Why arcades never actually died

The golden age was real and it was brief. The American arcade market went from nothing in 1976 to $7 billion by 1980 and $10 billion in 1982, then fell to $5 billion, $4 billion, and bounced between $2 billion and $9 billion through 1992. Not every game was a hit and genres went out of fashion fast.

The hardware was the business for the people who made it. A Pac-Man cabinet cost an operator $2,400 in 1982 — about $7,841 in today's money — a Street Fighter II cabinet $3,000 in 1991, and a Mortal Kombat II cabinet $4,000 in 1993, or $8,728 adjusted. Sega, Namco, Atari and Capcom competed on 3D, light guns, force feedback and co-op precisely because more elaborate hardware justified a higher price.

Then consoles took the spend and never gave it back. Arcades were worth $7 billion against $8 billion of consoles in 1997; by 2002 it was $2 billion against $12 billion, and by 2013 $2 billion against $15 billion. A home game does not need to be designed to milk quarters, which makes it a better game.

Chuck E. Cheese did not start as an arcade — it started as a pizza chain, and it was losing. Revenue stalled at $253 million, $272 million, $268 million and $263 million through the early 1990s while operating income collapsed from $26 million to $18 million, $3 million and finally $0. It could not beat Pizza Hut or Domino's on food, franchising or scale.

So it stopped trying. Every location was remodelled around a token-operated playroom aimed at 2-to-12-year-olds, and revenue went from $263 million to $294 million, $350 million, $441 million, $562 million and $728 million by 2004 — with operating income climbing from $0 to $134 million over the same stretch.

And the reason is margin, not preference. Games run gross margins of 85%, 91%, 93% and 92% against 79%, 77%, 75% and 77% for food. A token machine is a self-service asset with no ingredients, no kitchen labor and no spoilage — the closest thing to free money a restaurant can put on its floor.

The pivot worked on the bottom line for a decade. Operating margin went from 0% in 1995 to 7%, 12%, 15%, 17% and 18% by 2001, and the chain expanded from 314 locations to 519.

Then it stopped. Margins ran 18%, 16%, 14%, 13%, 13%, 13%, 12% and 10% through 2011, and average revenue per store has been essentially flat since 1997 — $1,634,000, $1,633,000, $1,632,000, $1,573,000. The chain got bigger; the stores did not.

Dave & Buster's is the same insight aimed at adults, and it scaled better because the room is bigger. Its average venue runs 44,000 square feet against 11,000 for Chuck E. Cheese — with Topgolf at 60,000, Round1 at 50,000 and Lucky Strike at 32,000. A bigger box holds more machines and more occasions.

Its margin split is even starker than Chuck E. Cheese's. Arcade and entertainment gross margins run 86%, 85%, 89% and 91% against 75%, 76%, 74% and 71% for food and drink. The restaurant exists to justify the visit; the midway pays for the building.

But it hits the same ceiling. Same-store sales ran 10%, 6%, −3%, 4%, −3%, −5%, −0% and 2% in its first growth era, then 6%, 4%, −3%, −8%, 7%, 9%, 3%, −1% and −2% by 2019 — negative well before COVID. Unless there is a birthday, an outing or a Super Bowl, nobody comes.

That is the structural problem with both: they are occasion businesses evaluated as a sum of parts, and every part loses on its own. The food does not beat a restaurant, the games do not beat a dedicated arcade, and the drinks do not beat a bar. Once the novelty of any new format fades, both chains fall back on discounting.

The independents are where the margins are. Annual revenue runs $9,630,000 at an average Dave & Buster's and $1,617,000 at an average Chuck E. Cheese — against $2,100,000 for OS in Manhattan, $1,082,105 for a haunted house, $351,000 for another, and $330,000 for a two-man game studio.

Invert that list for operating margin and it flips completely: 70% at the Slaughterhouse haunted house, 55% at Game of 1000 Boxes, 50% at Nightmare, 31% at Disney Parks, 13% at OS, 13% at Dave & Buster's and 6% at Chuck E. Cheese. Being small is the advantage, not the handicap.

Game of 1000 Boxes is the opposite bet: two people in Brooklyn who fabricate the hardware and write the software themselves, with no investors. Forty players, teams of four, a buzzer each and a live host — and rather than operate venues, they license the system to anyone who has one.

So the industry did not die, it changed hands. The turnkey era — buy a cabinet, collect the quarters — is genuinely over, because a console does that better. What replaced it is experiences a console cannot copy: a room full of strangers, a live host, a physical touch point for a friendship that already exists online.