Restaurant floor plan and table turnover: seat more covers without rushing anyone

Take two restaurants with the same square footage, the same team and the same menu. The first serves 40 covers per service; the second serves 70, with guests just as happy. The difference is not decided in the kitchen or on the menu prices. It is decided on the floor plan, and in the way the tables turn.

A floor plan is not a drawing you make once before opening night and then forget. It is a management tool that determines your real capacity, your guests' comfort and how many miles your servers walk per shift. And table turnover, often treated as a slightly shameful topic ("we're not going to put a stopwatch on our guests"), is in fact a standard, widely accepted practice when it is handled the right way.

This field guide covers the whole subject: what a good floor plan looks like, turnover explained without corporate euphemisms, how to calculate your real capacity per service, and the trade-off between comfort and volume.

In short:

• a good floor plan favors two-tops and four-tops that combine, rather than large fixed tables that sit half empty;
• table turnover is managed through reservation slots and, when needed, a dining time announced at booking, never improvised tableside;
• your real capacity per service is the number of genuinely sellable seats multiplied by the number of turns your service window allows;
• squeezing in extra tables eventually costs more than it earns: past a certain density, the experience degrades and guests do not come back.

What a floor plan is actually for

On paper, a floor plan is a map of your tables, their capacities and their position in the room. In practice, it does three jobs:

1. Assigning reservations intelligently: a two-top should not go to a couple when the only tables left for the next requests are four-tops.
2. Seeing the service in real time: which tables are occupied, which ones free up soon, where to seat the walk-in couple who just came through the door.
3. Calculating your capacity: without an up-to-date floor plan, you do not know how many covers you can really accept on a given slot, so you either refuse guests out of caution or overbook out of optimism.

That is exactly what a clear reservation schedule gives you: every booking is assigned to a table, the schedule shows the room hour by hour, and you see at a glance what is still sellable. If you are still running the room from a paper book, our guide to the digital reservation book explains what you gain by going digital.

The principles of a good floor plan

Zones that follow the rhythm of the service

Divide your dining room into zones and open them progressively. On a quiet Tuesday lunch, concentrate guests in the zone nearest the kitchen: service is faster, the room feels lively instead of empty, and your team saves hundreds of steps. On a fully booked Saturday night, every zone opens, each assigned to one server.

This zoning has a double benefit: it balances the load between sections, and it lets you close a zone to clean and reset it between two waves of guests.

The right table mix: twos and fours that combine

Most reservations are parties of two, then parties of four, then groups. A floor plan full of fixed four-tops therefore wastes space permanently: every couple seated at a four-top costs you two covers.

The field rule:

• a majority of two-tops that combine: two two-tops pushed together make a four, three make a six;
• a few fixed four-tops in the spots where combining is impractical (booths, corners);
• modular group capacity rather than one big fixed table: large parties are handled perfectly well by joining tables, and the rest of the time the space goes back to being sellable as twos and fours.

Square tables combine better than round ones; booths fix the capacity on one side but densify the other. The goal stays the same: the room's configuration should follow the reality of the day's reservations, not the other way around.

Circulation: think of your staff first

A dining room is judged with full hands. If a server carrying three plates has to turn sideways to pass between two chairs, the floor plan is too dense. The main aisles, the ones running from the kitchen to the service zones, must let two people cross, which means keeping them clearly wider than the gap between two neighboring tables.

For orders of magnitude, TheFork Manager's layout guide suggests around 1.67 to 1.86 square meters (18 to 20 square feet) per seat, with at least 91 cm (36 inches) between the edges of adjacent tables and 46 cm (18 inches) between occupied chairs. Below that, you gain covers on paper and lose them again in slowed service, grazed plates and cramped guests.

Accessibility is not optional either. In the US, the ADA Standards require accessible dining surfaces: tops between 28 and 34 inches high with knee and toe clearance underneath, reachable by a wheelchair without moving half the room. In the UK, the Equality Act 2010 places a duty on businesses to make reasonable adjustments where a physical feature puts disabled customers at a substantial disadvantage. Plan at least one generous route and table positions where a wheelchair fits naturally.

The handicapped tables: door, restrooms, pass

Every dining room has its bad tables: the one in the draft of the front door, the one along the corridor to the restrooms (or toilets, on the other side of the Atlantic), the one against the pass and its noise. Three rules for managing them:

1. Improve what can be improved: a partition, a plant, a curtain or simply rotating the table is sometimes enough to requalify a spot.
2. Assign them last: these tables fill up when the room is already buzzing, never first in an empty room. A solo diner seated facing the restrooms in a deserted restaurant will not come back.
3. Never give them blindly to sensitive bookings: the anniversary dinner booked three weeks ahead must not inherit the drafty table. Note the quality of each table on your plan and assign reservations accordingly.

Table turnover, explained without taboo

What "turning tables" really means

Turnover is the number of times the same table serves different parties during one service. A two-top occupied from 7 pm to 11 pm by the same couple turned once; the same table serving one couple at 7 pm and another at 9 pm turned twice, and that table's revenue for the night nearly doubled without a single extra square foot.

There is no universal "good" rate. TheFork Manager's own guide gives the formula (parties served divided by available tables) and notes that a common average is around 3 turns per service, while acknowledging it depends entirely on the concept: a lunchtime brasserie lives on turnover; a fine-dining room sells exactly the opposite, a long evening. What matters is knowing where you stand and managing accordingly.

Reservation slots: the heart of the system

Turnover is not decided in the dining room at 9 pm by staring meaningfully at slow guests. It is decided at booking time, through the arrival slots you offer.

Concretely: if your dinner service runs from 7 pm to 11 pm and a dinner lasts about two hours in your restaurant, offering arrivals only at 7:00-7:30 pm and 9:00-9:30 pm mechanically structures two turns. The guest who books at 7 pm knows, without being told, that another seating follows. Accepting an 8:15 pm arrival, on the other hand, condemns that table to a single turn: too late to serve anyone before, too early to serve anyone after.

This is a software setting, not an awkward conversation: in an online reservation system you define your arrival slots and the default duration of a table, and the schedule blocks or frees tables on its own.

Announcing a time limit: common practice, handled with tact

Plenty of fully booked restaurants limit dining time at peak slots, and say so at booking. The practice is routine in big cities on both sides of the Atlantic; what makes the difference is the wording.

What works:

• Announce the duration at booking, never at the table: "For this slot, the table is yours until 9 pm." A guest who books knowing the terms does not feel chased away; they accepted a framework.
• Reserve the limit for the slots that need it: the first Saturday-night seating, yes; a half-empty Tuesday lunch, no. A limit enforced on a slow night is a gratuitous insult.
• Offer an elegant exit: "If you'd like to linger, we'll happily move you to the bar for coffee." The guest keeps a choice, you get the table back.
• Honor the reverse promise too: if the second seating does not fill after all, leave the table alone. The limit is a peak-time tool, not a principle.

What does not work: clearing plates ostentatiously, bringing the check (or the bill) unasked, or sending an embarrassed server to deliver a message nobody prepared the guest for. If the duration was not announced before, you have no moral right to enforce it during.

Average meal duration: the reference points

To set your slots, start from the real duration of your meals, not from wishful thinking. As a reference, TheFork Manager observes that lunch guests often stay about an hour and 15 minutes, while dinner guests stay two to two and a half hours. Your own numbers may differ widely with the menu, the average check and the clientele: observe two weeks of services and measure.

You can then act on that duration without ever rushing anyone: take the order promptly at the start, serve desserts and coffee without dead time, bring the check quickly when it is asked for. Fifteen minutes trimmed from the dead time is regretted by no one, least of all the guest.

Calculating your real capacity per service

The advertised capacity ("60 covers") says nothing about what you can actually serve. The useful calculation takes three steps:

1. Sellable seats: start from your seat count and remove what does not really sell (the wobbly table by the door you only give as a last resort, the chairs that block a passage). Out of 60 theoretical seats, 52 to 55 comfortably sellable ones often remain.
2. Possible turns: service window divided by the duration of a table, clearing and resetting included. A 7 pm to 11 pm service with two-hour tables and 15 minutes of reset allows about 1.8 turns; with structured slots at 7:00 and 9:15 pm, you hold two clean turns.
3. Real capacity: sellable seats multiplied by turns. 54 seats and 2 turns give 108 possible covers, against 54 if you let every table drift into a single turn.

This calculation changes how you look at reservations: refusing a party of two at 7 pm because "tonight is full" while the room empties at 9 pm means leaving half your capacity at the coat check. A slot-based schedule shows you precisely what remains sellable, and a waitlist turns last-minute cancellations into reseated tables.

Comfort or volume: a trade-off to own

Adding tables increases covers up to the point where discomfort sets in: conversations crossing, chairs touching, service slowing because servers slalom. From there, every added table degrades the experience of all the others, and the guest squeezed in on a Saturday night picks another restaurant next time.

The trade-off depends on your positioning:

• Volume concept (brasserie, lunch canteen, diner): assumed density, fast turnover, contained prices. Guests come for efficiency and buzz; they accept proximity.
• Experience concept (bistronomy, fine dining): fewer tables, more space, open-ended dining time. The covers you give up are recovered on the average check.

The classic trap is the unchosen middle: a room as dense as a brasserie with fine-dining prices. Pick your side, and tune floor plan and turnover to match.

Double seatings: for whom, and how

The double seating, two clearly separated arrival waves (say 7:00 pm and 9:30 pm), is the most structured form of turnover. It suits fully booked restaurants that regularly turn people away, and it nearly doubles evening capacity without spending a dollar (or a pound) on construction.

The conditions for it to work:

• real demand on both slots: a half-empty second seating disorganizes the kitchen for nothing;
• a kitchen able to absorb two rushes close together, with a menu designed for it;
• an announced table duration on the first slot (see above), otherwise the second seating starts late every other night;
• effective reservation reminders: a no-show on a double seating costs twice, the lost table plus the delayed next slot. Our advice on reducing no-shows applies doubly here.

If the full double seating feels too rigid, the soft version exists: staggered arrival slots (7:00, 7:30, 9:00, 9:30 pm) that smooth the kitchen rush while still structuring two turns.

Setting all of this up in your reservation software

Everything above is configured once, then runs on its own. With ViteUneTable:

• you choose the arrival slots guests can book online, lunch settings separate from dinner;
• reservations land in your online reservation book: you view, edit and confirm in a few clicks, and you see what remains sellable on each slot;
• guests book through your link around the clock, only on the slots you decided to offer, and each booking triggers an automatic confirmation email;
• the free version covers all of that, with no commission on your covers and no commitment; paid packs add automatic email reminders and anti no-show protection (credit card holds) from €29 excl. VAT per month.

Let's be honest: no software will draw the right table mix for you or choose your positioning. But once your rules are set, it applies them to every reservation, including the ones made at 2 am while you sleep.

Frequently asked questions

How long does an average restaurant meal last?

As a reference point, TheFork Manager observes about an hour and 15 minutes at lunch and two to two and a half hours at dinner. The real answer is specific to your restaurant: measure your actual durations over two weeks of service before setting your reservation slots.

Is it acceptable to limit how long guests keep a table?

Yes, provided the limit is announced at booking, applied only to genuinely full slots, and paired with an elegant alternative (finishing coffee at the bar). What alienates guests is not the limit; it is discovering it at the table.

How do I calculate my restaurant's capacity per service?

Multiply your genuinely sellable seats (remove the tables you only assign as a last resort) by the number of possible turns: service window divided by table duration, reset time included. Example: 54 seats and 2 turns give 108 covers per service.

What is a good table turnover rate?

There is no universal figure: TheFork Manager cites an average of around 3 turns per service, but a lunchtime brasserie should beat that while a fine-dining room deliberately stays below it. Know your concept, measure your current rate, and improve it by trimming dead time, never by rushing guests.

Should I buy two-tops or four-tops?

A majority of combinable two-tops, a few fixed four-tops where combining is impractical, and group capacity created by joining tables rather than one big fixed table. The room then follows the day's reservations instead of suffering them.

What should I do with the bad tables, near the door or the restrooms?

Improve the spot when possible (partition, plant, orientation), assign these tables last once the room is already lively, and never give them to sensitive bookings like an anniversary. A floor plan that records each table's quality prevents these missteps.