-fired boiler. A cow will not join your social network or be called by an API. A cow is an animal which has been naturally selected, over millions of years, to turn feed into protein as efficiently as nature allows, lest it be out-competed by another, more efficient creature. A cow, thick with meat and fat, is dinner that humans have evolved to love.
Silicon Valley has convened to disrupt a thing that is solar powered, fully autonomous, and has — to borrow one of the Valley’s favorite terms — perfect product-market fit.
Is Silicon Valley insane?
Cows, eaten just a few meals per week, account for 10% of global greenhouse gas emissions including the contributions of methane from cow belches, nitrous oxide in manure, CO2 from fertilizer, and clearcutting to add pasture. Globally this is just 10% less than the contribution from passenger cars and light trucks, which account for about 60% of transportation emissions and 60% of the pollution from oil drilling itself. In the developed world transportation our cattle operations are more carbon-efficient thanks to factory farming. ‘Intensification’ of cattle ranching in the developing world will help, but it is not enough to get us to our carbon goals.
It’s an exhilarating narrative: The same technology-fueled capitalism that led us to pollute will be harnessed in the name of conservation, and driven in to save us.
From the perspective of a cow, air pollution is a feature — cows and other ruminants burp methane to clear their stomachs of chemical waste that otherwise slows digestion. Moving from ‘factory farms’ to grass-fed beef makes the situation worse, not better, by extending the time the cows spend emitting carbon before they are brought to slaughter. Incremental improvements are not up to the task: A century of breeding has not improved the feed efficiency of cattle. There is no such thing as a ‘clean cow’.
Silicon Valley is right: If we are going to save the climate, we have to remove beef from our menu. The challenge is to reckon how.
The campaign to replace meat with plants is emboldened by a new enviro-capitalist gospel forged in the growing triumph of renewable energy. Thanks to decades investment, local wind and solar produce energy cheaper than three quarters of US coal plants. LEDs and compact fluorescent bulbs occupy more than half of lighting sockets in the US, and have contributed to the first residential energy use decline in the country’s history. Electric vehicle costs will drop below those of internal combustion engines globally in a handful of years. Innovation births products that are faster, better, and (with increasing scale) cheaper than the original, and economic incentives have rapidly transformed behaviors that decades of activism have failed to budge.
The investment community’s embrace of Impossible and Beyond sings those same hymns: Surely a technology that makes vegetables taste like meat will be more environmentally efficient, and come at lower economic cost.Yet the parallels are not quite perfect. First, even factory farmed cows eat mostly grass, and grass is very, very cheap.
But more critically, the cow is not a photon or an electron. It is not an end in itself. To the consumer a cow is simply one answer to the question “What’s for dinner?” And while targeting the lonely cow for disruption seems easier than transforming eating, it may be that Silicon Valley is not thinking big enough.
US per capita beef consumption has been declining for decades, eclipsed by cheap chicken. Can environmentalists turn this into a rout?
My father was a man of entrenched habit, and he consumed nearly the same meal every night: Rice cooked in bullion, sliced tomato, and a roll stuffed with ground beef.
He was in his mid-forties when he had his first heart attack. A cruel combination of diet and genetics drew plaque into his arteries, and after returning from the hospital he resolved to change what he could by replacing his dry, crumbled beef with ‘Vege-Burger’ from Loma Linda. His cans of textured wheat gluten never appealed to the rest of the family, but vegan meat met my father’s admittedly low standards — it was bland, reliable, and would not kill him. Our house sustained an irreversible technological shift: Cows came off the menu and never returned.
Fake meat was a better product for my father. Beef threatened his health, and textured plant protein substituted seamlessly into the only recipes he knew how to cook. Fake meat was “faster” too — he could (and did) keep months of cans ready in the basement pantry. And what was true for my low-tasting father is becoming true for everyone today: Beyond’s patty is the culmination of decades of slow and steady improvement in food texturing and flavoring, a straight line of development from the original Worthington’s Fry Chik in the 1950s to Boca burgers and Morningstar Farms sausages in the 80s to “burgers that bleed” today. Vegan meat has proven to be the ‘Hey Siri’ of food, annoyingly imperfect yet improving every year, and now finally good enough to cross over from enthusiasts to the mainstream.
But both then and now, fake meat is not cheaper.
Historically, the market for fake meat has been miniscule. Before the rise of Beyond and Impossible, the entire ‘meat alternatives’ industry in the US accounted for $650 million in revenue, compared to over $200 billion for animal proteins. Ground beef outsells fake meat by a factor of nearly 100-fold, so it is no surprise that it is less expensive. But the opportunity to close the gap remains: a rough rule of thumb for manufacturing is that increasing scale by 10X will drop production costs in half. Make a burger that the masses are willing to eat and costs will come down naturally. The world’s vegan meat corporations will recoup their investments as the market grows.
But this framework underestimates just how streamlined the cow — and the global beef production system — really is.
There is more to the cost of food than just the raw ingredients — there is processing, packaging, marketing, transport. The “farm share” of a Beyond Burger — the total retail value that returns to the farm — is probably just 15¢ per pound of patty. The processed Beyond Burger sells for $12/lb at retail — a price at which the company still loses money — meaning that 1.25% of the shelf price returns to the farm to pay for raw ingredients. This is well below the farm shares of high volume commodity processed food like bread (4%), tofu (2–3%), and roughly on par with a branded product like Coca Cola (1.6%).
Meanwhile, the farm share of a cow is 45%.
Almost half the grocery store price of ground beef covers the cost of the live cow. Thanks to the magic of commodity capitalism trucks are filled consistently, less meat ages in inventory, and little is lost to waste. With volume high, everyone in the supply chain subsists on a sliver of revenue, and grocery stores keep retail prices at or even below cost to drive traffic into the store.
How much do economies of scale save? The farm share of beef is double that of apples, and triple that of lemons. We can slaughter beef, box it, transport it across the country and stock it on shelves with lower added cost than we can put a sticker on fruit.
More astonishingly, the “farm share” of cattle feed — the silage (ground grasses or corn) and soy that the cow processes into more cow — is over 25% (about 50–60% of the cost of a cow is feed itself). We convert hay into packaged, shelf-ready ground beef at 10X the economic efficiency of cooking soy into tofu.
Environmentalists rightly criticize the cow for being inefficient in terms of vegetable matter, but that ignores the rest of the food economy. The cow may lavish resources on bones and a spleen and a nervous system that we don’t consume, but it doesn’t have to invest in solvent extractors and mixing vats and social media campaigns to raise brand awareness. A cow is a commodity, which means that little money is frittered away on salaries or profits for the ranchers (much to their dismay). A cow with a belly full of corn, soy, and grass can walk itself to slaughter in a way that a soybean can’t.
How does the Beyond Burger compare? And how cheap could it get?
To match the price of beef on the retail shelf and to win over the bulk of its price-conscious customers, Beyond Meat has to scale. Automated equipment must be erected to purify and texture pea protein, and burgers must be slung at terrifying rates to spread that cost across hundreds of millions of customers. To meet its promise to the Earth, Beyond has to get bigger. Terrifically, implausibly bigger.
To understand how big requires a little math. Today, Beyond’s small production volume — 20 million pounds of plant burger per year, or just 0.2% of the US ground beef market — leaves each burger 4X more expensive to produce than beef. If we use my simple rule of thumb that a 10X increase in scale will drop costs by 2X, then Beyond will have to scale 100-fold — to 2 billion pounds per year, or 20% of US ground beef production — before it matches beef on production cost. But if it also wants profits — a standard 60% “gross margin” above costs to pay for marketing and provide a return to investors — another halving of production costs and another 10X increase in scale is in order (recall that it must compete with beef, a commodity that gets by with barely any profit at all). A single company would have to produce more fake meat than the entire cattle market sells today.
This kind of volume is, of course, absurd. There are plenty of customers willing to pay extra for a healthier substitute, such as the 5% of US consumers who voluntarily upgrade to organic milk or produce. Capturing these enthusiasts would allow Beyond to make profits at a shelf price $8/lb (down from $12/lb today, and in line with a realistic farm share of 2%). But the price of ground beef in the US is half this, closer to $4/lb, leaving the bulk of the market no incentive to switch.
To skim the most lucrative customers from a market is a perfectly acceptable business model, but it is not disruption. We will not save the planet by better serving the rich.
Cows are clearly bad. The contribution of transport to GHG is omitted; it’s a small effect. Data on “Plant burger” is for the Impossible Burger.
Cost of production of different proteins in the US, scaled to ¢ per gram of protein. Data on plant-based burger is inferred from Beyond Meat’s S-1. Beef prices use ground beef as a cost baseline (steak would lead to higher estimates). In the US tofu costs more than beef to produce on protein basis. Actual costs will vary depending on country; in Japan, tofu would be 33% cheaper to process (at larger scale), and beef twice the price.
Simply put, there is no Moore’s Law of meat. There is no exponential technological improvement that inevitably turns a cow into a relic of earlier civilization like a steam engine or photographic film. Our canonical vision of disruption arises in industries overturning early, crude human technology, not a machine that Nature has built. There is no universe where technology-enabled meat is half the price of the original, the way that solar is already half the price of coal in the sunniest areas of the world.
The technology that Silicon Valley wants to disrupt is one that is already evolved for its unique task, to turn prairie into protein using no other inputs than the sun. And this raises the question, both for Silicon Valley and the planet: If beef cannot be underpriced, can it still be disrupted?
In the 1960s, Americans stopped eating veal.
The disruption came as a surprise; there was no warning, no “veal substitute” entering the market, and even after consumption had dropped by 75% experts still expected it to rebound. They could not imagine why demand collapsed in the first place.
The industry did understand that veal was a luxury good, an expensive dish suitable for special occasions. But amidst the rise of suburbia and television and the first frozen dinners, Americans began to cook less. And the first dishes to fall off the menu were the ones that were expensive, complicated, and rare.
Veal was disrupted not because of one singular technological change. It was disrupted by all of them.
Today, we may recall veal as a cruel relic — in 1986 the Humane Farming Association began a national boycott to protest the inhumanity of tearing calves from their mothers; of housing them in crates where they could not turn around; of eating a baby. But by then it was easy for the nation to turn on veal because we had already forgotten how to cook it.
Technology is the greatest tool we have to better the condition of humanity. But it is still not as powerful as choice.
The “great forgetting” we witnessed for veal is repeating with steak today. Like veal, steak is high price, high skill stuff that we no longer have the competence to prepare at home. Since the 1970s, steak demand has plummeted 60% as an increasing portion of beef production has been routed to hamburger. There is no need for a high technology “lab-grown meat” to disrupt steak; steak has been disrupted because we have chosen to do things other than cook. Like veal, steak is slipping into oblivion, awaiting another national moral movement to finish the job.
In the 70s, only 40% of the cow was destined to become ground beef. We still eat the same amount of hamburger, but now it’s 60% of the cow.
Will the rest of the cow follow? Today Americans eat three burgers per week; surely that is not optimal. We may believe we eat hamburgers for flavor and culture and tradition, but the pedestrian reality is that the average adult knows how to cook only five meals. Burgers have retained their position in our lives for no better reason than they are all we know. Beyond and Impossible aim to revolutionize mealtime by disguising environmentally-friendly food inside our existing burger habit. But it’s the habit that is inefficient, not the cows.
This is why technology disruption seems to catch us by surprise: We expect disruption to befall a singular technology, like kerosene replacing whale oil. But it is more common that disruption will topple an entire system, where a new capability reprioritizes our choices, redirecting attention and cash before it even has a chance to be spent. Taxis are being replaced not by faster cars, but by a better user interface. Newspapers were disrupted not by responsive web content, but by the loss of classified ads. Steaks go uneaten because we spend our evenings watching cable TV.
The greatest danger to the beef market is not that we will invent a technology that makes fake meat cheaper than cow. The danger is that we will invent technologies that usher us towards something else to eat entirely.
The trends are evident today. Globally, we are moving away not just from cooking, but from assembling meals at all: The average number of ingredients used per meal has declined 40% over the last 30 years. We have more one-dish meals: pizza, sandwiches, yogurt, and protein bars. Over a quarter of meals today are frozen or ready-made. Only fifteen percent of meals are eaten within one hour of preparation.
The reason for the decline is simple: Building a balanced meal is a pain. It takes time and energy to shop for ingredients. We no more want to build dinner than we want to build a personal computer or maintain our car. If apps offering meal kits and food delivery and nutrition planning can increase our dinner repertoire from five meals to ten or twenty — or allow us to forget the handful of recipes we can still make without instruction — hamburger consumption will inevitably decline. Fake meat offers a replacement for beef on your plate, but a well-designed app can intercept beef before the desire even forms.
To say that software is the real threat to the cow — that AI algorithms and lifestyle branding and a chic user interface have the potential to deliver us from beef — is perhaps the most Silicon Valley thought of all. But the battle for the soul of bovinity began decades ago, and it is not a war over what we find efficient to produce, or even what is efficient for dinner. It is a battle over what we find efficient for life. The brilliant mimicry of the Impossible Burger and the rest of the fake meat pantheon will of course accelerate a transition. But they are spark, not fire.
The next waves of software and delivery innovation will lower the bar for food radicalism, and make it possible to consider reconfiguring diet without all the tangle of burger-forming robots, refrigerated supply chain, and plastic wrap. We will vest in apps the messy work of converting our personal beliefs into dinner. The price of ingredients will become subservient to the price of serenity.
Cynics may scoff that we could be liberated from the environmental burden of cattle by something so nebulous as choice. Yet in the developed world the greatest grievance with our food system is not price or access, but that we fail to meet the ambitions of our aspirations. We can clearly articulate personal goals for our health, our environment, or our community, but lack the skills to perform. The most powerful technology to change our behavior is the one that extends our reach.
The veal industry collapsed in less than a generation — the 15 years from 1958 to 1973 four of five calves went unsold, exactly the plunge in cattle consumption that GreenPeace believes is needed to save the planet today. In the last 15 years veal consumption has fallen 80% again. Technology can create more than economic inevitabilities; it can forge tools to change behavior. The planet can be saved not by one decisive technological change, but by committed activists fastening all the smaller ones together.
Cows are and will remain the technology we wish we could invent: Modular, self-powered factories able to build value from waste. Cows are the open source solution for food, low cost and available to all. Cows are what our technologies might aspire to after a few million years of our own technological evolution, herds of them quietly grooming the planet
We will not stop eating cow because we have built a better meat. When we stop eating cow, it will be because we have built a better us.
 Acolytes of Allan Savory and Holistic Grazing Management please don’t @ me. Savory put on a great TED talk on how more naturalistic grazing practices can save the planet, but the scientific evidence in support of his thesis is scanty. The biggest problem is that purely grass fed cow can take 6 years to come to market — belching methane all the time — while a factory cow requires only 18 months. And it would take a lot more cows to achieve the same output.
 This is technically “new market disruption”, where the needs of part of the market are underserved by existing technologies — most particularly here, you can’t buy climate-neutral beef. There is another form of disruption, low-end disruption, where new technologies start cheap to begin with. Clayton Christiansen pioneered both concepts; read his stuff.
 I estimate Beyond Burger’s costs from its public S-1 filing. Beyond has positive gross margins today, but it is in total unprofitable because of R&D, marketing, and other expenses.
 This uses a technique called a ‘scale factor’ (or exponent) to estimate manufacturing costs in a factory, and sets the factor at 0.7, a consensus estimate for operations like mixing. That scale factor is what gets you the 10X volume à 2X cost rule of thumb. THIS IS A SWAG; feel free to push back, but defend your assumptions if you do. With mine, Beyond is coming up a factor of 10 short. Even knowing the exact output is wobbly, I’d like to see better odds of success for a project that is supposed to save the planet.
 The analysis is US-centric because I live in the US, Beyond and Impossible sell here, and data is plentiful. In Japan or Switzerland, plant-based burgers would be competitive on price with beef today. But Japan and Switzerland are not dictating the fate of the planet.
 In the 1970s, about 40% of the cow ended up as ground beef; today it is above 60%. While overall beef consumption is down 40%, nearly all of that fall can be attributed to steak. The industry has been slow to react to this trend, still producing cattle for higher margin steak rather than for ground beef, and it’s not clear to me that they have fully registered the level of the decline they are in.
 Here I want to credit Glenn Edens, who has shared some brilliant observations about systems-level disruption with me.