I started giving tech talks at conferences at for corporate groups back in the mid-90s. Back then, I’d do several different formats: conference talks in which I’d have 45-90 minutes to talk, all-day workshops (7 hours), and corporate sessions (two to four days in which I’d be teaching for seven hours a day).
At one point someone asked which format was more difficult. You’d think that a four-day format would be, but it’s really not. The four day format is an endurance marathon, but you can do it if you’re very well prepared. And I can talk forever if I have to. Plus, in a four-day format you have a lot time to adapt to your audience, and if you screw up day 1 you can make it up on days two through four.
Shorter formats are much harder. I’ve often described the 45 minute conference talk as similar to landing on an aircraft carrier. The time window is tiny, and the margin for error is pretty much nonexistent. You have to do a lot more planning to ensure that your talk doesn’t go overtime, and if you inadvertently end early the conference logistics can get screwed up.
This week I’m working up a talk about the business side of CodeLesson that we’re proposing to a tech conference. (I’ll name the conference later, assuming my talk is accepted; it’s a competitive process and there’s a good chance we’ll be cut.) I have to give the talk to the conference organizers using a screen share on Friday, and then they’ll decide if it’s worth having me for the in-person conference. The format? Three minutes of me talking, followed by ten minutes of Q&A. Rough! This is going to be the smallest aircraft carrier I’ve ever landed on.
I’m a pretty good extemporaneous speaker, but there’s no way I’d do a three-minute talk without extensive preparation. For me, the big risk is that I’ll run overtime; three minutes is really nothing. So to prepare, I started by locking myself in my office. I went through my demo using the stopwatch function on my phone without really looking at it. The first run-through, I got up to four and a half minutes and I was about half-way through my slides. Whoops.
I backed up and deleted about 25% of my slides. Then I scripted what I was going to say on each slide. Normally I hate doing this because it slows the talk down and can make the talk seem stilted (you know, like you’re reading from a script). But even if your talk is pretty tightly scripted you can still make it sound like it’s extemporaneous. It’s not necessary to say every single word on the script, as long as the sentences that are coming out of your mouth still make sense. And, most importantly, varying the tone of your voice while reading (emphasizing the important words and the last word of a sentence) can go a long way toward making your talk seem more natural and unscripted.
After doing this, I got the talk down to about two minutes and forty seconds, which I think is where I’ll leave it. But there’s one more step to polish the talk up a bit more. I recorded the talk (using ScreenFlow) and put it on DropBox for members of my team to review. That way they’ll be able to tell me if my talk missed any important points, or whether I should take something out.
I’ll still need to give the talk live via screen-sharing (and in person at the conference if the talk is accepted), but having a recorded dress-rehearsed version of the talk will also help in case something goes wrong (like I get sick that day or the screen sharing doesn’t work). One of the keys to being a good public speaker is to use belt and suspenders on everything — you want to rely upon as few external variables as possible, particularly technical aspects that aren’t in your control.
Fingers crossed for Friday.