A talk is like a movie trailer for your paper — if you do your job, the audience should be excited to read it.

Movie trailers can rely on suspense, mystery, and movie stars to build excitement.  You, however, will have to rely on the clear, direct communication of the WHY, WHAT, and HOW of your work, and that requires careful thought and judicious choices about what keep in and what to leave out.

Teasers, Trailers, and Featurettes

Trailers come in different lengths.  Teasers are very short while normal trailers are quite a bit longer.  Longer still are “featurettes” that preview the movie and give  a peak behind the scenes.

Seminars

Seminar talks are like featurettes.  In 40-50 minutes you have time to cover a research paper in reasonable detail.  You can and should give a thorough and compelling description of your WHY.

You will also have time to discuss all or most your contributions (WHAT), and give a the audience a solid understanding of how all the components of your system work together.

Even with all that, you’ll have time to talk about some implementation details  (HOW) in some depth.

All that time can tempt you to include everything.  Regardless of how much time you have, never sacrifice clarity for completeness.  A real danger in a seminar talk is going into too much detail.  If aspects of your system are too complex to explain clearly in a talk, then you should leave them out or cover them at only a high level.

Conference Talks

To go from a 50 minute featurette to a 15-25 minute trailer, drop most of HOW, the less interesting parts of WHAT.  Keep all the WHY.

Take care in deciding what to keep and what to leave out.  Keep the most important, exciting, and interesting contributions and skip the others.  Don’t worry, they are still in the paper.

Trim the contributions you keep as much as possible.  Which details are essential?  Which parts are most interesting, surprising, or novel than others?

Keep only the most important and interesting parts of your HOW.

Above all, do not make cuts that reduce clarity. It is much better to leave the audience with a clear understanding of part of your system than with a muddled picture of all of it.  Examples and animation of your system in action can be very effective at conveying the essence of how your system works.  Keep them.

Lightning Talks

For the lightning talk, you’ll have to shrink WHY, so find a way to summarize it succinctly without sacrificing clarity or confusing your audience.

You may need to broaden your motivation.  At the highest level, the WHY for all of computer science research is “Computers are good, but not good enough to everything we want them to do.”   You can simplify your WHY by moving it up toward this very general statement.  You already have some practice at this — it’s probably how you describe your research to non-technical friends and family.

Your WHAT is probably reduced to a few bullet points and a block diagram — just enough to convince the audience that you address your WHY.  Your examples will probably have to go (unless they provide blinding insight in just a few seconds).

HOW may disappear completely, unless you have something really compelling to say.  If you built a working microchip or deployed your system in the real world you should mention and emphasize it.