Why nobody wants to own ice cream anymore

Ben & Jerry's was never a good business on its own, which is the part the brand story leaves out. As an independent through the 1990s its operating margin ran 7%, 9%, 9%, −3%, 6%, 4%, 4%, 5% and 2% — through the exact decade its cultural profile peaked. Popularity and profitability were never the same problem.

Twenty years later the top line has not moved. $8,422M, $8,127M, $7,623M, $7,045M, $6,947M, $8,284M, $8,386M through 2007, then $7,194M, $7,521M, $8,611M, $8,611M by 2023. Roughly 4% of growth across two decades in a category everybody on earth buys.

Set against the other divisions it is not close. In 2023 Unilever ran 10% in ice cream against 19% in beauty, 20% in personal care, 12% in home care and 18% in nutrition — and in 2022 nutrition hit 32% while ice cream managed 10%. Every euro invested in ice cream was a euro not invested in Dove, Axe, Knorr or Hellmann's.

Management blamed the weather for twenty years. Weak summers, cold seasons, the rising cost of sugar, dairy and cocoa — every year, the same explanations, on a sample size long enough to rule all of them out. Unilever committed to selling the entire ice cream business by 2025 and has struggled to find a buyer.

Its margins did improve — through shrinkflation. Nestlé's ice cream margin ran 11%, 12%, 11%, 12%, 12%, 11%, 12% and then stepped to 14%, 14%, 15%, 15% from 2010. The step change is 2009, the year it cut every Häagen-Dazs pint from 16 ounces to 14 and charged the same price for it.

Then it left. In 2019 Nestlé sold the division to Parisian private equity at a 20% premium for $4 billion, citing a desire to focus on high-growth verticals. Owning the second-best-selling ice cream brand in America did not qualify.

Kroger had owned Turkey Hill for four decades and it was the seventh best-selling brand in the country. The largest supermarket chain in America, with distribution and shelf space nobody else had, decided the ice cream business was not worth keeping either.

The pure-play vehicle was supposed to prove the conglomerates wrong. Froneri does nothing but ice cream and it has grown the top line — $2,631M, $3,973M, $4,240M, $5,074M, $5,292M — entirely through continuous acquisition, with no change in volume.

And the debt eats all of it. Froneri's operating income against its financing costs: $214M against −$193M, $276M against −$383M, $233M against −$392M, $413M against −$448M, $530M against −$586M. It has borrowed so heavily to fund the acquisitions that every dollar of operating profit and more goes to interest. The private equity partner is now looking for its own way out.

The franchise disclosures explain why nobody is opening one. It costs $584,000 to open a Cold Stone and the median store grosses $531,000 a year after royalties — against $1,386,549 at Dairy Queen, $794,620 at Handel's and $462,121 at Baskin Robbins. You spend more than a year's revenue to buy a year's revenue.

Baskin Robbins only survives by riding another brand. Its owner also owns Dunkin', and bundling the scoop shop into the donut chain is the only way it has held US distribution — 2,763 points in 2007 against 2,524 in 2019. It has fewer American shops now than a decade ago.

None of which has stopped private equity. Van Leeuwen has taken $25M, Oberweis $20M, Ample Hills $19M, Jeni's $15M and Salt & Straw $4M to scale production, blanket retailers with pints and open scoop shops — all on the bet that one becomes the Häagen-Dazs of the next generation.

The structural problem is that they cannot even reach the ceiling the conglomerates hit. These are single-product businesses with no shared manufacturing, no cross-sell, no distribution leverage and no adjacent categories — the exact things that got Unilever and Nestlé to double digits in the first place. It has happened before: Ample Hills, Milkmade and Phin & Phebes all imploded under outside capital in the 2010s Brooklyn wave.

The shop is barely the business. It grosses $15,000 a month in summer and as little as $4,000 in the other three seasons — $81,000 a year against $36,000 of rent. The money is at street fairs: over a hundred of them across the Northeast, 50,000 people a night, ten orders a minute, $2,000 in an evening against $500 on the shop's best day. Fairs bring $120,000 a year, for $201,000 combined.

His customers are chefs. He is the exclusive supplier to nearly 400 Manhattan restaurants including Minetta Tavern and Porterhouse, who buy in gallons and turn it into their own desserts sold year-round — which removes the seasonality that defines everybody else in this business. The average wholesale customer spends $550 a month; the average retail customer spends $12 a visit.

Which is why the one number that breaks the pattern is a wholesale one. A restaurant spending $550 a month beats a walk-in spending $12 by a factor of forty-five, all year, without a summer. Jon's moat is not a recipe — it is that he does every job himself, keeps inventory in a shorthand only he can read, and offers free delivery, no minimum order and midnight cut-offs that nobody with a normal org chart could match.