#btconf Berlin, Germany 07 - 09 Nov 2016

Sacha Judd

Sacha runs the Hoku Group, a family office combining private investments, early-stage tech ventures and a non-profit foundation. She is the co-host of Refactor (a series of events around diversity in technology), and runs Flounders' Club (a network for early-stage company founders). Sacha's first computer was a Commodore Vic 20, and she’s determined that the next generation of young women will have as much fun with technology as she does.

Want to watch this video on YouTube directly? This way, please.

How the tech sector could move in One Direction

Sacha will take us on a whirlwind tour of boybands, conspiracy theories, fake identities, appalling statistical method, and what it might teach us about the future of our industry.


Sacha Judd: Thank you for such a warm welcome. That’s my Twitter handle there, so keep in touch.

In 2010, five boys auditioned as individual contestants for the British reality TV talent show The X Factor. They didn’t make it through to the final round, but Simon Cowell, being an evil genius, put them together and formed a boy band and entered them in the groups category. Harry, Niall, Zayn, Louis, and Liam became One Direction. They didn’t win. They came in third. But they were signed as recording artists anyway.

Five years later, the band had performed over 360 shows in 36 countries and sold over 8.2 million tickets. Their 2014 tour was the highest grossing tour by a musical act generating over $280 million. They were the first band in U.S. history to have their first four albums debut at number one.

And despite being one of the highest grossing, most successful musical acts of all time, I didn’t pay them any attention at all. I couldn’t tell them apart. I certainly couldn’t have told you what any of their songs were. But, I like really strange rabbit holes on the Internet, like the stranger the better. Click holes that you can fall into for a day at a time and then emerge blinking into the light.

And a few years ago I fell in a really strange rabbit hole on the Internet about One Direction and, in particular, about Larry Stylinson. Larry Stylinson is a portmanteau of the names of band mates Louis Tomlinson on the left and Harry Styles on the right. And it’s the name or a conspiracy theory that says that the two boys have been involved in a secret, closeted, gay relationship since the band was formed.

I spent a full day clicking around and marveling at analysis posts and grainy zoomed in paparazzi photos and proofs that the boys may or may not have kissed or held hands or just looked at each other for more than a second. It was amazing. Like I was completely slack jawed. I raved to all my friends about it. And then I forgot all about the Larries.

Fast forward to December 2014, and the tech podcast, Reply All, does an episode about, of all things, the Larries and the theory at the time that Louis’ then girlfriend Eleanor was fake and a cover for the fact that he was still secretly dating Harry. So some of my friends heard this episode and they pinged me straightaway going, “Is that that weird conspiracy theory you were banging on about a couple of months ago?” I was like, “Yes, the Larries. I’d forgotten all about them.”

Meanwhile, in my real life, I’d made a seed investment in a company called Vestada, which is doing some really amazing stuff with log-on anomaly detection and contextual authentication for SaaS companies. But when they first started out the founders were offering a service that did a cloud-based backup of your cloud services. So you could back up your Dropbox or your Google Drive or your Tumblr in such a way that you could restore files if you’d accidentally deleted them.

And it’d only just started offering this service, and they hadn’t done much in the way of promo at all. And they woke up one morning to find that their sign-ups had gone off the charts. And when they put out the fires and tried to work out where the new traffic had come from, it turned out it was the Larries.

So Tumblr had started doing automated DCMA takedowns of infringing audio content. And a couple of Larries had had their blogs deleted and were panicking. And so some said, “I use this service to back up my Tumblr,” and so panicking fellow fans had hit the site in droves. I was amazed and delighted because how often does the Venn diagram of weird Internet rabbit holes you once fell in and companies you’ve invested in overlap? Like never.

So the next week I went off to Webstock, which is an amazing conference much like this one. It’s held in Wellington, New Zealand every February. And so if you’re going to come and visit us in New Zealand, you should come in February because the weather is really good, and you can come to Webstock.

So I went off to Webstock, and I wanted to keep telling people this story because I loved it so much, but it was a tech conference and everybody there new even less about boy bands than I did. So I had to start every time back at the beginning, like, have you heard of a band called One Direction? And then I became a bit like a weird door-to-door canvasser, like, have you got a moment to talk about One Direction?

Now, of course, I could just link to a Vox explainer, but I was an early adopter, so I didn’t have these kinds of fancy resources at my disposal. So my friends started teasing me about always talking about One Direction, and I like to embrace that kind of mockery, so I started replying to Tweets with nothing but GIFs One Direction. And from there things started to spiral a little bit out of control. I had a fight with one of my friends. He sent me this One Direction sticker book in the mail to make up. Even my mother asked me if I had embarked on some sort of elaborate practical joke, but she did buy me this for Christmas.

And underneath all of this hilarity there was this weird hint of concern trolling going on, like this was all fun and games if I was being deeply ironic. But people kept saying to me, “You don’t really like them, do you? Like you haven’t actually become a fan of a boy band.”

And unfortunately I’m also the sort of person that if you tell me I can’t do something, then I have a really knee-jerkreaction to that. That’s why I’m terrible at dieting. If you say I can’t eat carbs, then it’s basically all I want. And if you tell me that I can’t like a boyband, then it doesn’t matter that I still can’t tell three of them apart. I’m going to be the biggest boy bandstand you have ever seen.

It annoyed me. I realized we do this all the time: dismiss out of hand the things that particularly young girls get excited about. Laura Moss wrote a great article last year called Why Must We Hate the Things that Teen Girls Love? And she talks about the way we belittle things like One Direction and Twilight and Taylor Swift. How it’s okay to like a mindlessly bad action movie, but not a simpering romance. And she points out that that’s not a new phenomenon. At the height of the Beetles popularity, Paul Johnson writing in a New Statesman said that those who flock around the Beetles, who scream themselves into hysteria, whose vacant faces flicker across our TV screens are the least fortunate of their generation: the dull, the idle, the failures.

And the gendered aspect of it just kept bugging me because it seemed like it was okay to scream yourself hysterical about some things and not others. And it also seems like it was okay to be obsessive online about some things and not others. This obsessive project, which is amazing, took someone 18 years to complete. They tracked down every single song ever sampled by the Beastie Boys: 286 tracks, 22 hours of music, all beautifully encoded and available for download. And this project was shared really widely on social media. People were like, this is amazing, like, what an extraordinary archive.

But this collection of 3,419 video clips of One Direction, every live performance, every interview, every concert, well that’s just crazy, right? As an aside, when I needed a video clip to remake a GIF for this presentation, I sent an email out to Megan and asked her if she knew where, you know, which clip it was, and she sent it back to me in three minutes based just on a photo of Louis’ outfit. Like that’s an archivist right there.

But then something even more interesting happened. While I was spending all my time on Tumblr trolling for One Direction GIFs with which to troll my friends, I started marveling at the really creative work that the Larries were doing online. They weren’t just obsessively cataloging 11,000 photos of Harry Styles’s hair. They were, for example, cutting together video clips of their favorite moments.

[Video Played]

Sacha Judd: They were creating really gorgeous digital art of their favorite boys. They were manipulating photos so that their favorite boys appeared to be together even when they hadn’t been. They were creating their own memes. They were writing. Oh, my God were they writing. And this fiction was long, so these are the word counts of some novels that you will be familiar with. If you haven’t yet finished your first novel and it’s still sitting in a drawer, this is to give you an idea of the length of, you know, your first novel should be somewhere around the 100,000 to 150,000 word mark. And now so classics of the One Direction literary cannon. These are fully wrote novels with thousands of readers and comments and fans.

And then you go up a level because you get fan art about the fan fiction. So this is a fan produced trailer that one fan has made for a novel that another fan has written in which Louis is an English fashion designer and Harry is an American football player.

[Video Played]

Sacha Judd: I started to realize that this was an incredibly passionate, creative, engaged online community. Unsurprising, really, given that the band’s popularity was fueled in part by social media. Last year when Twitter turned ten, there were a lot of retrospectives written about the platform. And most of them overlooked the fact that six of the top ten most retweeted tweets of all time were from One Direction members. And the second most retweeted tweet in history, second only to Ellen DeGeneres’s famous Oscar selfie, is this one, a tweet in which Louis declares his love for Harry, retreated over 2.2 million times.

Last year project No Control started as a way to celebrate a song off the One Direction album Four, at a time when fans were a bit discouraged because Zayn had left the band and they felt that management wasn’t doing enough to promote the album. Four fans took it upon themselves to run with this idea of releasing the song, No Control, as a single. They used an online social media tool called Thunderclap, which describes itself as a bit of an online flash mob. So when you join a Thunderclap on the appointed day and time, the same message goes out over all your social media platforms, so Facebook, Twitter, Tumblr, and so on.

Within a week, these fans had organized the fifth largest Thunderclap ever. It had over 34,000 participants and a social media reach of 55 million. They oversaw a system of hashtags, mass streaming of the song on Spotify. They arranged for fans to buy the song and give it to other fans who couldn’t afford to buy it for themselves. They united fans to call into radio stations and request the song. And it was an amazing success. It resulted in worldwide airplay and mainstream media attention from the likes of Billboard and Pitchfork.

[Video Played]

Sacha Judd: I realize that this engaged, amazing, creative, online community also had enormous reach. Anna Leszkiewicz writing in the_Independent_recently said, you know, these girls are alchemists. They’re creating something out of nothing, uniting hearts and minds over miles. We might think of them as stupid or hysterical, but actually they are the sole engineers of this band’s unbelievable success.

One day it hit me. I spend a lot of time thinking about the pipeline problem as it relates to engaging young women in technology related careers. And a lot of the solutions that we’re playing with seem like quite blunt instruments: coding camps and programs in schools. I realized I was spending all this time thinking about how to get young women engaged with technology, and I was ignoring the fact that they already were. These girls were video editors and graphic designers. They were teaching each other CSS in order to make their Tumblr themes look more beautiful and using Chrome extensions in anger to force Tumblr to do what they wanted. They were essentially already front-end developers and social media managers. They were absolutely immersed in technology all the time, and we weren’t paying any attention because they were doing it in service of something that we don’t care about. Worse than that, because liking One Direction was embarrassing, most of them never told anyone in their real lives that this is what they were doing online.

So I decided I had to find out more. I became a Larry. I created my own online identity. I made a Tumblr. I started reblogging really cute photos of Harry and Louis. And I made friends, some really great friends. A shout out to my group chat Larries who have helped me with this talk.

It took a couple of weeks because you can’t just turn up on Tumblr and asked a question. It’s like being an egg on Twitter and hoping that someone sees what you have to say. But after a few weeks, I had a tiny number of followers who might have been prepared to share something that I had made, and so I threw together a survey. Those of you who do research for a living should look away now because you will weep at my scientific method. I’ve had over 600 responses to their survey and I promise, if I had any idea that the response rate was going to be that high, I would have agonized a great deal more over the questions. But hopefully you’ll get the idea.

First of all, I asked them what platforms they use to participate in fandom. And so you can see asking the question on Tumblr naturally skews the demographic. AO3 or Archive of Our Own is the platform that hosts transformative fanworks, and I’ll talk a bit about that again in second. So that’s where fan fiction is stored, and then a range of other platforms.

What fanworks have you created or contributed to? Again, you can see writing is the main thing, but then art and a range of more technical contributions as well.

So then I asked them, had they ever considered a career in technology, and overwhelmingly the answer was no. So I gave them a free text box and I said, “If not, why not?” And the answers nearly broke my heart.

  • “I’m tech savvy, but I’m self-taught.”
  • “I don’t have the education.”
  • I wasn’t even really made aware that tech was a field of work or study. I couldn’t name a job or position.
  • I don’t think I’m good enough to get hired over someone else.
  • “I’m intimidated by technology.”
  • I taught myself some HTML, but as a woman you have to be 100% certain and willing to fight for it.
  • I don’t think I’d be any good at it. A dumb, little, four-minute Larry music video won’t cut it in the real world.
  • “I don’t feel skilled enough to enter the industry.”
  • “I don’t think I’m talented enough.”

Bear in mind 70% of these respondents were age between 15 and 24. All but five of them identified as female or non-binary. And 65% identified as a sexuality other than straight. That’s our diversity pipeline right there, glued to their computers, and not telling anyone about it.

And of course it turns out they’re wrong. We’ve thought for a long time that our young women under-perform young men in STEM related subjects. But in 2014, the national assessment for educational progress in the United States undertook the first ever nationwide assessment of technology and engineering related literacy. Eighth grade students were presented with a range of real world challenges designed to test their competence with technology and engineering. And then they were asked questions that involved them solving problems and using technology to communicate and collaborate.

Eighth grade female students outperformed their male peers across the board. We also now know for certain that diverse teams build better software products. The New Yorker recently did an amazing profile of a hackathon held by the education startup AltSchool in the United States. And the startup uses a great deal of technology in the classroom, and so the hackathon was designed to come up with solutions to help teachers.

And at the end of the weekend, one of the teams was presenting their solution, which was designed to easily allow a teacher to timestamp the video feed so that they could come back and look at a clip later. And the team presenting their solution demonstrated how easy and unobtrusive the gesture was. All a teacher needed to do was knock twice on their phone and it would bookmark the video clip.

From the back of the room, someone put up their hand said, have you tested this with any women? Everybody laughed. She’s like, no, I’m serious. Most of our teachers are women and they carry their phones in different places. And the bookmark team, all of whom were men, looked deflated because, in coming up with this apparently elegant solution, they hadn’t envisioned the teacher having to slap her butt in class to bookmark a video clip.

We see it all the time with software products. We even now see it with hardware products. VR headsets that make women nauseous or touch screens that don’t take account of the fact that we have a low core temperature. Eighty-five percent of all consumer purchases are made or influenced by women, but 85% of all product designers and engineers are men. That’s just talking about gender.

When Pokemon GO swept the world recently, it didn’t take long for some players to work out that something was going awry. The PokeStops and gyms were appearing with much greater frequency in white neighborhoods. Think Tank Urban Institute of the United States found that the difference was really stark. Looking at data for Washington, D.C., they found on average in a white majority neighborhood there were 55 portals compared with only 19 in a black majority neighborhood.

Why? Because Pokemon GO’s starter set was based on the earlier game, Ingress, and Ingress was played by techs early adopters, who tend to be young, white, English speaking men. And Ingress’s portal criteria biased towards business districts and tourist spots, and the net effect of that was that the portals wound up in white neighborhoods. And in choosing to launch Pokemon GO off the same data set, nobody either noticed or cared enough to address the bias.

All studies show that companies with racial, gender, and ethnic diversity do better overall. And yet we’re still struggling to change this. And one of the things that we keep getting pointed to is this pipeline problem. That it’s really hard to find qualified, diverse candidates, and we don’t want to lower the bar. I love this epic typo fromHacker News. “If a great man comes along hire him. If a great woman comes along hire him.”

So here’s my challenge. How do we get these young women to see themselves as skilled enough and talented enough to enter our industry, and how do we change the way we think about hiring and qualifications so that there’s a path for them to join us? We’ve already seen fan writers enjoy runaway success as real authors. So what’s the equivalent of that?

A few moment ago I talked about Archive of Our Own or AO3. And it’s worth talking about that project for just a second. The archive is a noncommercial, nonprofit, central hosting site for transformative fanworks like fan art, fan fiction, and videos and so on. It went into open beta in November 2009. By the end of last year it held over two million uploaded fanworks and over 22,000 different fandoms and has over 900,000 registered users.

And the site was built entirely by volunteers from fandom, many of whom developed their first skills in coding, design, and development through their work on this open source project. When the project started, they asked a bunch of volunteers who had never coded before to read an introduction to both Python and Ruby and then to write a short program in each. And then they asked them about their experiences with each of the languages. And they chose Ruby as the language for the project based on that experiment. When the project went into open beta, it had over 20 contributors, all of whom were women, and it remains a shining example of an open source project that has managed to create an inclusive environment and welcomes people with nontraditional experience.

In a similar vein, this is my friend Maciej, who Marc mentioned. This is him in Havana when we were there a couple of weeks ago. And it was 34 degrees Celsius. But if you follow Maciej on Twitter, you know that he refuses to wear shorts. So this is him toughing it out in jeans in Havana.

A couple of years ago he gave a great talk called Fan is a Tool-Using Animal, and you can read it at his website. And it was about when Delicious, the bookmarking site, made some structural changes that made the site unusable for fans who had been using it to meticulously tag their fan fiction collections. Maciej also runs a bookmarking site, and so he tweeted out and said to fans, what features would you need if you wanted to come and use Pinboard?

And what happened next was amazing. You should read his talk, but over the next few days dozens of fans came together anonymously to collaborate on a Google document and produced for Maciej a 52-page technical spec of the features they were after. As he said, having worked at large tech companies where producing a tech spec involves shedding tears of blood in a room full of people whose only purpose is to thwart you and waiting weeks for them to finish, he couldn’t believe what he was seeing. It was like a mirror world to YouTube comments where dozens of people came together in love and harmony to produce a complex, logically coherent document off the back of a single tweet.

But I think the problem is most fans don’t value their skills in part because they’re self-taught. But let me ask you a quick question. How many people in this room, whether you come from a design or a div or a business background, how many of you started out on a self-taught pathway with computers? Right. Me too.

This is my first computer, a Commodore VIC-20. I had a 16K expansion pack and a tape deck, and it was awesome. I taught myself Basic out of the back of a magazine. I taught myself HTML several years later to build my first website. Check out those sweet tables. You want to know what it was for? A place to host my fan fiction. Want to know how many people knew about that in my real life? None. Zero. I have never talked about this publicly until today.

I was socialized out of using computers at high school. There were no girls in the computer science classes, and it wasn’t cool, and I was desperate to fit in. So I left school and became a lawyer and spent the next 20 years masquerading as someone who wasn’t part of the tech sector even though literally all of my time was spent online. And I can’t begin to tell you how common that is.

So what if your first experience of code is cutting and pasting something to bring back replies because Tumblr took that function away and broke your experience at the site? That’s what’s going on here. One fan is teaching others how to do that. Is that any more or less valid than divs who cut and paste from stack exchange all day long?

So what if your first experiences of online environments were places like MySpace or GeoCities? Or if as Auryn says, you started out in WordPress with simple themes and moved into more complex themes, and eventually into plugin development? Like is that any more or less valid than our standard hacker narrative?

Auryn gave a great talk recently about the language we use to describe roles in tech, how wizards became rock stars and then ninjas who crash code. But also, and crucially, how we manage to make people who haven’t followed a traditional path feel excluded because they haven’t been programming since they were four or because they don’t know the right programming language or, God forbid, because they use the wrong text editor? Yeah.

It turns out there might be some really compelling reasons why someone hasn’t been coding since they were a kid. I’m not sure if you can see that? Oh, no, it’s big enough. So the cost of a VGA card and a compatible CRT monitor in 1989, in today’s money, $2,800.

So when we’re hiring for junior graphic designers and junior front-end developers and photo and production editors, what do we need to change? I think the first thing we hit is a problem with pattern matching, which is when we look for traits in candidates that we think point to successful hires because we’ve seen those traits work before. But as Anil says, it’s like trying to put together the Avengers and only hiring Hawkeyes.

Pattern matching crops up when we’re reviewing CVs because we recognize certain companies that people have worked at or computer science programs that we’re familiar with. And it crops up again when we’re interviewing. It’s not so much that we try to hire people who actually look like us, although we do that too. It’s that we try to hire people who have been on a similar journey to us, who are interested in the same things that we are, who pass our invisible culture tests like would we want to get a beer with them or sit on a plane next to them for 12 hours or, in my case 24.

And we’ve started to use some words to describe this. We talk about unconscious bias and implicit bias, and already those labels are seen by some people as code words for being called out for being sexist or racist. But actually they’re cognitive tools that our brains need just to get through the day. Like we’re bombarded with so much information that our brains are making all of these tiny assumptions faster than even we can recognize. So if we think about implicit bias as an insult or as something that we can fix by going to a single HR seminar, then I think we’re missing the point.

Project Include is a group effort aimed at improving diversity and inclusion in tech companies, not just in terms of gender, and not just in terms of hiring, but company culture and remuneration and other things. It’s a great site. It’s got lots of amazing detailed resources, ands it’s got very specific recommendations, even if you’re a very small team, a young company. If there are only three or four of you, there are still some things you can be doing now.

And there’s a couple of recommendations that they make that I think are helpful in this context. The first is about trying to eliminate bias when we are evaluating CVs. So studies have shown systemic, unintended bias when you look at a bunch of CVs that are identical in all respects other than a name that signifies a racial background or gender, or an entry that signifies LGBT status.

Venture capitalist John Greathouse got in a pile of hot water recently for this tone-deaf op-ed that he wrote for _The Wall Street Journal_in which he acknowledged this unintended bias, but then suggested the way around it was for women to hide their gender in order to get ahead. I’m not sure how long he thought that would last because presumably you’re going to get busted when you turn up for work.

The article met with predictable ridicule. And to be fair, he’s apologized for it unreservedly. He called it dreadful once it was pointed out to him how terrible it was.

But we can use technology to take those kinds of signifiers off CVs, and there are services that will do that for you. But I think the more interesting question is, rather than just removing these signifiers, what would it look like to really hire blind? Not just to take those things away from CVs, but to do away with CVs all together and just hire people who can do the job.

There’s a startup out of Y Combinator called Triplebyte that’s trying this with programmer interviews on the basis that they think most programmer interviews are broken. And as they say, credentials can’t be a proxy for talent. Education and work history might be meaningful, but if you rely on that, you’re going to miss out on good programmers. So Triplebyte sits between the companies that are hiring and the candidates, fits the programmers themselves, and then recommends them to the businesses that are hiring. And importantly, they’re not that interested in your CV.

GetJumpers is a company that lets businesses who are hiring kind of do this themselves: host blind auditions. And what they see in their data is that women are taking blind auditions at a rate comparable to their representation in the general population, so 54% of people taking blind auditions are women. Women make up 60% of the top performers in blind auditions. And they find their blind auditions are resulting in candidates applying who come from really diverse educational backgrounds, so state schools and universities, community colleges, online colleges, coding boot camps and, yes, self-taught candidates.

So GetJumpers says, you know, if you interrupt the simplicit bias in real time, you’ll wind up hiring from a more qualified, more diverse slate of candidates. But if we’re serious about changing this, we also need to think about the location of our job advertisements. Where are we looking for candidates? So project Include notes that if keep advertising in traditional networking communities and the same sites, you’re going to get the same kinds of candidates applying.

The point is not to hire diverse candidates just because they’re diverse. The point is always to hire amazing candidates. But if only men respond to your job advertisement, you have to ask yourself, did I place it somewhere that only men saw it, or was there something about the ad that meant that only men wanted to reply to it?

What about instead of saying “seven years experience required,” we said something like “experience solving these kinds of problems,” and then listed them? Do you really care if a candidate has acomp-sci degree, or is that just on your wish list? Because as soon as you put in the advertisement, you’re going to be turning a bunch of people away.

Textio is a service that will run your job advertisement using machine learning against a database of 40 million other ads and give you a score on their platform as to how well it will perform before you publish it. And what their data shows is that a high score on their platform will result in you getting more qualified, more diverse candidates and then your positions being filled more quickly.

Buffer changed only one word about their job advertisements. All of their developers used to be called hackers, so they had front-end hackers and back-end hackers and iOS hackers and traction hackers. And they chose the word deliberately when they were a young company starting out because they felt like it was the best way to describe what they did. Their CTO has said that, to them, the word “hacker” meant someone who just could do the job well and fast. They didn’t necessarily have a computer science degree. They weren’t necessarily skilled at mind games or puzzles. So to them the word “hacker” was an inclusive label.

But as the company grew, they saw the number of women applying for jobs at Buffer declining, and it got as low as 2%. So they dug into this challenge and tried to work out what was going on. And what they found was that word “hacker” was turning people away. They didn’t feel like they could identify with that. So the company discussed it internally and came up with a long list of ways to describe what it was that they did, words like builder and maker, engineer, architect, code experimenter. And in the end they went with developer, but they also recognized that they had quite a lot more work to do. Now when you go to the Buffer website, you can find this real time diversity dashboard. So as a candidate, you can see at a glance what kind of company you might be joining.

If there’s one thing that I know about diversity, it’s that it doesn’t have a single solution. We need to work on culture and unconscious bias, but we also need to make our computer science degree courses practical and welcoming and interesting so that people want to take them. And we need to improve technology education in our schools so that our kids are as excited about a career in tech as they are about anything else. And we need to improve access to technology in our communities in the first place. But maybe when it comes to just this one part of our problem, our pipeline, maybe we also need to recognize that it’s already full of passionate, dedicated, creative, hard working, awesome young women online, and we just need to find a way to include them.

Thank you.