How do servers remember a whole table’s order without writing it down?

Practice plan

Say you have a four-top ordering drinks, starters, and mains (illustrative). A routine built on the two findings above:

1. 1Pick a fixed starting seat. Always start from the same spot, say the seat nearest the door, and go around the table the same direction every time.
2. 2Tie each order to a seat. “Seat 3: salmon,” not “the woman in the green jacket.” Seat position is what the Buenos Aires waiters used.
3. 3Use the same order categories every time. Drink, starter, main, temperature, side. The same slots in the same sequence every table.
4. 4Repeat it back. Read the table’s order back seat by seat before you walk away.
5. 5Write it down when it’s big or complicated. Large parties, allergies, and lots of modifications are worth a pad. Your house rule comes first.

Not tax or legal advice. This routine is practice advice built on what the two studies describe, not something they tested.

Plain English

The waiter who took 20 orders without notes

In a 1988 study in the Journal of Experimental Psychology: Learning, Memory, and Cognition, Ericsson and Polson studied one waiter, called JC, who could take up to 20 complete dinner orders without taking notes. Here’s what the abstract reports:

• He used the same mental structure to store dinner orders even when the order of presentation was changed a lot.
• His method carried over almost completely to memorizing other material with the same category structure as dinner orders.
• His performance dropped when that category structure was taken away.
• The authors read the results as fitting skilled-memory theory, and as rejecting a theory based on chunking information in short-term memory.

In plain terms: the authors tie his skill to a consistent framework built around the parts of a dinner order, not to chunking in short-term memory.

Buenos Aires waiters and seat position

A 2008 study in Behavioural Neurology by Bekinschtein, Cardozo, and Manes tested Buenos Aires waiters, who memorize all orders without written support, on the job. Each waiter got eight different orders. In some trials customers stayed in their original seats; in others they changed seats. The match between orders, customers, and places dropped only when customers changed seats. The authors describe the waiters’ strategy as linking each client to a position at the table, and then to the drink.

What this doesn’t tell you

One study looked at a single waiter, and the other at waiters in one city. Neither tested a training method for new servers. If you’re still learning the menu itself, start with how to memorize a menu fast.