Looking back on a year of speaking at conferences, I'm rather flabbergasted that I managed to make it without my pants falling off mid-sentence or my laptop exploding in a fireball of self-doubt and international power cords.
I'd read all the blog posts and heard all of the advice: slow down, speak loudly, tell a story. But goddamn, no one told me I'd have to put my laptop down on the ground twenty feet away behind a couch because that's the only place the projector's VGA cable would reach.
Turns out there's all this other stuff to worry about, too!
When you're on stage in front of a few hundred people, your brain is screaming at you DON'T SCREW UP IN FRONT OF ALL THESE PEOPLE. But these people aren't your main audience.
Run the numbers: most technology talks are in front of 50-300 people. That's your most important audience; after all, they came to listen to you. But if you post your talk online afterwards, you could reach thousands more. It's a different audience. You can choose to ignore this audience completely, or you can choose to embrace them.
Most conferences are a crap shoot when it comes to video. Half the time they won't record your talk, and the other half of the time it may take months before your talk is published.
Something I've been doing recently is making a screen recording of my talks using QuickTime on my Mac. It'll record both my voice and my slides as I flip through them. It's a much better experience than just posting contextless slides.
There are a number of things you'll want to take care of before jumping into your talk:
I've done one or two talks that I thought were great ideas initially: projects that weren't finished yet, things that I thought would be interesting to talk about, and so on. They're talks I regret. Talks should always be reactionary rather than anticipatory: they're going to come off as more natural, more interesting, and above all, more valuable.
Think about writing a detailed blog post before you speak on any subject. That way you'll know your comfort level with the information, and, more importantly, you can gauge the reaction of everyone else. Is it interesting? Are you wrong? Does anyone care?
My favorite part about talks are the questions afterwards (if your conference allows it). Don't shy away from them. They're a great barometer of how well you know your material, and, most importantly, whether your audience "got it". If you plan on giving your talk again, take special note of what happens in the question-and-answer section. It's a goldmine for improving and changing your talk for the next time through.
There's one mistake I consistently see made by speakers both novice and experienced: they're not excited about their talk. Changing this one thing will completely, utterly change the perception of your talk for the better.
You're on stage, ostensibly talking about something that interests you or makes your life better or saves you time. Be fucking excited about that. Let your enthusiasm bleed through. People absolutely react to emotion. The opposite is just as true: a passive voice on-stage will lead to a passive response by your audience.
These are some of the many things I've picked up on while giving talks, but it's certainly not a comprehensive list. Public speaking's all about thinking on your feet, of reading people. Like everything else, it all gets better with practice.