soledad penadés

I have been doing a number of talks this year, and each conference has different parameters for what each talk should last. They range between 15-60 minutes. My personal preference is between 20-30 minutes. Those should give enough time to lay out the concepts without rushing it too much, do a bit of a deep dive and finally close down with some conclusions---even allowing for the speaker to hesitate a little bit if need be. 45+ minutes start verging on the "attention span" limit, and I hate both doing those (I end up showing silly stuff because I just cannot be speaking for so long non-stop), and listening to those (I just can't; my brain zones out unless they are REALLY GOOD).

It's funny, at the beginning most of the conferences I attended would have sessions of one hour, so they could fit relatively few speakers. They started doing parallel tracks and then they had more speakers and attendees had to choose between speakers. So I guess then single-track conferences pushed for shorter talks so they could have more speakers and people would not need to choose who to listen to.

But... there is a limit on how short you can go. 15 minutes is nothing when your audience is big and diverse--you want to make sure most of them can follow, so you need a moderately sized introduction. Then you don't have 15 minutes anymore, you might have 9, or 8. You cannot fit too many complex concepts in just 10 minutes. Or well, you can, but most of the people won't follow, and you might also die because you didn't breathe in order to fit more words per minute.

Or the speaker, used to longer talks, might feel the topic too complex to really dive into it in such a short amount of time, and ends up just giving a very shallow overview of a topic. If many talks are like that, a conference can feel dull and meh: what did you tell me I didn't know already?

There were a few comments about this after and before dotJS. The general feeling between speakers was that under 20 minutes is just too short for technical stuff (unless you assume everyone has knowledge of the topic, but then you're excluding a huge bunch of the audience). The feedback I heard from attendees was that either the talks were too vague, or things were unclear (because there was too much content in too little time).

I would like to encourage some debate on this. Conferences should be about quality and not quantity. If I wanted quantity, I could just lean back on my couch and open YouTube---there are so many talks we can watch nowadays, it's awesome!

I'll even go further: what about agreeing with speakers what they think could use, instead of going for a one-size-fits-all solution upfront? Yes, it will be harder to build the timetable, but you're already curating the content and taking your time---why not spend some more to make sure your speakers don't stress about talk length and can focus on delivering great content?

Let's discuss.