Lightning talks: five minutes, one idea, a hard stop

A lightning talk is a short slot of five minutes, sometimes ten, with a hard stop. It is the cheapest thing a conference can put on its programme and, for a first-time speaker, the lowest risk way to stand in front of a room. It is also the slot most often treated as filler by the people scheduling it, which is where most of the disappointment with the format comes from.

What five minutes actually holds

One idea. Not one idea with context, three examples and a conclusion, one idea, shown.

The arithmetic is unforgiving: five minutes is roughly 700 spoken words, and an introduction that would be reasonable in a conference talk eats a fifth of that before anything has been said. The talks that work open at the point a longer talk would reach at minute eight, and they usually have a single thing on screen at a time.

The reliable shapes are few. Here is a thing that surprised me, one finding, the evidence, and what changed. Here is a tool and what it replaced, the problem in one sentence, the tool doing it, the trade-off. Here is a small piece of a big thing, a single function, a single regulation, a single failure, which is the safest of the three because it has a natural end.

One idea. Shown, not introduced.

The hard stop is the format

Lightning talks work because the clock is not negotiable, and every conference that has allowed overruns has ended up with a block that runs half an hour long and a room that leaves. The usual mechanisms are a visible countdown, a chair who stands up, and slides that auto-advance, the last being the strictest and the one that changes how a talk is written rather than how it is delivered.

For a speaker, the practical consequence is that the talk has to be rehearsed at least once against a real timer. Reading a five-minute script aloud is the only way to discover that it is actually seven, and no amount of speaking faster on the day recovers two minutes.

Why they belong in the schedule on purpose

Three reasons, and only the first is about the speakers.

They are the lowest-risk entry point onto a programme. A five-minute slot does not need the confidence, or the free weekends, that a full session does, and a good lightning talk is the most common way a first-time speaker becomes a returning one. Some conferences allocate them on the day from a sign-up sheet, which removes even the proposal step; that is covered under first-time speakers.

They absorb the good proposals that do not fit. A single-track conference rejects most of what it receives, and a lightning block gives the committee somewhere to put a strong idea that does not carry forty-five minutes. Offering one instead of a flat rejection is a materially different message: see after a rejection.

And they reset a room. A block of five-minute talks after lunch does more for attention than a coffee break does, which is why both documented editions of this conference used them at the low points of the day rather than as an afterthought: see programme and tracks and the 2008 schedule, which attached them to breakfast.

What goes wrong

Almost always one of four things. The talk is a conference talk compressed, so it arrives at its point as the timer stops. The slides carry paragraphs, which cannot be read in the seconds they are on screen. There is a live demo, which needs setup time the format does not have, a recording of the demo removes the risk entirely. Or the block is scheduled last on the final day, when a third of the room has left for a train.

Questions