Modern MBA

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

$24.46 → $0.66Oatly share price from its 2021 IPO to the end of 2024
3 of 23Years the average US dairy farm made money without subsidies
$347B vs $40BSubsidies to the top 20% of farms versus the other 80%

By the numbers — swipe or use arrows

01Five years up, five years goneValuation in billions of dollars. Impossible Foods ran the same arc — $5B in 2019, $8B in 2020, $4B in 2021.
Five years up, five years gone — Why the dairy industry is failing$0.0B$2.0B$4.0B$6.0B$8.0B$10.0B$12.0B$0.2B$2.0B2019$2.0B$4.0B2020$10.0B$7.0B2021$0.4B$1.4B2024OATLYBEYOND MEATModern MBA
View data
OatlyBeyond Meat
2019$0.2B$2.0B
2020$2.0B$4.0B
2021$10.0B$7.0B
2024$0.4B$1.4B
02Oatly is a penny stockAdjusted closing price per share since the June 2021 IPO. There has been no recovery in any half-year since.
Oatly is a penny stock — Why the dairy industry is failing$0.00$10.00$20.00$30.00$24.46Jun2021$7.96Dec2021$3.46Jun2022$1.74Dec2022$2.05Jun2023$1.18Dec2023$0.94Jun2024$0.66Dec2024Modern MBA
View data
Value
Jun 2021$24.46
Dec 2021$7.96
Jun 2022$3.46
Dec 2022$1.74
Jun 2023$2.05
Dec 2023$1.18
Jun 2024$0.94
Dec 2024$0.66
03Nobody in plant-based makes moneyOperating margins of the two listed leaders. Beyond Meat got worse every single year.
Nobody in plant-based makes money — Why the dairy industry is failing−120%−100%−80%−60%−40%−20%0%−14%−13%2020−33%−39%2021−55%−83%2022−49%−97%2023OATLYBEYOND MEATModern MBA
View data
OatlyBeyond Meat
2020−14%−13%
2021−33%−39%
2022−55%−83%
2023−49%−97%
04The vegan tax on milkAverage nationwide price per gallon, 2024. Oat milk costs 70% more than the product it is replacing.
The vegan tax on milk — Why the dairy industry is failing$0.00$2.50$5.00$7.50$4.10Cow milk$6.56Almond milk$6.97Oat milkModern MBA
View data
Value
Cow milk$4.10
Almond milk$6.56
Oat milk$6.97
05And on meatAverage nationwide price per pound, 2024. Impossible costs almost twice what ground beef does.
And on meat — Why the dairy industry is failing$0.00$2.50$5.00$7.50$10.00$5.35Ground beef$6.70Beyond Meat$9.32ImpossibleModern MBA
View data
Value
Ground beef$5.35
Beyond Meat$6.70
Impossible$9.32
06Revenue grew 26x and the losses grew fasterOatly revenue and operating income in millions of dollars. The losses arrive exactly when the factories do.
Revenue grew 26x and the losses grew faster — Why the dairy industry is failing−$500M$0M$500M$1,000M$30M$1M2014$39M$0M2015$43M$0M2016$58M−$1M2017$118M−$10M2018$206M−$31M2019$421M−$47M2020$643M−$207M2021$722M−$352M2022$783M−$344M2023REVENUEOPERATING INCOMEModern MBA
View data
RevenueOperating income
2014$30M$1M
2015$39M$0M
2016$43M$0M
2017$58M−$1M
2018$118M−$10M
2019$206M−$31M
2020$421M−$47M
2021$643M−$207M
2022$722M−$352M
2023$783M−$344M
07Half the company is sales and adminOatly revenue against SG&A spend, in millions of dollars. It never grew into the overhead it built.
Half the company is sales and admin — Why the dairy industry is failing$0M$250M$500M$750M$1,000M$206M$93M2019$421M$167M2020$643M$354M2021$722M$412M2022$783M$373M2023REVENUESG&AModern MBA
View data
RevenueSG&A
2019$206M$93M
2020$421M$167M
2021$643M$354M
2022$722M$412M
2023$783M$373M
08It won the shelf and still lost moneyOatly sales by channel. Two thirds is retail — people asking for it by name and buying it by the box.
It won the shelf and still lost money — Why the dairy industry is failing0%25%50%75%65%35%202164%36%202265%35%2023RETAILFOODSERVICEModern MBA
View data
RetailFoodservice
202165%35%
202264%36%
202365%35%
09Milk is a worse business than every other drinkTrailing-twelve-month margins for the global dairy companies against the non-dairy beverage giants. Perishable and essential against shelf-stable and optional.
Milk is a worse business than every other drink — Why the dairy industry is failing0%10%20%30%40%50%35%10%Global dairy47%25%Global non-dairyGROSS MARGINOPERATING MARGINModern MBA
View data
Gross marginOperating margin
Global dairy35%10%
Global non-dairy47%25%
10Oatly was worse than the industry it was disruptingThe same comparison with Oatly added. It makes its product from oats, water, oil and enzymes — no cows required.
Oatly was worse than the industry it was disrupting — Why the dairy industry is failing−50%−25%0%25%50%35%10%Global dairy47%25%Global non-dairy19%−44%OatlyGROSS MARGINOPERATING MARGINModern MBA
View data
Gross marginOperating margin
Global dairy35%10%
Global non-dairy47%25%
Oatly19%−44%
11The gross margin was cookedOatly excluded customer distribution from cost of sales, breaking beverage convention. Adding it back gives the real number.
The gross margin was cooked — Why the dairy industry is failing0.0%10.0%20.0%30.0%24.0%16.4%202111.0%2.6%202219.0%12.8%2023AS REPORTEDWITH DISTRIBUTION ADDED BACKModern MBA
View data
As reportedWith distribution added back
202124.0%16.4%
202211.0%2.6%
202319.0%12.8%
12It spends like a fashion brandOatly SG&A as a percentage of revenue against the industry. Once everyone sold oat milk, the marketing had to sell Oatly specifically.
It spends like a fashion brand — Why the dairy industry is failing0%20%40%60%40%202055%202152%202253%202327%Global dairy26%Global non-dairyModern MBA
View data
Value
202040%
202155%
202252%
202353%
Global dairy27%
Global non-dairy26%
13Twenty years and milk barely movedAverage nationwide retail price via the Bureau of Labor Statistics. Chicken doubled. Milk did not, because retailers price it as a loss leader.
Twenty years and milk barely moved — Why the dairy industry is failing$0.00$1.00$2.00$3.00$4.00$5.00$3.23$1.032004$3.00$1.062006$3.68$1.312008$3.32$1.282010$3.58$1.482012$3.82$1.542014$3.29$1.462016$3.54$1.622020$4.22$1.832022$4.10$2.062024MILK, PER GALLONCHICKEN, PER POUNDModern MBA
View data
Milk, per gallonChicken, per pound
2004$3.23$1.03
2006$3.00$1.06
2008$3.68$1.31
2010$3.32$1.28
2012$3.58$1.48
2014$3.82$1.54
2016$3.29$1.46
2020$3.54$1.62
2022$4.22$1.83
2024$4.10$2.06
14A century of promotion and the decline never stoppedMilk consumption per capita in the United States, in pounds. School mandates, Got Milk and fast-food cheese deals did not reverse a single decade.
A century of promotion and the decline never stopped — Why the dairy industry is failing0 lbs100 lbs200 lbs300 lbs400 lbs335 lbs1950275 lbs1970235 lbs1990215 lbs2000181 lbs2008177 lbs2010169 lbs2012158 lbs2014153 lbs2016145 lbs2018141 lbs2020130 lbs2022Modern MBA
View data
Value
1950335 lbs
1970275 lbs
1990235 lbs
2000215 lbs
2008181 lbs
2010177 lbs
2012169 lbs
2014158 lbs
2016153 lbs
2018145 lbs
2020141 lbs
2022130 lbs
15Half the dairy farms in America are goneNumber of milk farms in the United States. Output hit records over the same decade — family operations replaced by corporate herds.
Half the dairy farms in America are gone — Why the dairy industry is failing010,00020,00030,00040,00050,00044,809201443,584201541,819201640,219201737,468201834,817201931,652202029,842202127,932202226,2902023Modern MBA
View data
Value
201444,809
201543,584
201641,819
201740,219
201837,468
201934,817
202031,652
202129,842
202227,932
202326,290
16Subsidies are awarded on size, not needUS farm subsidies awarded in billions of dollars, 1995 to 2023. The bottom 80% of farms split $40 billion between them.
Subsidies are awarded on size, not need — Why the dairy industry is failing$0B$100B$200B$300B$400B$101BTop 1%of farms$231BTop 5%of farms$296BTop 10%of farms$347BTop 20%of farms$40BBottom 80%of farmsModern MBA
View data
Value
Top 1% of farms$101B
Top 5% of farms$231B
Top 10% of farms$296B
Top 20% of farms$347B
Bottom 80% of farms$40B
17Without subsidies, milk loses moneyAverage profit per hundredweight sold, excluding subsidies, 2024. Cotton and peanuts are worse. Rice and soybeans are the exceptions.
Without subsidies, milk loses money — Why the dairy industry is failing−$300−$200−$100$0$100$200$300$247Rice$39Soybeans$12Corn−$4Milk−$76Wheat−$139Oats−$193Peanuts−$243CottonModern MBA
View data
Value
Rice$247
Soybeans$39
Corn$12
Milk−$4
Wheat−$76
Oats−$139
Peanuts−$193
Cotton−$243
18Dean Foods proved scale works, right up until it did notMargins at the largest milk processor in America through the 2000s. Consolidation delivered exactly what it promised.
Dean Foods proved scale works, right up until it did not — Why the dairy industry is failing0%10%20%30%40%26%200226%200324%200425%200525%200627%200723%200828%2009Modern MBA
View data
Value
200226%
200326%
200424%
200525%
200625%
200727%
200823%
200928%
19The biggest processor in the country went bankruptDean Foods revenue and operating income in billions of dollars, to the 2019 filing. It owned Horizon, TruMoo, DairyPure and Land O'Lakes milk.
The biggest processor in the country went bankrupt — Why the dairy industry is failing−$2.5B$0.0B$2.5B$5.0B$7.5B$10.0B$12.5B$11.0B$0.6B2009$10.8B$0.3B2010$9.7B−$2.0B2011$9.3B$0.3B2012$9.0B$0.1B2013$9.5B$0.1B2014$8.1B$0.1B2015$7.7B$0.3B2016$8.0B$0.1B2017$7.8B−$0.3B2018$7.3B−$0.4B2019REVENUEOPERATING INCOMEModern MBA
View data
RevenueOperating income
2009$11.0B$0.6B
2010$10.8B$0.3B
2011$9.7B−$2.0B
2012$9.3B$0.3B
2013$9.0B$0.1B
2014$9.5B$0.1B
2015$8.1B$0.1B
2016$7.7B$0.3B
2017$8.0B$0.1B
2018$7.8B−$0.3B
2019$7.3B−$0.4B
20The winners are conglomerates and farmer co-opsHighest-grossing dairy companies in the world, 2023, in billions of dollars. Three of these are owned by the farmers themselves.
The winners are conglomerates and farmer co-ops — Why the dairy industry is failing$0B$10B$20B$30B$25BDairy Farmers of America$19BLand O'Lakes$11BSaputo$11BNestlé$10BLactalis$7BDanone$4BPrairie FarmsModern MBA
View data
Value
Dairy Farmers of America$25B
Land O'Lakes$19B
Saputo$11B
Nestlé$11B
Lactalis$10B
Danone$7B
Prairie Farms$4B
01 / 20

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

01

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.

02

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.

03

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.

04

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.

05

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.

06

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.

07

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.

08

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.

09

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%.

10

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.

11

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.

12

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.

13

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.

14

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.

15

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.

16

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

  1. 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?

  2. Oatly 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?

  3. It 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?

  4. Subsidy makes the incumbent product artificially cheap. If you cannot change the policy, what are your actual options as a challenger?

  5. You are running Oatly in its best year. What do you do differently, knowing what the growth assumption is about to do?

Related case studies