Where a restaurant layout is about the whole room, a restaurant table layout is about one decision repeated: how many tables of each size, and where. It is the decision most restaurants inherit from whoever fitted the room out and never revisit, and it is the one that silently sets the ceiling on what the booking book can accept. This page is about how to read your own book to find the right mix, and what changing it costs and returns.
Start from the parties that ring you, not the parties you imagine
Take a month of bookings and count them by party size. Almost every independent restaurant finds the distribution is not what they assumed: usually more twos than expected, a solid block of fours, and a tail of sixes and above that is small in number and disproportionately valuable in spend. The right table layout is the one that accepts that distribution with the least joining and the fewest refusals, and it is discoverable in an afternoon from records you already have.
Flexible is not free
Two twos that can be pushed together to make a four sounds like the best of both, and it costs something every time you do it: a pause in service, a disturbed neighbouring table, cutlery relaid, and a table that is now harder to turn because the party is bigger than the space was built for. A layout that requires joining for a quarter of its bookings is working hard. One that requires it for one in twenty is flexible. The difference is worth measuring before deciding you are flexible.
Where each size goes matters as much as how many
The six top belongs somewhere it can be joined and somewhere a party of eight will not have to be walked through the whole room. The twos belong where a couple is comfortable and a server can reach both sides. The counter seats, if they take bookings, belong in the pacing arithmetic rather than as a vague overflow. A restaurants layout that gets the counts right and the positions wrong will still refuse bookings it could have taken.
Test a change on paper before you move anything
Changing a table layout is a day of work and a week of arguments, so it is worth an hour of arithmetic first. Put the current mix through the worksheet on this site, then put the proposed mix through it with the same turn times and fill rate, and look at the covers a service and the money a week. If the difference is inside the noise of your own fill estimate, the change is a preference rather than an improvement, which is fine as long as everybody knows that is what it is.
Questions people ask about restaurant table layout
What is a good mix of table sizes?
The one that matches your own bookings. As a starting point, most independent dining rooms find that roughly half their tables want to be twos and the rest fours with one or two larger, but the honest version of this answer comes from counting a month of your own book rather than from anyone's rule of thumb.
Should the counter seats be in the plan?
If they take bookings, yes, and they should be in the pacing too. Counter seats that exist only as overflow tend to be either invisible when you need them or oversold when you do not, and both happen because they were never counted properly in the first place.
How much does the table layout really change the takings?
Enough to be worth an hour. The mix changes which bookings you accept and the positions change how fast tables turn, and both multiply into the week. Work your current and proposed mixes through the same worksheet at the same fill rate and the difference is yours rather than an industry average.