Where the usual ratios come from

They are not a standard: they are the average of what works. The classic one waiter per fifteen covers comes from plated French service with three courses in a compact room. As soon as any of those three conditions changes, the ratio stops working as a single figure.

The mistake in using them blindly is not over-hiring: it is hiring the right number and having service go badly anyway, because the problem was never how many people there were but how far they had to walk.

What changes with the type of service

Plated service concentrates the effort into short, very demanding moments; a buffet spreads it out but needs fixed staff who never leave the line; a reception turns it into continuous circulation.

  • Plated: the number of courses multiplies the peaks. A five-course menu with the same covers asks for one or two more waiters than a three-course one.
  • Buffet: you have to count two separate teams. The floor team, on a looser ratio, and the buffet line, which is fixed: two to four people depending on the hot stations.
  • Reception: the ratio depends on how many service points there are and how they are spread across the plan. A reception in two areas needs more people than one in a single space with the same guests.

How the plan comes into it: distances, office and gangways

The distance from the office to the furthest table is the variable most often ignored and the one that weighs most. In a long forty-metre room with the office at one end, the last table is twenty seconds away on every trip, there and back. At ten trips per service per waiter, that is a whole waiter’s worth of lost time.

The service gangway is the second. With ninety centimetres clear two waiters pass each other; with sixty, they wait. The difference does not show in the kitchen’s timings, it shows in the temperature of the plate when it reaches the table.

The first-course peak

This is where any calculation made from averages breaks. Every table is served at once because the whole room sat down at once, and for those ten or fifteen minutes the ratio actually needed is close to double the average for the service.

There are two ways to solve it and both get decided on the plan. One is to stagger the service by zones, which requires the zones to be drawn and assigned. The other is to reinforce that stretch alone with support staff who come in for two hours. What does not work is applying the average ratio and hoping.

How to work it out with the plan in front of you

You start from the ratio for the type of service, then correct it with three figures the plan gives you directly: how many metres from the office to the furthest table, whether the service gangway lets people pass each other, and how many service zones you can define.

The rule of thumb that has held up over the years: for every fifteen metres of average office-to-table distance above twenty, add one person per hundred covers. It is not exact, but it corrects in the right direction and avoids the staffing error, which is always on the low side.