How to Improve Table Turnover Without Rushing Guests

Turnover is a capacity problem, not a speed problem

When a full restaurant turns away business, the instinct is to look at the room and conclude there are not enough seats. Usually there are plenty of seats; they are just occupied by tables that finished eating fourteen minutes ago, or sitting clean and empty because nobody told the host they were ready.

Table turnover is the count of how many times you sell the same seat in a service. It is the only revenue lever that does not require raising prices, adding covers to the building, or spending money to attract new guests. The demand is already at the door on your busiest nights. Turnover decides how much of it the room can physically take.

The arithmetic is more generous than most operators expect, because table time works as a ratio. A 90-seat room doing 152 covers at 82 minutes per table, with a $38 average spend, gains about ten covers and $375 by moving to 77 minutes. Twelve services a week makes that roughly $234,000 a year, and nearly all of it is margin because the rent and the fixed labour do not move.

Five minutes is not a demanding target. In most dining rooms there are twenty to thirty minutes per table that have nothing to do with eating.

Measure table time, not dining time

Before changing anything, get the number right. The figure that governs your capacity is not how long the guest sits. It is how long the table is unavailable, from the moment the previous party is seated to the moment the next one is.

That includes the turnaround: clear, wipe, reset, reassign. POS reports give you seated-to-paid and stop there, which means the number most operators quote is eight to fourteen minutes short of reality. On the nights where turnover matters, the gap is at its widest, because that is exactly when resets go unowned.

Timing it is a clipboard job, not a software project. Write the seated time on the docket, note the reset time at the pass, and do it for twenty tables across two busy services. Then break the total into four stages:

• Seated to first order taken
• First order to first course on the table
• Last course cleared to bill requested
• Bill requested to table reset and reassigned

The value of splitting it up is that the recoverable time is almost never evenly spread. One or two stages will hold most of your minutes, and they will not be the ones you expected. Guests eat at more or less the same pace everywhere; what varies wildly between restaurants is everything around the eating.

The four places minutes actually hide

The greeting gap

A party is seated and then waits six or seven minutes to be acknowledged because their server is buried in another table's mains. This costs you twice: it extends the table, and it delays the drinks order to the point where a second round becomes unlikely. Acknowledgement within sixty seconds — even without taking the order — starts the meal on time and reads as attentive rather than hurried.

Two or three slow dishes

A single long ticket item holds the entire table, including everyone who ordered something quick. It is worth pulling ticket times by dish and asking whether the slowest ones can be re-engineered so the long step happens in prep rather than at order. Ticket time belongs in your menu engineering alongside margin: a high-contribution dish that adds nine minutes to every table it lands on is not the star it appears to be.

Payment friction

Bill requested, bill printed, bill delivered, card machine located, payment taken, receipt returned. Six steps, four of which are waits, and twelve minutes is common. This is the single most recoverable block in the meal and the one guests are most grateful to lose. Handheld terminals or a scan-to-pay option remove it almost entirely, and no guest has ever experienced paying promptly as being rushed out.

The unowned reset

Nobody specifically owns the table, so it is cleared eventually, wiped when someone passes, and reset when there is cutlery in the station. Assigning tables to named people and stocking the side stations before service turns a fourteen-minute turnaround into four. Then close the loop with the door: a clean table the host does not know about is worth exactly nothing.

The line between efficient and rushed

Everything above removes waiting. None of it removes time from the guest's meal, and that distinction is the whole game. Guests resent being hurried, but they resent waiting considerably more, and the two feel completely different even when they cost the same number of minutes.

The interventions that cross the line are recognisable and worth naming, because they are also the tempting ones:

• Clearing plates before everyone has finished. Reads as a signal to leave and is remembered long after the meal.
• Dropping the bill unasked. Saves two minutes and can cost you the dessert and the coffee, which are usually the highest-margin items on the check.
• Hard time limits announced at seating. Workable for a genuinely in-demand concept with clear communication at booking; corrosive when applied to a room where guests expect to relax.
• Dimming lights or stacking chairs to move people on. Guests notice, and the online review is more expensive than the turn.

There is also a revenue trap. Compressing the window between the last course and the bill is the fastest way to reduce table time, and it is precisely where dessert, digestifs, and the second bottle live. You can shorten the table and shrink the check at the same time, which is a poor trade. Track average spend per cover alongside turns whenever you change something, and stop if the two move in opposite directions.

Fix speed where you are actually full

Extra capacity only becomes revenue when there is unmet demand at that exact moment. Turning tables faster at 6pm on a Tuesday produces a quieter room, not a busier one, and the effort would have been better spent on filling it.

So separate your service into peak and shoulder before doing anything. A dinner service averaging 1.7 turns is often running 2.4 between 7pm and 9pm and 0.6 either side. Those two halves have opposite problems: one needs minutes, the other needs guests. Averaging them hides both.

Then check your reservation grid against your real table time. Booking on 90-minute intervals when your measured table time is 82 minutes builds eight minutes of guaranteed idle into every slot by design — a scheduling decision, not an operational failure, and one that costs a full turn across a long service.

What to check after the change

Faster turns move more than one number, and not all of them in the same direction. Three worth watching for a full month after any change:

• Average spend per cover. If it falls while turns rise, you compressed the wrong stage and traded margin for volume.
• Labour cost percentage. More covers across the same staffed hours should lower it. If you hired a runner to buy the minutes, recalculate honestly — the labour cost formula will tell you whether the extra wage was covered by the extra covers.
• Contribution against fixed costs. Additional covers on an unchanged cost base move your break-even point more than most operators expect, because the incremental cost is only food and variable wages.

If you want the full picture of how the labour side responds, the labour cost formula guide covers how to attribute the hours properly rather than dividing a weekly payroll by weekly sales and hoping.

The goal is not a faster restaurant. It is a restaurant where guests spend their time eating and drinking rather than waiting for someone to notice them, bring the machine, or clear the table they are no longer sitting at. Do that and the turns follow on their own.

Frequently asked questions

How long should a table take in a casual dining restaurant?

Most casual dinner tables run 75 to 95 minutes from seating to reset, and lunch runs 45 to 60. What matters more than the absolute figure is that you measure it including the turnaround, because that is the part reservation grids are built on.

Does bringing the card machine to the table really matter?

It is usually the largest single saving available. The payment sequence commonly costs ten to twelve minutes across four separate waits, and removing it is the one speed improvement guests actively appreciate rather than tolerate.

Should I use timed seatings?

Only if demand genuinely exceeds capacity and you communicate the limit at booking rather than at the table. Applied to a room that is not consistently full, it costs goodwill without gaining covers.

Will faster turns reduce average spend per cover?

It depends where the minutes come from. Speed taken from resets, greetings, and payment has no effect on spend. Speed taken from the window after the last course removes the dessert and second-drink opportunity and will show up as a lower check.

How do I get the kitchen involved in turnover?

Give them ticket times by dish rather than a general instruction to be quicker. Two or three items usually account for most of the delay, and reworking those so the slow step moves into prep is a specific, solvable task.

Do guests who linger after paying count against my turn time?

Yes, because the table is still unavailable. Handle it with the room rather than the guest — offering to move a lingering party to the bar frees the table without asking anyone to leave, and it often adds a drink to the check.

Is it better to add seats or turn faster?

Turning faster, almost always. Extra seats add fit-out cost, more staffed sections, and often a slower kitchen at peak. Recovering five minutes per table costs nothing and delivers a similar capacity gain.

Run the numbers

Use the free Table Turnover Rate Calculator to apply everything above to your own figures.