My friends and family are under attack in Ukraine. Donate to protect them directly or help international organizations.

Should Conference Talks be Written in Advance?

March 3rd, 2015

I had a discussion on Twitter today regarding a conference's selection process. It was suggested that speakers should be forced to write their entire talk before submitting it to a conference, perhaps even presented at a local event. This might work for speakers who give the same talks over and over again, but that's not the majority.

I organize a conference where I select nearly 150 talks each year. I also speak around the world (South Africa, Spain, Germany, USA, etc.) That gives me a unique perspective of the relationship between speakers and organizers and how they can better collaborate.

How hard is it to write a talk?

It's extremely time-consuming. I personally spend between 20 and 60 hours to prepare a conference talk, depending on how much code and research is required. It used to take me more when I was still new to the speaking scene.

I submit new abstracts all the time: different abstracts depending on the type of conference. I know these subjects very well, but I can't possibly write 10 or more talks each year if only a handful of them will be picked. First, it would take too much time. Second, those that will not be selected would be wasted and might not be relevant next year.

Is it fair to demand that speakers potentially waste so much time? After spending hundreds of hours preparing the talks, it would be even more crushing to receive a rejection e-mail from the committee.

How do organizers know if the talk will be good?

We look at both the speaker and the topic. A speaker who has a history of giving great talks would normally write subsequent talk of equal or superior quality. So if a speaker is good and has experience to show, we don't need the slides or the code. I have occasionally asked speakers to provide an outline when they had nothing else to show and that worked out well.

To determine whether the talk is good, we read the title and the abstract, and check whether the speaker has experience related to the topic. We discuss the interest in the topic, and in our case, check how many people voted for it. There are other criteria that are unrelated to this question. I blogged about them earlier here and here.


We know that the job of a speaker is hard enough. Even harder for those who are still trying to get into their first conference. We don't want to give speakers arbitrary work to counterbalance our own laziness. This is why ConFoo will continue asking only a title and an abstract for each proposal submitted.

Previous: Anatomy of a Bully Next: How skiing made me a better developer