Talk Pretty: The Challenges of Public Speaking

Defining the Problemspace

Before getting into some of the challenges, allow me to first setup the problemspace. There are a LOT of smart people out on the various talk circuits giving talks. With over 200 security conferences each year (of varying size), it's unsurprising that there are going to be talks with really good or important content that are simply delivered inexpertly. And, let's face it, putting together a good talk is really hard work!

If you've attended a conference, I'm sure you know what I'm talking about. Some people just aren't naturally gifted speakers, and very few people have ever had formal training or coaching on public speaking. A lot of geeks, in particular, seem to think that talking at cons is a trivial thing: have a good idea, throw stuff onto some slides, and go talk. Unfortunately, these talks oftentimes completely suck! Not only do they tend to be boring, but they also often are disorganized, over the heads of attendees, or simply bad talks. Being smart does not mean you know how to give a talk.

Perhaps one of the biggest challenges overlooked by new speakers is that, regardless of their content, they must be entertaining and engaging. If the audience doesn't feel like they're enjoying their time in your talk, then you're either going to lose them, or you'll get panned in reviews later (if your con has such a thing). On the flip side, if you're good at entertaining an audience, then you'll have a much easier time giving a successful talk.

One need only look at Charlie Sheen to see what I mean. The guy is circulating the country on what is effectively a speaking tour, and he's being very successful. Yet, I think it's safe to say that his brand of #winning is very, very simple. People are fascinated and entertained by him, even if nothing he says is of any real value. If you're a super-smart geek, then think about that for a minute: a complete farce of a preso will run circles around your high-tech-quality talk merely out of entertainment value. That may suck and be unfair, but it's very true.

My Own Preso Failures

As already noted in the introduction above, I recently gave a failed talk. As it stands, I have one major problem, and that is with talking too fast (it happened to me at AppSec DC 2010 in addition to BSides Austin 2011). As a refresher, here are some of my other failures:

- Botched projector setup: I couldn't get my slides to display properly because the projector wouldn't go down to 800x600.

- Poor deck colors: I thought I'd done a good job with high contrast images and text, but I didn't get a chance to test it. Turns out my assumptions were largely wrong.

- Too negative: My talk started with 5 negatives, followed by 3 tips. It felt very negative overall and needs to be flipped around. It should have been a fun and humorous talk, but instead it just sucked.

- Not funny: Was there a joke in there someplace? Only insomuch as my talk was ironic in talking about "not sucking at speaking" while I, in fact, sucked at speaking. D'oh.

My Preso's Negative Examples

In my firetalk, I covered 5 main examples of bad practices. Those are:

- Death by Text: As a general rule, slides should not be heavily saturated with text. Slides are merely props to provide visual points to support the oral points articulated by the speaker. They should help the audience track the presentation, as well as provide useful supporting information that is not particularly distracting. If you have more than a couple dozen words on a given slide, then it's time to think about how you can cut that down.

- Microfonting: If the slides are unreadable, then that's also a bad thing. The most common mistakes I see people make is in using screenshots (or other graphics) that are too big for the screen, and are thus unreadable. Or, the speaker has included so much text in the slide that they felt compelled to reduce the font size to a point that becomes unreadable. A font face smaller than 20pt should generally be avoided. Anything smaller than that quickly gets into jeopardy of being unreadable. And, an unreadable slide pretty much defeats the purpose of having the slide at all.

- Bad Slide Schemes: There are some truly hideous color schemes in various template repositories. People can also do some incredibly hideous things with graphics; especially animated GIFs. In general, try to focus on relatively neutral color schemes that maximize contrast. Remember: Just because it looks good on your high-res screen does not mean it will look good on a much lower resolution projector. Brightness and contrast can vary greatly between projectors, as can the light output (lumens). It's not always possible to be fully prepared for projector conditions, but one should do their best to maximize contrast levels between background and fonts in order to have the best possible chances for success. Also, don't forget to consider slide size. If you design in Keynote, as I have for a while now, then watch out for resolution disparities. My failed talk was built at an 800x600 resolution, but the projector would not go that low.

- Reading the Slides: As a general rule, you should not read directly from the slides. There are exceptions, of course, such as when you need to get a quick right. However, overall, the text of your speech should not be a 1-for-1 match to the slides. As noted above, the slides are just a guideline for your talk, providing a visual outline and highlight for your points. If you want people to read your talk, then you should publish it somewhere and save people the time of sitting in a presentation. That may sound harsh, but the key is this: People are sitting in your talk because they want to hear you speak, not because they want to hear you read a selection from your recent writings, as displayed on the slide. Moreover, it's very hard to be entertaining when you're simply reading from the slides; even if your performance qualifies as a "dramatic reading." ;)

- Unprepared: The biggest mistake any speaker can make is being unprepared. Even if you're being thrown into something last-minute, you need to take time to make sure that you're familiar with your slides, that your logistics are known and sorted, and that you can walk into the room feeling confident and ready-to-go. In the most ideal scenario you will have a chance to give your talk as a dry run, either to a video camera, or even a small audience of your peers/coworkers. Practice makes perfect (or, at least, better). Moreover, practicing your presentation will help reveal areas for improvement, such as broken or cumbersome transitions, slide flow issues, etc.

My Preso's Positive Tips

In addition to covering the 5 negative points above, my firetalk also covered 3 positive tips. I tried to convey this breakdown as "don'ts" and "dos," but I don't think it was properly balanced. There's probably a much more entertaining way to present this information. Anyway, without further adieu, my 3 tips.

- A picture is worth a thousand words. (Less Text): One of the best presentations I've ever seen had almost no text whatsoever. It was primarily pictures that matched with the speaker's points. It was very entertaining, plus memorable, which made it effective. In another talk that I saw more recently, the speaker literally turned the slides off at several points (he said he was sick of looking at his own slide deck), which also worked very well. People attend talks to hear speakers on topics, not to read slide decks. Moreover, you can say a ton in a single picture that would otherwise require myriad words. To top things off, pictures can oftentimes get a message across far more effectively than trying to describe it orally or in text. Consider, for example, a magazine article for a Jeep. They're almost universally a large, glossy picture of the truck out in the wild being driven for adventure. It evokes many positive sentiments and provides quick correlation in a single image, which would almost never be accomplished simply through writing. That's effective.

- Guy Kawasaki's 10-20-30 Rule (Know Your Audience): I had the opportunity to hear Guy Kawasaki speak a few years ago. One of the things he talked about was his "10-20-30 Rule" of VC pitches. As a VC, he was sick of seeing long, droning, boring pitches that started with the history of computing before getting to the point. Instead, he suggested getting right to the point, using no more than 10 slides over 20 minutes in at least a 30-point font. For me, there are a couple key takeaways from his recommendation. First, it's important to know your audience. If you're speaking to a room primarily filled with senior techies, then you don't need to explain the history of IP because they (should) already know that. On the other hand, if you're speaking to a room of managers or executives, you need to modify your content and delivery accordingly so as not to lose them in the technical weeds, or otherwise alienate them. Humor and cartoons might play well with one group, while only professional stock photos should be used with another. In general, know who you are speaking to, what their experience levels should be (on average), and then construct a minimalist deck to support that delivery.

- Practice! Practice! Practice! (Practice): This point was highlighted in the negatives, but it's also a major positive tip. Practice your talk. Practice in front of a mirror. Practice in front of a camera. Practice on a projector with a remote. Practice with a small, friendly audience. Practice with a small(ish), less familiar audience. Run through the talk a few talks before you give it "for real." You should know how your slides flow and transition without having to stare at the screen. You're talking to the audience, not the project slides! Also, practicing will help you detect speech ticks that you might have (e.g. "umm," "ahh," "errr," "right?"). There may not be such a thing as perfection, but practicing a fair amount will get you closer to that lofty goal than simply wandering in unprepared.

Thoughts Going Forward

The question, then is whether or not this talk should be revived and revised, or if I should just let it die as this blog post. What do you think? Here are some of my thoughts...

- Be funnier...: If I'm going to revise and re-deliver this talk, then it needs to be A LOT funnier than it is today. Is this possible? I'm not sure. I think I'd need to come up with some humorous examples or pictures to better highlight the points. Moreover, maybe the points need to be wordsmithed to be less damning and, well, boring.

- Be less negative, more positive...: The talk itself is way too negative, which is evidenced by having 5 negative points vs. only 3 positive tips. Is there a way to merge them down into a small, more manageable number? If so, what would those points be, and how can I make them all positives instead of just negatives?

- Or, maybe dump it altogether?: The easiest course of action would be to stick a fork in it and call it done (as in, over, complete, final, bashed to bits). Is there even interest in this topic? Maybe not. It certainly did not go over very well, and left a lot to be desired. What do you think?