Attending technical events, from the local after-hours meetups to the high-caliber and well-known conferences, becomes the usual part of a developer’s life. Generally, those events are packed with 45-minute talks, often also to the full one hour. I argue that there are more benefits of limiting such tech talks to a shorter duration, say 20 minutes (or even 18 minutes, in the style of TED talks). The most important is that it will lead to a more thoughtful, lean, and balanced content.

First of all, the presenter needs to cut to the chase. Long-winded introduction is out of the question. In the last few years, we have seen some tremendous improvement in the way the self-introduction was conducted. There is no more wasted minutes of credentials flashing (meticulous list of projects, certificates, affiliates), irrelevant funny story about someone’s Twitter handle, or any other conversational icebreakers longer than necessary. But of course it is even better if there is simply no room for those. After all, the audience can use their favorite search engine to look for more information about the speaker.

As a corollary, the speaker also needs to filter the content and condense it. During the preparation, careful thought needs to be exercised to pick the most relevant topics only, given the time constraint. It will be less of "How do I fill this 40 minutes" and more of "What will be the three important take aways for the audience". Remember that the speech is for the benefit of the audience, it is not a vehicle to show off various domain expertise of the presenter.

A shorter allocated time naturally promotes to a more balanced composition. If the presentation is in the format of problem-solving (typical for a lot of tech talks), the speaker will not ramble too much on the initial sales pitch. How many times have you witnessed talks about web performance in which the first 15 minutes were spent only on explaining the benefit of performant web sites? Let’s get straight down to business, no need to convince us that we are in the right room.

Last but not least, packing different talks into an hour means more diversity and variety for the attendees. One hour is not enough anyway to make someone become really competent in a particular field. That time is better spent on giving the audience some use-cases, inspirations and pointers for further self-exploration. Even one convincing story will be more memorable than a long checklist of tips and tricks. Now, imagine getting three inspirational and memorable stories in one hour!

What if the topic can not be packed into a 20-minute session? Well, there is always a choice of splitting them into two parts, conveniently mapped (among others) into beginner vs intermediate level. Those who are not novice anymore but still want to enrich their knowledge can choose to only participate in the second part. Other curious minds who just want to get the taste of the field could focus on the first part, possibly even skipping the sequel. It is a better value for everyone.

There will be always the need for longer talks. However, those in-depth coverage are better handled in the form of workshops. Many conferences have this extra one-day, usually before the usual technical sessions start, dedicated for those long, lecture-style of tutorials.

These days, conference organizers often offers some extras to the standard technical sessions, usually in the form of lighting talks, unconference, Ignite, and other similar variants. In some cases, like JSConf, many talks finished in 30 minutes of less. At the last O’Reilly Fluent, we have seen two 20-minute talks packed into one usual slot. In the near future, hopefully more and more organizers will consider and give priorities to those short, high-quality talks.

Remember Pericles, Time is the wisest counselor of all.

  • Andrew Betts

    This is one of the key reasons why I started Edge (edgeconf.com, at which Ariya is participating in two months’ time ). Giving tech talks is actually a very difficult skill to master, and I found myself thinking that we need a way of extracting knowledge from people in a public forum without them needing to become accomplished speakers. Panels are often very dull and dry, so we sought to reinvent the panel concept. Strict time limits, aggressive and proactive moderation quick fire audience reaction and great AV support so we avoid long pauses for mic-running. It’s a ton of small things that make for a great session.

    OK, so that whole thing was a plug. Sorry. The point is I completely agree, and as the web platform gets bigger, more complex and more specialised, we need to be better at matching the right talks to the right audiences as well as ensuring that the speaker is able to deliver the topic well.

    • I couldn’t agree more. I watched past videos of Edge and the combination of short material from each speaker and rich Q&A is highly attractive! On top of that, there is a lot of useful take aways for further follow-ups. Kudos! 🙂

  • I completely agree. That’s also why we chose to limit dotJS.eu talks
    to 20 minutes and as far as I know it worked great 😉 Sylvain and
    Ferdinand used the same format for the other dotConferences* and now
    they have a bunch of videos you can watch during your commute (I suspect
    less than 30 minutes commute are the most common, at least in Europe).

    * http://www.youtube.com/user/dotconferences/videos?view=0&flow=grid&sort=p

    • Very true! I quite enjoyed the videos from those dot Conferences: high quality, useful, and compact.

  • OldETC

    At a conference there are many levels of individuals, and there are also some people who are the experts who have gaps in their knowledge that they don’t know about or can’t acknowledge, either because they don’t know about the gap, or because they choose to ignore it. The longer talks often cover points that include these gaps.

    Speaking also requires getting the audience’s attention. You may have good attention, and you may be eager to get to the meat, but others do better if there is some warm up, to get their mind to switch from the last topic to this one, to finish up some notes, or just to descend from thought spaces.

    Some useful information requires that the audience, the whole audience, be along for the ride. I can say that the best presentations do have a bit of warm up, some background and then the full topic, followed by principal points, then a summary. This takes some time to be fully effective for most people.

    And the time required for a talk is relevant. If you are talking about C code for a specific function, 10 minutes might be too long. If you are discussing the correlation of complex data sets and physical interpretation of the results, a day might be too short.

    I think you are on a good track about length of a presentation, and if you choose a time, you must also arbitrate the size of the topics in some fashion and establish moderation rules that are acceptable to participants, and audiences. If you go too far in one direction or the other you will lose part of your conference, either attendee’s or presenters.

    • Many tech conferences nowadays are quite scoped. Coming to a JavaScript conference without knowing much about JavaScript is hardly the norm. And yet, many JavaScript-related tech talks often spend too much time reiterating stuff known to the majority of the audience.

      The usual stage of intro-analysis-conclusion doesn’t negate the fact that each stage shouldn’t be longer than necessary. Get straight down to business is the name of the game.

      Here’s a nice exercise (for all of us): pick some long videos of a tech conference, watch it from the beginning to the end, and then think of which portions can be cut without affecting the gist of the talks.