The reason a long queue is an advertisement

A poster says the food is excellent. Forty people standing outside say nothing at all, and the second statement is believed more. Neither group has inspected the kitchen. The difference is in who is speaking and what it would cost them to be wrong.

What everyone sees

Both are read as evidence of quality, with the queue simply feeling more convincing. The reason for the asymmetry is usually left unexamined — attributed to trusting people over posters. The structural point is sharper: one message comes from a party with an interest in it, and the other does not.

What is actually happening

Cheap talk from an interested party carries little information, which is why advertising must be costly to signal anything, as Nelson argued. A queue is different in kind: it is a record of decisions taken by parties with no stake in persuading a stranger, and each of them paid in time. Third-party evidence of this sort is treated as more diagnostic, though it reports what others chose rather than whether they were satisfied.

Why it stays hidden

The gap between the two hides because the queue overstates what it can carry. It reports arrival decisions, and those decisions may themselves have been copied from earlier arrivals rather than formed independently. A queue can therefore be simultaneously the most credible signal available on the street and a poor guide to the food — credible about demand, silent about quality.