On diverse speaker lineups at conferences

There’s been a lot of discussion on Twitter about diversity in speaker lineups at web conferences.

Lea Verou wrote about the blindness of blind reviews, and while she was looking at the process of anonymized reviews in general, some tweets in and around this conversation debate whether anonymizing the speaker selection process eliminates bias. Zach Leatherman tweeted:

I don’t believe an anonymous conference speaker selection process eliminates bias any more than “not seeing race” eliminates racism.

I think Zach makes a good point. Not only does an anonymous selection not eliminate bias, but the only thing it guarantees is that the curators don’t know who they’ve selected.

This does not guarantee a diverse speaker lineup.

What anonymizing does is put a responsibility-absolving blindfold on biases, so one can inadvertently end up with a totally non-diverse lineup and then can say, “Hey, it was a fair selection because we anonymized.”

That’s a weak position, in my opinion. And yes, I know that some people “de-anonymize” in round 2 of their selection process. Kudos. For now, we (the team I work with) prefer invite-only for events with a small lineup, but that’s a post for another time.

We all have biases. But when it comes to biases regarding people, it’s what we do with them that counts. One approach could be to face those biases and consciously act against them in order to get the type of speaker lineup we want. What lineup that is might be different for everyone. You want an all-female lineup? Great. 50/50 split? Also great. Choose what you want and take responsibility for it. Own the selection process rather than hiding behind it.

One possible approach

I don’t have all the answers, but I’ve been a part of conference organizations for many years and maybe the approach we used for two editions of dsgnday and this year’s CSS Day could be useful to some. Again, I’m not saying “we did it right”. I’m just saying that we owned our process and got the lineups we wanted, and someone out there might benefit from our approach. It might not fit with your idea of the right approach. That’s fine. Anonymized curation is okay; I simply have reasons for not preferring it. Please read that again. Also, note that our team consists of four men. This is purely coincidental, something we’re acutely aware of, and is part of the reason we’re thinking so much about these issues.

We started with the goal of the conference itself, of course, which is a range of topics on web design in a broad sense of the word, since the day is meant to be something of a counterweight to the larger number of developer conferences we have here in the Netherlands. However, that doesn’t mean “no code”, so we have a huge base of speakers to pull from, including non-technical speakers.

Right away, our longlist included people we knew would be a good fit for the conference. Krijn, PPK, Martijn, and I kept a simple text file in a Dropbox that any of us could add to at any time when we thought of someone we thought would fit. (Disclaimer: I’m speaking at the event, and while I hope my talk is well-received, that’s in part a financial move since it’s not the most-attended conference in Europe and paying for speakers and their travel is very expensive.)

Setting the goals

While some conferences have a primary or secondary goal of creating a platform for new or less experienced speakers, we currently do not. It’s perfectly valid to require a certain level of speaking ability or experience for an event. There are plenty of calls for papers and other platforms for willing new speakers at other events.

We had also set a specific gender diversity goal for this conference: at least a 50%/50% male/female split. An uneven split in favor of female would have been fine as well. Note that this is a goal, not a quota.

I have to admit that several years ago, I would have found this approach to be ridiculous. I’ve even openly opposed this type of thing. But years in the industry and hearing more viewpoints and a bunch of other factors led to me completely changing my mind. If I just say to myself, “Write out a list of potential speakers”, mostly men come to mind. That’s not a bad thing, but it’s a true thing. Call it a built-in bias. I might just happen to know more male speakers. But it is a fact, at least right now. And to get the lineup we want, we have to consciously combat that tendency. Setting the goal is the first step.

Battling built-in bias

So while I had a couple of people I really wanted to speak (not only men, BTW), I left them on the longlist and did the following thinking exercise:

What if the entire world only consisted of women? Who would I like to have come and speak at this conference?

Names came easily, because we had consciously removed our natural bias by way of a simple thinking exercise. And no one can say that these are “token” (I hate that term) female speakers designed to fill a quota, because they were the people we honestly thought would do a great job and fit the subjects we wanted handled at the conference. Note that it’s getting easier and easier to think of female speakers without resorting to bias-removal thinking exercises. This is partly due to us knowing more female speakers, and also because when you consciously and consistently work to remove bias, it slowly starts to go away. At least, that’s how I experience it.

Anyway, we had several names, and we curated the final lineup based on the quality of each speaker’s work and presentation abilities, and how their varied expertise would combine into a day-long program.

This process might not be great, and the conference is not a big one, but we are really proud of the three lineups of fantastic, qualified women and men we’ve put together so far (we love you, speakers!). And there is no way we could have guaranteed this result with an anonymized process. Without taking responsibility and making conscious choices.

It’s hard to do. That’s why they call it work. :-)