Why buffets went extinct

Unlimited had worked for decades — as an add-on. Olive Garden gave away breadsticks, Pizza Hut ran a lunch buffet to fill dead hours, Ponderosa piled on self-service sides to sell beef at full price. Every time the industry tried it, it worked, and every time it sat beside something that made money. The leap of the 1980s was removing the something.

The buffet did not create its customer, the economy did. Wages had been flat since the mid-1970s, manufacturing peaked in 1979 and never came back, and by 1990 more than half of women had jobs. Households covered the difference on credit, taking debt from 60% of income to over 90%. Dinner had to be fast, hot, filling and cheap enough for a family. The buffet was that, with the price on the sign.

The whole bet was that scale would compound. Old Country went one location to 200 in ten years, doubled again in five, and at its peak opened more than one a week. The 1996 merger with HomeTown made a near billion-dollar operator at 17% restaurant-level margins — ahead of Wendy's, level with Outback. It was McDonald's playbook, and nobody questioned it.

Then the restaurants started shrinking and nobody could say why. Same-store sales went negative in 1994 and stayed red five straight years; after inflation, weekly sales fell every year from 1996. Nothing had changed — 40% of the estate was new or remodeled, the price rises were marginal. Expansion never outran the decline. It just built more of it.

Advertising stopped being marketing and became life support. Cut to under 1% of sales in 1995 to lift profit, it produced the worst same-store sales in company history the next year. Only tripling the budget turned them positive again. The chain had to buy its own traffic, and hand over the margin to get it.

Labor rose six years running even as staffing was cut. The only cost that improved was food, and only because meat and produce got cheaper — 34.7% of sales down to 31.6%. When your one working lever is a commodity cycle you do not control, you are not managing the business. You are waiting on it.

Private equity found a cash machine and drained it. On $90 million a year and almost no debt, Buffets Inc. went private in 2000 on over half a billion borrowed. Traffic never improved in six years. The owners raised checks 20%, took $90 million in dividends, and paid for Ryan's by selling the land under the restaurants — turning the one asset that did not depend on traffic into rent.

Fresh Choice had the cheapest food in the industry and it was not enough. Selling salad instead of meat held food cost ten points below Old Country and HomeTown, and that gap was the entire margin. Customers still would not pay more without an entree. National expansion cannibalized itself until 90% of stores were back in California, and it filed first, in 2004.

Sweet Tomatoes did everything right and it only bought time. It owned its land or signed 30-year leases, ran 19 central kitchens, and posted the best margins in the category through the 1990s. Traffic collapsed anyway after 2001. Private equity sold the land, and with no ground to hibernate on the chain died within weeks of the pandemic.

The model runs on a hedge, and the hedge dies with the traffic. A buffet earns its margin on light eaters and gives it back to heavy ones, and the law of averages carries the room. Narrow the customer base and what remains is disproportionately the people who eat the most. Falling traffic does not lower food cost. It raises it, per head.

Waste is not a failure of the format. It is the format. Abundance has to be visible, so the trays stay full until closing — which means cooking food you know nobody will eat and binning it at the end of the night. Every other restaurant treats waste as an error to engineer out. Here it is the cost of the thing being sold.

There was no ladder out, so the turnover was structural. Heating pans, filling trays, bussing and washing transfer only to institutional kitchens — hospitals, prisons, catering — where the pay and hours are worse. Even good wages would not have stopped the churn, and wages were the one thing these chains could never afford.

What took the demand was not a restaurant. It was the deli counter twenty feet away. Supermarkets matched the buffet on value and beat it on convenience, because the deli never had to carry the store: the customer already came every week, labor and overhead were shared, the hours were already long, and unsold rotisserie chicken became chicken salad instead of going in the dumpster.

Fast casual then took the top. The family that piled plates at Hometown buys burrito bowls at Chipotle; the girl who went to Fresh Choice for salad upgraded to Sweetgreen. Fast casual was built for takeout, and a buffet only makes money when you get full on the premises. There was never a world where it could box the food up or deliver it.

Every restaurant reaches profit four ways — raise the price, shrink the portion, cut the labor, shrink the room. McDonald's pushes cost onto franchisees and kiosks. Cheesecake Factory stays small and charges more. Shake Shack engineers out the staff. Wendy's cut too deep and lost customers faster than the checks could cover. The buffet had none of the four: variety needs the square footage, so a chain that started big had to stay big, locked into 10,000 square foot boxes as the malls emptied. Chapter 11 was the only way out of the leases.

Golden Corral survived by selling the format instead of operating it. $50,000 a franchisee, a 4% royalty, equipment and development sold at a markup, and three company-owned restaurants left. It runs the highest food cost in the category on purpose, which is why the average store takes $3.6 million. It lost a quarter of its restaurants through the pandemic without ever appearing in court, because the leases belonged to somebody else — and it was recruiting replacements inside the year.