Twitter didn't cause the SXSW audience revolt


While it's stupid that the biggest story to come out of South by Southwest (SXSW) Interactive was the gossip over the interview of Mark Zuckerberg by Sarah Lacy, the one "hook" that has kept the story going is the suggestion that it was the use of twitter, in particular snide comments on twitter, which turned the audience against Lacy, the interviewer from Business Week.

There have even been comments (from those who weren't even there) suggesting witch hunts and misogyny. Other bloggers used hyperbolic terms like "train-wreck" and "career-ending" which are serious exaggerations.

Short summary. In a "keynote" interview, Lacy, who has just finished a book about Facebook, was on stage to interview Zuckerberg. Zuckerberg was, as usual, a difficult interview subject, but for a variety of reasons the character of the interview changed as the audience turned against Lacy, cheering criticism of her. Most agreed they had not seen somebody lose an audience like this in some time. Many people at SXSW were making above-average use of twitter, both for commenting on panels, and to find out what parties people were at. I liked it for that use, as the evening activity is very spread out. Most people only follow a small set of friends on twitter, though it is possible to track keywords and hashtags, but the volume on that can be very high, especially at a very twitterying event. There are a few twitter-celebs who get followed by thousands.

And indeed, there were what amounts to "heckles" on twitter during the interview, including some nasty ones. This is not unusual. People seem to treat a twitter as more like nudging your neighbour to make a comment than a public declaration.

However, I have also been to conferences where there is far more backchannel. There are conferneces where everybody has a desk and laptop, and many use an IRC channel to chat about what's going on. SXSW had official chat rooms using Meebo. There have even been conferences where the backchannel is displayed on one of the projectors, so everybody but the speaker sees it, and this becomes a channel of constant heckling, jokes and comments.

But even these much bigger backchannels don't result in what happened to Lacy. She lost her crowd the old fashioned way. Those who were in the room, rather than those who watched the video, know that her mistakes generated good old-fashioned murmurs, neighbour to neighbour in the audience. An upswell of "why is she conducting the interview this way" had already built. Thus when Mark Zuckerberg made his first mild negative reaction to one of her questions, the audience broke out into cheers.

That first cheer broke the dam. Now everybody in the audience knew, "OK, it's not me, a lot of the room finds this interview odd." There was no recovery without a complete about-face of interview style, but Lacy was surprisingly oblivious to how she had lost the audience.

Now, if you are curious as to why she lost the audience, mainly it's a combination of personal style and simply doing the wrong interview for the crowd. SXSW is full of web developers and designers and plain old users, with far fewer venture capitalists and business reporters. But she did the interview you might want in Red Herring or TechCrunch. The questions seemed to be all about valuations, and deals, and what it's like to be a young billionaire. About management style and the personal styles of other executives in the company. The questions were daring (ie. on the edge of rude) and strangely flirty at the same time. It was bizarre, but might have gone over at an investor/CEO conference.

However, this audience wanted to know about Facebook the product and Zuckerberg the product designer, not Facebook the company and Zuckerberg the boy billionaire. They wanted hard questions, indeed, but about Facebook flaws, and privacy issues, and applications. When they got the mic, that's what they asked. Where they would have expected Lacy to be their advocate and proxy, digging out the information they wanted to know, they instead found her asking about what they didn't want to know, and frankly talking about herself far more than they should.

So after a dozen rounds of "Wha?" in short mutters -- from both women and men -- it eventually broke out as a swell, and twittering played a minor role, if any.

Was it sexism? As I noted, I saw as many women as men complaining. She did a few things that raised this spectre -- a strange flirtatious style that a man could not use, and even the women commented on her choice of a short skirt with associated leg flashing while up on a raised stage. But SXSW has a very solid female population, far more than the typical computer conference, and would be less likely, in my opinion, to errupt in a populist sexist witch hunt than the the typical conference.

As backchannels grow, they might well make it easier to lose an audience. They can embolden hecklers who now will have the courage to know that others share their view. But to lose a whole audience you have to have to turn them against you in bulk.

I'm not a big fan of the backchannels. They are fun, but largely distracting. Indeed, today we seem to see rooms where only a small fraction of people are paying attention to the speaker. Of course if conferences had fewer boring speakers, people would not feel the need to divert their attention to E-mail and chat. I built a backchannel system once called the "Question Authority" that lets people submit questions for the Q&A, and other people can vote the questions up and down, with the top voted questions getting the mic. It's a good way to avoid the really stupid questions. We used it several years ago at the Foresight conference. I have heard that Google has recently developed a tool like that for their internal all-hands.


The Google tool's been used for over a year now, both for all-hands (although that's a bit more recent) and for submitting questions to Tech Talks (particular from remote sites) and major visitor talks such as the Presidential candidates (for the latter, the top-ranked were passed on to the interviewer, rather than being asked by proxies at the mike). It also allowed you to comment on a question, although those were only visible if a user chose to make them so.

One problem with it, at least for the version being used when I was last there, was that there was a definite advantage to submitting your question early on, since more people would then see it and have a chance to vote on it. If you submitted something right before (or during) a presentation, you probably wouldn't get enough votes to break into the top tier. I did suggest to the person working on the next version that perhaps something could be added so that if you'd voted, you'd get a reminder email N minutes before the event's start to go back and vote for any newer questions that'd been submitted since your vote.

Add new comment