Seating Preference

Every reservation platform has a place to put a seating preference, and every restaurant has more preferences than it has recorded. The gap is not a software problem. Preferences surface in the middle of service, in front of whoever happened to be standing there, and the recording step competes with everything else happening at that moment.

How do you find out where a guest likes to sit?

There are three routes, and only two of them are ever written down. A guest states the preference, at booking or at the door, and it goes on the reservation. A guest gets moved during service, which the server notices and almost nobody records, because the moment passes and the note never gets made.

The third route is the one nobody uses. A guest seated at the same banquette on eight of their last ten visits has expressed a preference without saying a word, and that pattern sits in the reservation platform’s own history as table assignments nobody reads back. Preferences inferred from where a guest has actually been seated cost nothing to collect, because the data was generated by the act of seating.

What kinds of seating preference are worth recording?

Three kinds, and they are not equivalent. Positive requests name a place: the banquette, a booth, a window, the patio, the bar, the corner table. Avoidances name a thing to stay away from: the kitchen door, the restroom corridor, the service station, a speaker, a draught. Occasion preferences depend on the night rather than the guest, such as a quiet corner for a proposal or a round table for a conversation.

The distinction that matters most is between a preference and a requirement. A guest who needs step-free access, room for a wheelchair, or a quiet table because of a hearing aid does not have a preference, and filing that in the same bucket as “likes the window” means it gets traded away on a busy Saturday by a host doing exactly what the system implied was fine. Restaurants that keep the two apart treat one as a constraint on where the party can be seated at all and the other as a goal to hit when the room allows.

Can a restaurant actually honour a seating preference?

Not always, and being clear about that is part of using the information well. A preference is a claim on a finite room. Two banquette regulars booked at eight o’clock with one banquette available means somebody sits elsewhere, and a couple placed on a four-seat booth during peak costs the room two seats it could have sold.

So preferences function as inputs to a seating decision rather than as promises. They change the order in which a host assigns tables, and they pay off most where they are cheapest: off-peak, in a partly empty room, or for the guest whose value justifies the seat. The version that works is a floor plan built before service with the night’s known preferences visible on it, rather than a note discovered as the guest is already walking through the door.