January 15, 2014

1. Every talk needs a story.

The introduction should introduce your audience to the problem you’re trying to solve. By the end of it, your audience should know what the problem is and be convinced that it’s an important problem worthy of your time and effort. Establish what you hope to accomplish.

After the introduction comes the method. You should describe your approach to the problem, describing not only what you did but how it differs from other approaches. If your method has any caveats, be honest.

Now that the audience is familiar with your approach, discuss your results. The amount of time you spend on a particular result should be proportional to how important you believe the result to be.

Finally, the conclusion briefly summarizes your results and discusses the wider implications of your results. Be sure to distinguish between conclusions and speculations. How do your results contribute to solving that problem you talked about in your introduction?

2. Tell 'em what you are going to tell 'em. Tell 'em. Then tell 'em what you told 'em.

Give your audience guidance. Do not assume that your audience is with you at every step - repetition of key points at a high level is essential.

3. Talks are written with the audience in mind.

Take your audience’s level of interest and experience into account when dialing in the breadth and depth of the talk. Likewise, be sure that (for any audience) you’re not casually using acronyms or concepts two slides before you actually explain them.

4. Every slide needs a segue.

The end of every slide should flow naturally into the beginning of the next. The last thing you say on slide (n) should suggest the content on slide (n+1).

For example, if you’re delivering a slide comparing existing approaches to embiggening, and your next slide discusses your new approach to embiggening, you might say something like:

“Approach A is too expensive, Approach B has a high failure rate, while Approach C is prohibitively explosive. What is required is an embiggening procedure which is low-cost, reliable, and safe.” [Click to next slide] “My approach to embiggening is...”


“As you can see, the field strength is linear, but this equation assumes that the viscosity of the luminiferous aether is negligible.” [Click to next slide] “As you can see, the results are very different when the viscosity of the luminiferous aether is considered.”

5. If the text is only there to remind you of what to say, cut it.

The slides are there for your audience, not for you. If they’re reading the slide, they’re not paying attention to what you’re saying.

6. If there is a diagram, chart, or equation on a slide that you don’t plan to walk the audience through, cut it.

There’s a natural tendency to throw everything you’ve done into a talk. Resist it. If it’s not important enough to spend thirty seconds of your presentation time on, it doesn’t belong there.

7. Avoid the following common mistakes.

Using more significant figures than you have. 2.35689745632 +/- 0.3659487652 g/mL should be 2.3 +/- 0.4 g/mL. All of the numbers smaller than the error are meaningless.

Using a table when a graph will do. Graphs are almost always easier for the audience to parse.

Graphs with unreadable text. Excel tends to default to extremely small text that is indecipherable in a presentation.

Graphs that don’t know if they’re constructed from measured data or an equation. Measured data should be shown as symbols (like ▵,╳, ▢, or ◯) while the use of a line generally implies an equation (even if that equation is just a fit to the data).

Inappropriate use of statistics. When in doubt, consult an expert.

8. Plan to use a laser pointer sparingly, if at all.

Nine times out of ten, using an animation to make elements on the slide appear is more effective and less distracting than using a laser pointer.

9. Get a presentation remote.

Otherwise you’re stuck hiding behind your laptop.

10. Practice the talk.

 No, seriously. Practice the talk, in front of a small group if possible. Everybody plans to; few succeed.