Table Turn

Turns are how operators talk about throughput, and the number describes the room rather than the guests in it. Two turns on a Friday and three on the same covers are different nights for the floor, the kitchen, and the payroll. Turns are also where guest records are most likely to get attached to the wrong person.

What is the difference between a table turn and a cover?

A turn is a seating cycle at a table and a cover is a guest, so the same volume produces different turn counts. One hundred covers might be twenty-five turns of four-tops or fifty turns of two-tops, and the room that did fifty worked much harder for the same guests.

Turns are also quoted two different ways, which causes real confusion when operators compare notes. One is seatings divided by tables across a service, giving “two and a half turns.” The other is turn time, meaning the minutes one cycle takes, so “a sixty-minute turn.” Someone saying “we do three turns” could mean either, so ask which before treating two venues’ figures as comparable.

One detail is left out of both versions. A turn is not finished when the party leaves but when the table is reset and sellable, so a room with thin bussing coverage loses turns that never appear in any dwell-time figure.

Why does a high-turn room make guest matching harder?

Because table and time stop being a unique key. Matching a check to a reservation relies on both records agreeing about which table, at roughly what hour. A table hosting one long seating in an evening matches easily. A table hosting three parties two hours apart produces three checks that all look plausible against a booking at that table, and the candidates get harder to separate when a party runs long, a check opens late, or a bar tab transfers into the dining room.

So attribution errors concentrate in the rooms that turn fastest: bistros, bars, and high-volume neighborhood places with tight pacing. The fix is a tighter key rather than a better guess, meaning normalized table names across both systems, a matching window sized to how long parties actually sit, and a preference for the check’s open time over its close time. A tasting-menu room can match on almost anything. A room turning three times a night needs the join configured to how it actually runs.

Why did my turns drop?

Turn counts fall for floor reasons far more often than for demand reasons. Dwell time drifts as menus get longer or more dishes are meant for sharing. Bussing hours get trimmed, so tables sit dirty for eight minutes each. Average party size grows, which raises covers while lowering seatings. Pacing stacks arrivals into a wave the kitchen answers slowly, pushing every subsequent seating back. Late-service campers cost the last turn outright.

The useful comparison is the same service period in the restaurant’s own history, since turns are not comparable across service models.