Case study — Food & beverage · 13 min read · 5 questions
Why the dairy industry is failing
The thesis
The vegan defense of Oatly is that the game is rigged. Cow milk is subsidized in every country on earth and oat milk is not, so plant-based brands are asked to build demand, build supply and match an artificially cheap price all at once. The first half of that is true. The conclusion is wrong.
Oatly's numbers are its own doing. Revenue went from $30 million to $783 million while operating income went from $1 million to a $344 million loss, and the shares went from $24.46 to $0.66. It committed half a billion dollars a year to owning factories on every continent on the assumption that 20–30% growth would hold, then growth fell to low single digits. It also excluded distribution from cost of sales, which the beverage industry includes; add it back and 2022 gross margin was 2.6%. Marketing ran 40% to 55% of revenue.
But the industry Oatly wanted to replace is not worth capturing. Milk is a political product: retailers sell it below cost to pull people through the door, politicians need it cheap, and the squeeze lands on producers and processors. Excluding subsidies, the average American dairy farm has been profitable in 3 of the last 23 years, and Dean Foods — the largest processor in the country, owner of the biggest brands — went bankrupt anyway. The subsidies do not reach processors. They reach farmers, and mostly the biggest: $347 billion to the top 20% of farms against $40 billion for the rest. Oatly is a processor.
How do you think about this? 5 strategy questions this case raises and does not answer.
Read the comments, or add yours →
The statistics
By the numbers — swipe or use arrows
Oatly revenue, operating income, SG&A, gross margin and channel mix from Oatly Group AB 20-F filings and quarterly reports; share price from Nasdaq closing data; Beyond Meat, Very Good Butchers and Meatless Farm margins from company filings; Dean Foods and Silk figures from Dean Foods Company 10-K filings through 2019; per-capita consumption, milk production, farm counts, price per hundredweight and profit per hundredweight from USDA; subsidy totals by crop, program and farm decile from the Environmental Working Group farm subsidy database, 1995–2023; retail prices via the Bureau of Labor Statistics; global dairy and non-dairy beverage margins are trailing-twelve-month composites of the public companies shown
Key takeaways
The plant-based bubble inflated and burst in five years. Oatly went from a $0.2 billion valuation in 2019 to $10.0 billion in 2021 and back to $0.4 billion; Beyond Meat went $2.0B, $7.0B, $1.4B; Impossible Foods went $5B, $8B, $4B. Nothing about the products changed. The story did.
Oatly shares have gone one direction since the IPO: $24.46, $7.96, $3.46, $1.74, $2.05, $1.18, $0.94, $0.66. At that equity value the company cannot raise another war chest except through debt, which is why the strategy is now cost-cutting rather than growth.
Nobody in the category makes money. Operating margins ran −14% and −13% for Oatly and Beyond Meat in 2020 and −49% and −97% by 2023. The Very Good Butchers posted −256%, −420%, −264% and −313%. Meatless Farm posted −208% and −183% before it stopped reporting.
The vegan tax is real. Cow milk averages $4.10 a gallon nationwide against $6.56 for almond and $6.97 for oat; ground beef runs $5.35 a pound against $6.70 for Beyond Meat and $9.32 for Impossible. Every plant-based brand is priced against a floor it cannot reach.
Oatly grew revenue $30M, $39M, $43M, $58M, $118M, $206M, $421M, $643M, $722M, $783M — and operating income went $1M, $0M, $0M, −$1M, −$10M, −$31M, −$47M, −$207M, −$352M, −$344M. The losses arrived exactly when the factories did.
Selling out of every drop was never the problem. Retail was 65%, 64% and 65% of sales across 2021 to 2023 — people were buying it by the box, not by the splash at a coffee shop — and the company still lost money on all of it.
Milk is structurally a worse business than every other drink. Global dairy companies run 35% gross and 10% operating margins against 47% and 25% for the non-dairy beverage giants, and 6% net against 12%. Milk is perishable, refrigerated and essential. Soda, energy drinks and alcohol are shelf-stable and optional.
Oatly was worse than both. Against dairy's 35% gross and 10% operating and non-dairy's 47% and 25%, Oatly posted 19% gross and −44% operating — while making its product from oats, water, oil and enzymes rather than from cows.
And that 19% was flattered. Oatly excluded customer distribution from cost of sales, breaking beverage convention. Add it back and reported gross margins of 24%, 11% and 19% become 16.4%, 2.6% and 12.8%.
Marketing is the other hole. Oatly's SG&A ran 40%, 55%, 52% and 53% of revenue against 27% for global dairy and 26% for non-dairy. When Oatly was the only oat milk it advertised the category; once PepsiCo, Coca-Cola and Nestlé shipped their own, it had to advertise itself as the expensive one.
Consumers do not reward milk with price. The nationwide average went $3.23 a gallon in 2004 to $4.10 in 2024 — roughly flat in real terms — while chicken went $1.03 to $2.06. Retailers price milk as a loss leader and put it at the back of the store, because it is the only product guaranteed to pull a household in multiple times a week.
Meanwhile Americans keep drinking less of it. Per capita consumption has fallen every decade since the school-milk mandates: 335 pounds in 1950, 275, 235, 215, 181, 169, 158, 145, 130 by 2022. A century of federal advertising, Got Milk and fast-food cheese partnerships did not reverse a single decade.
Production went the other way. The US made 201 billion pounds of milk in 2013 and 226 billion in 2022, a record, into falling demand — because subsidies reward volume. When prices rise farmers milk more cows to earn more, and when prices fall they milk more cows to cover the shortfall.
So the farms disappear instead. The country went from 44,809 dairy farms in 2014 to 26,290 in 2023 — nearly half gone in a decade — while output hit records, as family operations were replaced by corporate herds that processors prefer because it is one pickup instead of ten.
Subsidies do not save the small farm; they are awarded on acreage and production. From 1995 to 2023 the top 1% of farms took $101B, the top 5% $231B, the top 10% $296B and the top 20% $347B, against $40B for the remaining 80%. Dairy is only the fifth most subsidized commodity at $8B, behind corn at $48B.
Dean Foods is the whole argument in one company. The largest milk processor in America ran margins of 28% in 2009 and rode them down to 20% by 2019 as retailers used private-label milk as a loss leader, with revenue falling $11.0B to $7.3B and losses at the end. It owned Horizon, TruMoo, DairyPure and Land O'Lakes milk, and it went bankrupt anyway — which is what happens to a processor when the subsidy stops at the farm gate.
Common questions
Why did Oatly fail?
Over-expansion into a ceiling. Oatly committed roughly half a billion dollars a year to building its own factories on five continents on the assumption that 20–30% annual growth would continue, then growth fell to low single digits by 2023. It also excluded distribution costs from cost of sales, which inflated its reported gross margin — add them back and 2022 gross margin was 2.6% rather than 11%. Sales and marketing ran 40–55% of revenue because once PepsiCo, Coca-Cola and Nestlé shipped their own oat milks, Oatly had to justify being the most expensive brand in a commoditized category. Operating losses went from $47 million in 2020 to $207 million, $352 million and $344 million, and the shares fell from $24.46 to $0.66.
Is it true that plant-based milk is unfairly disadvantaged by subsidies?
Partly. Cow milk is propped up by school-lunch mandates, minimum price supports, the Milk Income Loss Contract, Dairy Margin Coverage and taxpayer-funded promotion through Dairy Management Inc. Oat milk gets none of that, which is why it can only ever be a premium and never the floor. But the subsidies flow to farmers, not processors, and they are awarded on acreage and production — the top 20% of farms collected $347 billion from 1995 to 2023 against $40 billion for the other 80%. Oatly buys oats, oil and enzymes and processes them, which makes it a processor. Handing it every dairy subsidy would not change the economics of the position it occupies.
Why is milk such a bad business?
Because its price is political rather than economic. Consumers watch the price of milk more closely than bread, eggs or meat, so supermarkets sell it at or below cost as a loss leader and place it at the back of the store. Politicians need it cheap, retailers need it cheap, and the pressure lands on farmers and processors. Milk is also perishable and needs refrigeration, so it has to move fast, and cows produce year-round rather than on a harvest cycle, so supply cannot be throttled. Excluding subsidies, the average American dairy farm has been profitable in only 3 of the last 23 years.
What happened to Dean Foods?
It bet on consolidation and got commoditized. As the largest dairy processor in the US it expected scale to deliver a cost advantage no rival could match, and through the 2000s it ran the best margins in the industry. Then the Great Recession made retailers desperate for foot traffic, and they found it in private-label milk sold at a loss. The gap between Dean's branded milk and store-brand milk widened every year while Dean still had to pay farmers the government-set floor price. Revenue fell from $11.0 billion in 2009 to $7.3 billion in 2019 and margins fell from 28% to 20%. Horizon, TruMoo, DairyPure and Land O'Lakes were not enough, and the company filed for bankruptcy in 2019.
Who actually makes money in milk?
Conglomerates and farmer co-ops. The largest dairy companies in the world are Dairy Farmers of America at $25 billion, Land O'Lakes at $19 billion, Saputo and Nestlé at $11 billion each, Lactalis at $10 billion, Danone at $7 billion and Prairie Farms at $4 billion — and three of those are co-ops owned by the farmers themselves, whose economics rest on collecting subsidies rather than on selling another gallon. For everyone else, a milk brand only works as one line in a large processed-food portfolio, carried for shelf coverage rather than for profit. That is why Silk ended up inside Danone.
Did Silk do any better than Oatly?
Slightly, and it still could not command a premium. As the first nationwide soy, almond and cashew milk brand, Silk grew revenue from $1,821 million in 2010 to $4,198 million in 2016 at operating margins of 6.9%, 8.7%, 7.9%, 6.2%, 7.8%, 8.6% and 9.6% — better than Oatly ever managed, but below the dairy companies it competed with, and this was during a period when retailers were too busy fighting over private-label cow milk to bother with plant-based store brands. Danone acquired Silk in 2016 and it stopped reporting separately.
Discussion
The vegan defense is that cow milk is subsidized everywhere and oat milk is not, so plant-based brands must build demand, build supply and match an artificially cheap price at once. The first half is true; the case says the conclusion is wrong. Where exactly does the argument break?
No answers yet — be the firstOatly grew revenue from $30 million to $783 million while operating income went from $1 million to a $344 million loss. What kind of growth is worth having, and how would you have known the difference at $200 million?
No answers yet — be the firstIt committed half a billion dollars a year to owning factories on every continent on the assumption that 20–30% growth would hold. When is vertical integration the right call, and what should trigger you to stop?
No answers yet — be the firstSubsidy makes the incumbent product artificially cheap. If you cannot change the policy, what are your actual options as a challenger?
No answers yet — be the firstYou are running Oatly in its best year. What do you do differently, knowing what the growth assumption is about to do?
No answers yet — be the first
Related case studies
Food & beverageWhy Taco Bell had to get expensiveThe most profitable chain in fast food is being squeezed hardest, and the reason is KFC and Pizza Hut.8 min read
Food & beverageWhy your favorite brands got so much worseRecipes got cheaper and portions smaller across American brands at the same moment. Not a coincidence.8 min read
Food & beverageWhy Crumbl won't lastMrs. Fields had its decade too. Crumbl store margins fell from 30% to 10% while corporate went from $4M to $152M.12 min read