A better way to prepare speakers for the room

Stand backstage with someone five minutes before they go on and they will not ask you about their argument. They will ask you where to stand.

The questions are always the same four, and they are all logistical: where am I, when exactly do I start, who introduces me and what do they say, and what happens the moment I finish. A speaker who knows those four things has all their attention available for the talk. A speaker who does not spends their first two minutes visibly recovering.

This is good news, because logistics are answerable in advance from data you already hold.

The standard answer is a document, generated once, emailed, and wrong within a fortnight.

It is wrong in the most damaging possible way, too: confidently and specifically. It gives a time and a room in clear type, and the speaker believes it, because why would they not. The programme moved; the PDF in their inbox did not. That is the drift problem pointed directly at the person least able to detect it.

Give a speaker a live view instead of a document and the whole failure mode disappears.

Their page reads from the same record as the showflow, so it is correct at the moment they open it — including at nine o’clock on the morning of, standing in the foyer, which is when it actually gets read.

For the speakers who work from a script — and for every host, who always does — the script should be attached to the session rather than living in a separate document with its own version history.

When it is, the prompter knows which session it is in and what it is showing. A late change to a name or a time reaches the words on the glass. And the operator and the presenter are looking at the same thing, which removes the specific horror of a presenter reading a line the operator does not have.

The last thirty seconds before someone walks on is the highest-stakes information transfer at the whole event, and it is usually done by a stressed person whispering.

It goes better when nothing in it is new. If the speaker already knows the four answers, the handover is a nod and a countdown, which is all it should ever have been.