Restaurant table turnover time: measure yours, ignore the averages

Why published averages mislead

You will find turnover averages quoted for fast casual, casual dining and fine dining. They are nearly useless for running your floor: they mix party sizes, dayparts, menu lengths and cultures. A taquería turning in 35 minutes and a steakhouse turning in two hours can both be perfectly run. The only benchmark that matters is your own floor last month, measured the same way every time.

Split the turn into three segments

A turn is seat-to-first-visit, the meal, and vacate-to-reset. Segment one is a service promise: water and a hello inside a couple of minutes. Segment two belongs to the guests, and the honest moves there are pacing the kitchen and bringing the check promptly when asked. Segment three is pure operations: how fast an empty, dirty table becomes a sellable one. When owners say they want faster turns, segments one and three are almost always where the recoverable minutes live.

How to measure without software

• Pick your two busiest nights and one quiet one, so you see both shapes of service.
• Log four timestamps per table: seated, first staff visit, guests left, table ready. A host with a phone timer can do it for a section.
• Separate by party size. Two-tops and six-tops live in different worlds, and mixing them hides the problem tables.
• Look at the spread, not just the average. One table that took 25 minutes to reset matters more than a tidy mean.

Two weeks of this gives you a baseline. It is tedious, which is exactly why cameras end up doing it better than clipboards.

What the number is worth: a worked example

This is a worked example, not a measured customer result. Take a 50-table floor, a $100 average check per table and a 60-minute average turn. One table-minute is then $100 ÷ 60 ≈ $1.67. If better greeting and resets trim 6 minutes off a turn, each turn that happens while guests are waiting frees about $10 of seating capacity. On a night where 40 tables turn once with a line at the door, that is roughly $400 of capacity, from minutes nobody was using. Your inputs will differ, which is the point: measure, then multiply.

How TableWave measures it for you

TableWave reads seated, waiting, vacated and reset events from the CCTV cameras you already own, on a computer inside your restaurant, so every table gets the four timestamps automatically, every night, split by party size. The 14-day payback report prints week one against week two with the formula on the page, so your accountant can check it. See pricing or book a 30-day pilot to get your own baseline measured.