The seating plan is not the problem. The room is.

Couples lose entire weekends to the seating plan. Most of that pain is not caused by who sits with whom. It is caused by a room that was never going to work, and no arrangement of names fixes a room.

The room decides more than the chart

Long banquet tables and round tables produce completely different evenings. Rounds seat ten people who can all see each other and hear roughly half. Banquets seat people in a line, where you speak to four people all night and cannot see the far end.

Neither is better. But choosing rounds and then agonising over "who will they talk to" is solving the wrong problem — on a round, they will talk to everyone. On a banquet, the two people either side matter enormously.

Sightlines beat social engineering

A table behind a pillar is a bad table regardless of who you put there. A table beside the kitchen door has staff passing all night. A table at the far end of a long room cannot see the speeches and will talk through them.

Walk the room, identify the compromised positions, and accept that they exist. Then allocate them deliberately — to the people most likely to enjoy themselves anyway — rather than pretending they are equivalent.

Distance from the dance floor is a decision

Older guests near the floor will feel the volume all night. Put them furthest from the speakers and nearest the exit, and they will stay longer and enjoy it more. The people who will dance want to be close, and they will get up sooner if standing up means being on the floor in three steps.

This one change does more for the energy of a night than any amount of careful pairing.

Tables, not seats

Assign people to tables and let them choose a chair, unless there is a specific reason not to. It removes most of the agony, and it removes the problem of a guest arriving to find their exact seat next to someone they would rather avoid.

The exception is the top table and anywhere you are actively separating people.

Build it last, and build it once

Do not start until RSVPs have closed and the final numbers are in. A plan built at sixty per cent response is built twice. Give it one focused evening rather than six anxious weeks, and ask somebody outside the family to look at it — they will spot the thing you have stopped seeing.