Choosing a talk topic

1. Pick a topic in which you have deep expertise.

Particularly when you're getting started, your best bet is to start with something that you know inside out.

The first talk I ever gave was on Vim, a text editor I'd been using for five years. I'd seen my coworkers using it far less efficiently and figured that if they needed some pointers, other developers would too.

Writing the outline for that talk was a breeze, because I'd built up so much useful knowledge about the topic over the years. And when I eventually gave the talk, I was confident I could handle any questions from the audience.

2. Pick something you rant about.

Counter-intuitively, you can give a successful talk about something you hate.

Most organizers like to have a little excitement in the schedule. Since most talks are about why technology X is great and should be used, one that says the opposite about technology Y can be a nice change of pace. Picking a fight can get your proposal noticed.

One caveat: if you convince your audience something sucks, the first thing they'll want to know is what they should do instead. Make sure you spend the bulk of your talk time on this.

3. Pick something you know nothing about.

For the longest time, I wanted to learn Haskell and Erlang, but had never found the time.

When my talk "What Rubyists should steal from Haskell and Erlang" got accepted, I finally had the motivation I needed to really dig into these languages.

This tactic is more risky than the previous two, so it's probably best to only consider it if you're an experienced speaker. (And, if I'm being honest, the talk I described above didn't turn into one of my best.)

4. Ask your colleagues what they've learned from you.

Sometimes it's hard to tell that you know something that would be useful to others.

Before publishing my most popular blog post to date, I asked in Campfire "doesn't everybody know this already?". It can be hard to remember what it's like to be a beginner, and what kind of things you didn't know. It's also easy to discount your own expertise (Dunning-Kruger anyone?).

One good hack around these challenges is to ask your co-workers what they've learned from you. You might be surprised. Marvel for a minute, and then mine that stuff for a new talk idea!

Next
Proposing successfully