If you leave every talk to the last minute, something dangerous happens. You stay up late four nights in a row, you get it done, you deliver it, and everyone seems to like it. So you start to believe you work better under pressure.

The talk always happens. That is the trap. You never get to see the version you would have given with more time, and almost always it would have been better. We tell ourselves we are special unicorns who thrive on adrenaline. It isn’t true.

So I use a work-back schedule. If the talk is ten weeks away, I decide what happens in week nine, week eight, week seven, all the way down. If it is tomorrow, I do the same with hours. Writing the schedule is often the moment you realise you need to start now: the research needs doing, and the people you need to talk to are about to go on holiday.

The more prepared you are, the better you are. It is the least exciting advice I know, and the most reliable.