Read More about Anil Dash in “To Remake Tech, Remake the Tech Company“
MATT MULLENWEG: A lot of tech companies talk about prizing “people over profits,” but Glitch is a startup that is serious about these ideals, and holds itself publicly accountable for sustaining this commitment as the company grows.
That’s partially because Glitch’s CEO is Anil Dash. Anil’s an old acquaintance of mine — he’s one of the early pioneers of blogging. Over the last twenty years he’s developed a reputation as something of a tech prophet — not just for predicting what’s going to happen next, but for holding the industry’s feet to the fire.
Glitch is a partially-distributed company that runs a social platform for building and sharing web applications. To do that, they’ve developed a workplace environment that centers around employees’ well-being. I’m interested to hear from him how his company has aimed to go beyond platitudes and create a genuinely equitable and respectful workplace, and to learn where their semi-distributed structure fits into that goal.
Alright, let’s get started.
MATT: Anil, you have been blogging forever.
ANIL DASH: [laughs] Roughly, yes. In geological time, it’s short, but in human years it’s 20 years.
MATT: How did you start and what keeps you going?
ANIL: People in my life were tired of hearing me rant about things. So they were like, “Go put it somewhere else.” [laughs] And at the time I had a really long commute. I was commuting by train an hour and a half each way. I mean it was really — it was like three hours a day on a train and I was going nuts.
ANIL: You didn’t have Wi-Fi back then. So I had a giant Dell laptop and I was like, “I’ve got to learn how to do more with HTML.” I knew the basics but I wanted to do it. I would do it, literally just practicing on local on my laptop, on Internet Explorer 5 or something, whatever it was at the time, and I thought “Oh, I could take these rants in my head and put them out here onto the internet.”
Right about the time that I had that idea in maybe summer of ’99, I saw the first couple sites. I was like, “Oh, this is what I could do. I could organize it this way.” I saw Peter Merholz’s site, PeterMe, and then very quickly discovered a couple of others. And so these pioneers were doing it, and I thought, I don’t think I can write like them but I think I’ve got something to say.
It really felt like it was the right time too because I had started in maybe July and by September the Pyra team had built Blogger and the Danga team had built LiveJournal. Even just the fact that there was software to me meant “OK, this is legit.”
MATT: So 20 years later?
ANIL: 20 years later.
MATT: Why do you blog now?
ANIL: One, it’s part of how I think. My wife will always say, “You’re staring off into space like you’re writing something.” She just knows that it’s this thing where I’m collecting my thoughts. Certainly one of the most important things to me is, I think better and organize my thoughts better and share my ideas better when I write it, and it introduces a rigor to what I’m sharing. I love that push to accuracy and push to quality. It makes my thinking stronger.
Some of it’s just, I like to write. For a long time, I had no other place to do it. I was lucky, after I had written a million words online people asked me to write for things. [laughter] You know? I got a column in Wired and I was like, “Where were you all ten years ago?”
ANIL: But nobody was trying to hire me to write so I might as well put it out there. And that’s still true.
MATT: What would make you stop blogging?
ANIL: Well I’ve slowed down. So I would say the thing that slows me down is, well, life, right? So I’ve got to spend time with my child and I’ve got a company to run and I’ve got — the priorities have shifted. I can’t just say I’m going to stay up all night and finish this 3,000-word piece like I used to.
But at the same time, the biggest thing chipping away at it is having other venues and other platforms. I resisted doing a podcast for the first 15 years of the medium. [laughs] I think the week podcasts were invented somebody was like, “You should go do one,” and I was like, “Ah, I don’t know.”
MATT: And now you have your own little studio.
ANIL: Yeah, exactly. Then I started thinking about the craft that… The same is true actually of other social media. Like Twitter in particular I spend a lot of time on. And this is a strange thing to say but I think you’re probably one of the few people who can appreciate it — I care about being good at it. I think people are like, “That’s an absurd thing to say,” like, “Isn’t this disposable, isn’t this ephemeral?” I don’t feel that way at all. I have Twitter threads that have been going for six years. I am very mindful of how I use my audience, who I retweet and who I amplify.
So I think very much of that as a body of work too. I never delete tweets. Again, same thing, I am certain somebody is going to go back and be like, “You said this thing that’s terrible,” and hopefully I’ve learned since then. But I very much want there to be a body of work between all the things I’ve done that I look at on a years — and now decades — timescale. I don’t think very many people look at their YouTube channel or their Snapchats [that] are like “Yeah, how is this going to age in 20 years?” But I have that luxury so I try to do what I do with that in mind.
MATT: As you mentioned, you’re running a company now — Glitch.
MATT: How is that culture of blogging or writing part of Glitch’s culture?
ANIL: So Glitch is the latest name and current incarnation of a company that started as Fog Creek Software in the year 2000, founded by Joel Spolsky and Michael Pryor. Joel is, to me, one of the all-time legends of blogging, one of the greatest people to ever do it. Joel started Joel on Software in ’98, I think.
MATT: Super early, yes.
ANIL: Yes. I would read Joel’s blog and he would talk about a company where they cared about the software they made, cared about how they treated their people, were very thoughtful about the work that they did. They did not want to be just another dot-com, which at the time that was what was in vogue. I thought, “Wow, that would be an amazing place to work.” And also, “I would never pass their coding test.” That was the other thing I thought about the company.
And in late ’99 — December ’99 — they had a blogger’s dinner in New York City, which is a funny thing to say because the premise was —
MATT: Of all the bloggers.
ANIL: All the bloggers, right. And we fit around two tables in a Mexican restaurant. You fast forward over the next 10 years and they had built some products and really established a culture, and Joel reached out and it was just like, “We’ve got something to show you.”
I saw the prototype of what became Glitch. You could have live-in-your-browser code. And as you coded and typed your code out, it would live-deploy without you having to run anything, do anything, touch anything. You didn’t have to ask somebody down the hall for the access to the AWS account. It just worked. I still have the notes from that first meeting, and we were looking at them not long ago, we had had some folks join and I wanted to show them where it all came from.
That first meeting that we had of the demo of it was — we talked about, well, we need to have a social network wrapped around this so you can find the apps, and you need to be able to remix the apps so you can redo it, and we need to have multiplayer editing, so more than one person can edit at the same time. And we had this list, and we made these bullet points in the first hour, all of which we did, all of which are the heart of the Glitch experience.
It was really like few moments in my career. It actually felt a lot like when we first saw blogs. I first saw the blogging tools and it was like, “Oh, this is going to be it. This is… I don’t know if what we’re building is going to be the way it happens, but this way of creating the web is going to be how we make the web.” It was really, really clear.
I remember really early on, when I was talking to Mena Trott, who had led the creation of Moveable Type, and was one of the bloggers that influenced me most, with her voice and her tone. I had said, “Someday there’s going to be a million blogs.” And she just looks at me and she’s like, “You dumb ass, there’s going to be like a hundred million,” you know? [laughter] And it was just such a great, both affirmation, and also “This is so much bigger than we can imagine.”
I had that feeling with Glitch, which was — I think a lot of us had lamented the web we make. That we have apps on our phones, and what I always experienced with the apps on my phone. Everybody complains about [how] the algorithms aren’t fair, and there’s all kinds of awful content being shared, and misinformation and these things. What I miss most is, I look at my phone and I look at the apps and I don’t know who made any of them.
ANIL: It used to be [that for] every website I went to, I knew the person, and I knew what they liked, and I knew why they made it the way they made it, and what their peccadillos were. I mean, I still go to Daring Fireball and I’m like, “Damn it, John Gruber, the font is still too small. [laughter] I mean we are so much older now, just nudge the type size a little bit bigger,” you know? I’m sure he’s been finessing the font for 17 years to find exactly the right size but he’s not ready to commit yet. [laughter]
But ultimately I come away and I’m like “That’s fine, I know exactly why it is what it is, I know his personality is captured in every pixel on that screen.” And then the stuff I spend my time tapping on, on my phone, has none of that. It doesn’t have people behind it, I don’t know the peccadilloes.
I always use the analogy of food. Every meal I remember in my life was made by people I love who were — maybe it was a part of a tradition, like our culture. I know where the ingredients came from, I certainly know who was in the kitchen. And in none of my great digital experiences do I know who made it and can I thank them.
You and I both have had this experience where we see an app or a tech that meant something to us, or a website that meant something to us, and we get to meet the creator.
MATT: It’s so exciting.
ANIL: Yes. So I felt that all like a flash the first time I saw the prototype of Glitch. Then we spent a year and a half, almost two years, really remaking the company in that vein. So I joined, which was incredibly intimidating, to be the first non-founder CEO of this company after 17 years and 18 years in business. We had shared an office with Trello and we moved out on our own, which felt like being 18 years old and moving out of mom and dad’s house. We renamed the company Glitch. Joel was incredibly supportive.
And then the team, the people here… It’s so interesting because we’re not trying to be retro. I don’t have any idealistic view of the good old days, but I think there were things on the web that we used to like that we miss. So we imagined a modern version of that and everybody aligned around that. It’s really interesting where there are people certainly of my cohort who are like “Oh I remember the old internet and I used to go to MetaFilter,” or whatever the old thing was, but most of them are like, “Oh I’ve never known that web. The web I’ve grown up on was — the wildest part of it was MySpace and everything else has been this blue box we shove our photos into.”
And the exceptions are around the edges. Definitely I had a WordPress theme that I used a hack on, or definitely I had a Tumblr thing that I would try to mess with the theme of. But those were the edges. It is satisfying to make something, and very little of their experience of the web had been that, and so they wanted to unleash that for as many people as they could.
We’ve been fortunate. We came out of beta a little over a year ago. People have built well over three million apps on Glitch, and it’s going up very quickly, which is fun. [laughs] They are all full stack web apps, they are 99.99% open-source, they are all remixable. You can view source on all the apps — and full stack, you can even view the server code. It feels like a bit of the promise of the web that got so many of us excited in the beginning.
MATT: So someone could think of it like a largely open-source code repository where all the code executes as well?
ANIL: Mhm, yes.
MATT: You can see how things work and if you build something, I can click a button, remix it, essentially fork it, software terms, and start my own version?
ANIL: Yes, yeah, exactly. So for everybody, if you come to Glitch, I comment it’s almost like an app store. There will be something fun today that’s like “Build a Slack bot” or “Make a fun game for your friend,” or whatever it is — very simple stuff. And if you just want to go, take that and play, you’re great, it’s fine, you don’t have to be a coder, you don’t have to dig in.
What we find is most people are like, “This is cool, but I want this button to be blue instead of green,” or “My boss wants me to add this one feature to this little tool,” or whatever it is. Then that idea of, “I’m going to remix this and customize it and make it my own.” The fact that you could do it in the browser, you click remix, you get your own copy, and behind the scenes it’s just a cloning a git repository.
But what’s happening is the editor — it’s like Google Docs. You can do live, real-time editing together on the code, and you can share the link with somebody. As soon as you get in that mode where rather than “I’m going to figure out what this is on my own,” but I can collaborate with somebody — and especially that you can just do it in your browser, you’re not installing a dev environment, you’re not configuring an editor, you’re not doing all the complicated things, you’re just doing the creative part — I think that has been very, very freeing for people.
There is this magic happening behind the scenes where that app runs and then if you type in code and you make a change, it immediately is running with that newer version. Those are things that have empowered people to sit there and say, “Now I’ve got permission to do this again,” maybe in a way that if I’m an older-timer and I’m used to doing it on GeoCities or on MySpace or something back in the day, but that has gotten hard. I think you and I both started FTP-ing pages up to a website. That’s hard as hell to do now.
MATT: Oh my goodness. [laughs]
ANIL: There’s deploy scripts and there’s cloud configuration and —
MATT: It’s gotten more complicated, actually, now.
ANIL: Yeah, yeah.
MATT: It actually always surprised me, Moveable Type had this feature in WordPress too, where you edit the template in the browser.
MATT: How popular that would be and actually still is today. Like, we still invest in — I guess we build it on CodeMirror or something, which is an open source library.
ANIL: Yeah, so do we. And it’s essential.
MATT: Oh cool. People love those features.
ANIL: If you’re an experienced dev, you are going to have your whole toolkit, and you’re going to connect it all up. It’s the difference between a weekend woodworker or hobbyist, and the person who’s got the wood shop.
MATT: But collaboration is a super feature that none of the desktop tools that I’ve used have.
ANIL: Yeah, exactly. I think the biggest thing we saw is that you could — people already have the behavior. If you’re editing a Google Doc or whatever tool you use, you send them a link, or you Slack them a link, and you’re just in there. You don’t even think about it, you know, “What do you think of this,” “You might not —”
MATT: You’re not resolving versions, they’re just happening in real time, which I think is really powerful.
ANIL: Yeah. In the olden days when you would email Word doc final_final_REALLYFINAL_versionthree… [laughter] and you’re approving changes, it’s like, “Oh god, this is hell,” right?
MATT: Not old days, that’s current lawyer days, by the way.
ANIL: Yeah, yeah if it’s expensive, right. Legal contracts, sure. But the internal, for some announcement or something —
ANIL: — you’re not doing that. And then code was still feeling like that. I love GitHub, we use it every day, it’s an essential tool, but I — probably every single day — type “man git.” I’ve been using it for as long as its existed. [laughter]
ANIL: I don’t think I’m dumb, I think it’s hard. [laughs] You know?
MATT: Yeah, yeah.
ANIL: Especially when you have other things to do. If that’s all I did, I think I could probably learn it, but I have to do other things. So that idea of, while you’re just remixing something, it’s already handled, and then while you type and both of you or all ten of are typing in your Glitch project, we’re auto-committing into git behind the scenes.
Then the biggest thing for me, after we launched in beta, that turned the corner for Glitch, was we introduced a feature called Rewind. It’s a timeline slider at the bottom, you can pop it up at the bottom in a window, and it gives you the same thing you see on a YouTube video, but if you want to rewind back to old versions of your app, you just slide the slider back. And again, could I undo a commit and revert and do the right thing? Yeah, I can do it, I’ve done it. I don’t feel confident at doing it, and I certainly don’t do it casually. I wouldn’t just be like,
Let me try a different version of this and undo these things.”
MATT: Play around with it.
ANIL: Right. And again, I’m a spreadsheet nerd, in spreadsheets we do scenarios all the time. “What if it grows by 1.2% instead of 1.1%?” It’s trivial, right? So you are able to think more broadly. It un-bounds your experimentation, it changes the way that you’re willing to try something because there is a lower cost to it.
Rewind is about that. It’s “Let me try a different version of this app. I’m going to remix my own app with one click,” [and] three seconds later I’ve got another copy. Now I’m going to rewind back to last Thursday and “What if I’d used this other framework, tool kit, API, instead of this one? Now I can have them side by side and look at the difference.” That kind of creation, I have that in Garageband when I’m making music, I have that in turning off a stack of layers in Photoshop, when I’m working visually. I have that in all my other creative tools, and in code it’s way too hard.
MATT: The only thing I’ve seen close is maybe the computational worksheets from Wolfram.
ANIL: Yeah. I think they’re great. I look at the Jupiter notebooks, and in Python they get a little bit of that feel. Those are all adjacencies. I don’t think any of this is new. People have been exploring these boundaries for decades in coding culture and in computer science, but productizing it for something you could actually put a live web app on today instead of in the lab — I think that is the leap, that’s the thing we want to do, is take this out of…
We can imagine a million things, but can you do something that — you can use this and then ship a production app you’re using at work, like the actual… The thing we have a lot of — Slack bots for example — people are like, “I just want to get some data out of this whole legacy system and dump it into a Slack channel, and I know the connecting pieces are here.” So Slack, they created a bunch of examples that do the basic mechanics of it. Actually, Matt Haughey, an old mutual friend of ours —
MATT: Oh wow.
ANIL: — yeah, had built some of these examples.
MATT: And there is an amazing connection as well that the predecessor to Slack was Glitch.
ANIL: Mhm, yes.
MATT: In a different incarnation.
ANIL: Yes. Exactly. So the team had originally built Flickr, a lot of the same folks, and that started as Game Neverending, and they pivoted from the game into Flickr, which was a feature of what they were doing. The second incarnation of the game was called Glitch. And we had started…
Actually the first prototype that we had built was called Hyperdev. We went through a lot of iteration but I reached out to Stuart Butterfield and the team at Slack, and I was like, “Listen, you’re not using the Glitch name, we’ve got something good that we think is worthy, not just another generic photo sharing app, we’ve got some idea here,” and they believed us and they thought that was worthwhile, so they were very generous in cutting us a deal to get it. And we got the Glitch handle on Twitter and all that stuff, and that was it.
It’s funny because I’m always a little skeptical — names are names — the product, it’s always about the product in these things. But the truth of it is that changing the name was the biggest inflection point we had. As soon as that happened, people could understand what it was in a way. That was really surprising to me. In retrospect, I said, well if you get a Prince fan to run a company, they’re going to rename things, so… that’s just how it goes. [laughter]
MATT: I like it. Formerly known as.
MATT: One of the things that was really an undercurrent there was a real optimism about what could be built and everything. You were also an early contrarian, perhaps a pessimist or realist, however you want to define it —
ANIL: Or at least critical, for sure.
MATT: — on the neutrality or goodness of technology and communities and things.
MATT: Tell me about how — you have the receipts, so what were the receipts there, and what happened like you predicted?
ANIL: Like a lot of people who are changed by something, you have a crisis of faith. So I had been working on building blogging tools at the point I left that work, for seven-and-a-half years. I had joined a company that my friends had started that made a product that I loved. And by the time I left, I felt like none of that was true. They weren’t in charge, the company was unrecognizable, the products had fallen into disarray or decay, I wasn’t proud of a lot of the work.
More broadly, I had gotten very disillusioned with tech in general. I had been in San Francisco for, at that point, three years or something, and I would walk my dog around South Park where I lived, and hear three different people talking about their podcasting startups and I — [laughs] I was like I’ve got to get the hell out of here, I’ve got to go home, I’ve got to go back to New York, as I think happens.
There’s not some epiphanic moment where the clouds parted and all the sudden I saw this thing. I always follow my gut and my gut was like, “This is not good for me personally but something is wrong here, something is really wrong.” For context, that you will remember but that I think has been erased because our industry doesn’t have a great history, there had been a lot of effort around just before that — a thing called OpenSocial.
MATT: Sure, yeah.
ANIL: So Facebook had had its platform that had succeeded in people building apps for Facebook, which is funny because they killed that off too, but at the time it was the hot thing. Everybody else freaked out and said we’re going to respond to this. So the company I was at, we participated at Six Apart, and also MySpace was big in it and Google.
MATT: Evo, Google, everybody.
ANIL: Yeah, exactly.
MATT: Everyone except Facebook was a part of this.
ANIL: Right, exactly, right, this is a coalition of the willing. They had some big event at Google and all kinds of — I remember at the time thinking “Wow, these folks have a lot of money. And they did some big event and launched it.”
I had been very excited because I wrote about it. I was like, “Listen, the open web wins, and this is about the open web and Facebook is closed and it’s like the new AOL and it’s just siloed,” and all these things. The thing that got me excited was I wrote that blog post and it got linked to by a blog I had helped recently set up just before that, which was Marc Andreessen’s blog, and he said, “This is the smartest take of the day.” I was like, “Wow, the Netscape guy thinks I’m smart.”
MATT: That’s awesome.
ANIL: You fast forward not much later and Marc’s on the board of Facebook and OpenSocial is dead and in fact all of the work I had tried to do on OpenID and pretty much everything except OAuth of that era died. A lot of that was the user experience was bad, the technology was too hard, it wasn’t user-centric enough — there’s a lot of legitimate criticisms. But the motivations behind them were all legitimate, that no one company should own identity, that we should have ways to share identity pseudonymously between sites — there were a lot of really key pieces there that were dead on.
And that we should be able to control our information and where it goes and have a record. One of the important ideas that totally got dropped that was an underpinning of things like OAuth, was that you would have a record of where you signed into and where your data got sent to, and it would be this audit log. And you imagine how different that world would be and one of the —
MATT: TypeKey was the identity provider that had an open —
ANIL: Yeah, yeah that was… And we built a product around —
MATT: It was great because it would show you where you had —
ANIL: Yeah. And it would show you the history and also that we documented the protocol and said you can implement your own, and we never… It was a thoughtful design. I think, again, the user experience was too hard. There were a lot of legitimate reasons it didn’t work but there was a very informed, thoughtful perspective, and one of the last big projects I had pushed on was —
So the company I was at, at the time, we had Moveable Type and Typepad, which were two big, serious publishing platforms. And then we ran LiveJournal, which at the time was still one of the biggest social networks in the world, but it was probably the first to get to 10 million users.
MATT: I think also, in addition to pioneering so much, was largely open-code or open source?
ANIL: Entirely open source.
MATT: Open source. You created Memcached, a lot of things.
ANIL: Created Memcached, yeah, yeah, exactly.
MATT: That Mogile and everyone uses today.
ANIL: Yeah, exactly. There is this hugely generative platform. They invented a lot of — a functional social model and of the underlying technology and open source, it was really thoughtfully done. But we were adapting to the new world. And so Facebook reached out and said “Will you use this new feature we have, where people can share their activities on” — I think it was still “the Wall” at that point.
MATT: Was that Beacon?
ANIL: This was Beacon. So we pushed back and I leaned on our lawyer and we said “We’ll do it but it’s opt-in and our users have to be informed.” And it was actually, to my mind, the basics that you do. People don’t share their information without consent. We were the only launch partner of Beacon out of dozens, and the commerce sites and everything, who made it opt-in. And Facebook had pushed back really, really hard.
I was shocked because I had just assume everybody would do this. I thought, “Oh they messed up in their defaults but they… of course they want users to consent to sharing data with them.” I was astoundingly naive. In retrospect, I didn’t even feel stupid. I should have felt stupid, and instead I was just like, “I have misunderstood where I am.”
MATT: I wouldn’t put that on you. I would say that there was a general zeitgeist of openness and sharing with good intentions.
ANIL: Mhm, and that was the social expectation that we all had of each other.
MATT: Yeah, yeah.
ANIL: I mean, we were competitors, but we both believed, “Oh this is how you do these things.”
MATT: And we support the same standards and things like that.
ANIL: Yeah, yeah.
MATT: Facebook bootstrapped on exploiting that, whether that’s address book APIs or things like Beacon or whatever, it was that kind of — [they] took advantage of some of that good will in a way that wasn’t ultimately two-way. Although they tried to be two-way as well. Let’s be fair. Their platform, the API, a lot of the stuff they now get in trouble with, was to try to open up Facebook.
ANIL: Well, right. I think part of it was the reason they exposed data that people didn’t expect is they collected data people didn’t expect to share. Right? And talking to people who were there early at Facebook, one of the things that is interesting is they weren’t even intentionally exploiting — it had not occurred to them that we had these norms, they weren’t of the community.
We had all come up reading Joel and Software and Zeldman and Shelley Powers’ Burningbird. There were these people who were I think in retrospect we would call it “tech ethicists,” who were thinking about the impacts. And also, I think you and I both were like, “We want to be cool to these people.” You want the nod and you want them to link to your blog and all these things. And that healthy impact of them being skeptics and thoughtful, as well as those of us who were coming in, saying we want to get the nod from them, all led to, in good and bad ways, a set of norms, and they didn’t understand the import of — Zuck was not from that. So it wouldn’t even occur to him to try to conform to the norms of somebody he’d never heard of. Like, who would? You know? And so it was just like, “I’m a stranger in a strange land, I’m just going to do what it takes to grow or… I’m not saying, “Why didn’t you have the same values as us?” because, like I said, we got a lot of things wrong, but that was telling.
To your starting point here about how did I go from the cynicism to the optimism, and the truth is, I still have a lot of both. At that point, I reached a breaking point. I genuinely thought “I will never work in tech again, certainly never work at a startup again, I am done with social media as a business. I have to be here because I need it for my career,” but I got as far away as I could.
I came back here to New York, I started working at a non-profit, doing research, I spent a couple years trying to get a grasp on, honestly, what I felt I had been culpable in, like, “What have I helped cause to happen?” and “Is it good?” And then I also had thoughts about my — I had spent a lot of time when I was younger as an activist and involved in a lot of really morally and ethically-based movements. I didn’t feel like what I was doing contributed to the values that I had come up with, and the people that had opened doors for me and inspired me.
So I just looked around and, in particular, I looked at who was in the room when we built these things. And it didn’t look like who was in the room when I had worked in the music industry or when I had worked in media or I’d worked in activism. I thought, there is no way we’re thinking about these things.
I wrote this series of posts on my blog in maybe 2009. I think I had said I was leaving Six Apart, but I hadn’t left yet or something. You know there’s that senioritis moment.
MATT: In between, yeah.
MATT: Oh, senioritis. [laughs]
ANIL: Right? I wrote about basically every major tech company and what I thought was wrong. I did one a week all summer long that summer. I genuinely thought I had blown up my career. I got PR people that were like basically, “Who the hell do you think you are?” [laughs] And the end of that was — The Social Network had just come out.
MATT: The movie about Facebook?
ANIL: Yeah, yeah, exactly, the movie about Facebook. And they had a story tied to that in the New Yorker, and I gave a quote basically saying what I thought. I just said, “If you’re Mark Zuckerberg you don’t know what you don’t know, and of course you think everybody can have one public identity that’s yourself because if you’re him, you’re not going to get vilified for it, you’re not going to get kicked out of your house for it, you’re not going to be violently abused for it.”
All the back channels lit up. People were like, Facebook is not happy that you are quoted in this and that you’re seen as somebody credible. I really thought, “I’m never going to work in tech again.” That was my belief.
You fast forward anyway to 2017. By the time I had, you know, dabbled, even what Gina and I were doing, we very assiduously avoided the conventional startup world. I didn’t go to Sand Hill Road and look for VC funding and all those kinds of things.
But I had a really great conversation with my wife and I was saying “I still want to change this industry for the better, I still want to have impact.” I haven’t given up on the potential but tech is itself one of the major social pillars that influences culture. The other is media and entertainment, broadly, and obviously I deeply care about pop culture and those things. And then the third was policy and governance.
I had spent a lot of years when I was doing that nonprofit work, working with the Obama administration, and [I’d] gotten access in policy and government, and I started to see “Okay, there’s some things that government can do here but it’s by design, slow. And also very corruptible or capture-able by interests.” Okay, we know those flaws but we didn’t maybe know the extent of them, but we knew that was there.
Then media and entertainment, I had a little bit of an in, and people knew me. They didn’t know why. They’re like, “I don’t know why this guy writes about Prince all the time but like… [laughs] and he seems to know what he’s talking about, this is weird.”
MATT: I think you’re a worldwide Prince expert.
ANIL: Yeah, you know, I’m not the most knowledgeable Prince fan in the world but there aren’t ten ahead of me. [laughter] So people were like “OK.” And not just that, but they’re like “OK, he understands how the entertainment industry works and that’s interesting.”
I thought about how you triangulate fitting those pieces together into shaping tech to be something better. And the truth is, I had as much access as you could have, and I hadn’t been able to have any affect. I had been there, banging the drum, and to talk to the Chief Technology Officer of the United States, and be like “Here’s an issue about data privacy,” and they’re like, “We’re trying our best.”
To see people in his office that spoke up on net neutrality and had their lives destroyed by companies trying to lobby to keep net neutrality from happening. This all happens behind the scenes, but I would see good people who were targeted by the AT&Ts of the world, by the Comcasts of the world, who were like, “Do not let this happen.”
And then even in media. At the time, not long after that, I got a column in Wired. So for me as a writer that felt like the pinnacle.
MATT: That’s making it.
ANIL: Yeah, exactly. In tech influence, it’s like, this is this opinion column in a magazine that at least has the impact. I would write there and it would just get dismissed and bounced off. They’re like, “Oh that’s nice, you brought your blog into a magazine.” It didn’t land.
What I realized is that even seeing the most credible academics, like I think about dana boyd, who is just brilliant at this, and her research institute, Data & Society, would document all this stuff, and it’s incredibly valuable as the underpinnings but it doesn’t cause the change to happen. So ultimately the theory of change I arrived at was you only change the tech industry by being a founder and by having a hit. There is nothing else you can do. Anybody else will get dismissed.
MATT: That’s true, yeah.
ANIL: “You’re not in the game, it doesn’t count, you’re just an outsider, you’re just jealous.” I see VCs say this all day long on Twitter to every valid criticism. It’s like, “You’re just jealous because we’re rich,” basically. They’ll frame it different ways but…
As we’re speaking right now, there is a lot of criticism of WeWork ahead of their IPO, and there’s a lot of fundamentals there that are a little iffy. Now people are turning the corner into “Well, okay, maybe there’s some stuff here that’s not as rigorous as it should be,” but the first wave of defense was all “You’re just jealous, they’re a unicorn, you don’t recognize their vision and da da da.”
I’m like, yeah, there are times I miss a vision of something, but the reality is, this is real estate, this isn’t magic. And software doesn’t make real estate magic, and you have to be able to say something is what it is. Interestingly, there is a willingness to hear it if you are at a certain title — CEO, founder, and if your product is a hit.
MATT: And here we are.
MATT: That’s a good full circle. On Glitch I noticed the careers page had some interesting stuff. A series of pledges, including no endless meetings.
MATT: Allyship, best practices, reasonable work hours.
ANIL: Oh yeah.
MATT: Do you ever worry about being held to these things in the future or is everything on there something that you feel like is…?
ANIL: I hope to be. Yeah, the good news is a lot of these date back to Fog Creek, and even to the earliest days, which at this point is 19-20 years ago.
MATT: Which are the ones you think are most important for people?
ANIL: Oh that’s a good question. I think the fundament is respect for the people on our team. Joel Spolsky, as our founder and our chair, said from day one, “We treat our people well and we give them a great working environment.” And it is a very deeply held view of his. I agree but that’s the ultimate authority, that’s my boss. And that has been true from day one.
MATT: Orient it a little bit. How many people do you have here in the office, how many remote?
ANIL: The company overall is 56 people. We just had two people accept, so soon we will be 58. Of those, 30 are remote. So 26 here.
MATT: The majority, yeah.
ANIL: Yeah. And what’s interesting is that is shifting. This has never happened in the 12 years the company has been doing remote. People are starting to move to New York to be in HQ, to balance it out. And it’s always been the other way, which is the remotes were the one-offs and so people were like — obviously you don’t have to twist people’s arm to be in New York, but it was an interesting sense of people — the balance of both.
The majority are remote. There’s an interesting mix of people that have been in the commute-to-work-everyday thing and don’t want to do that anymore. I really love that here in New York, we all take mass transit, but nobody at the company has to drive to work and be stuck in a commute. Nobody around the world does that, so that’s really powerful.
Then importantly, you know, we make sure we pay attention to — there are people in leadership that are remote. There are people in every role and every team, in marketing, in engineering, in infrastructure, [for] everything we do, there are remote team members, in media, all the things that we do, there’s no one side or the other.
MATT: I’m excited to hear your thoughts and learn a little bit more about how you operate in a distributed fashion at Glitch.
ANIL: Joel Spolsky, who is our founding CEO and co-founder of the company, he had a lot of things to say about how you should set up your tools, your software and [how] everybody should have a really nice, big monitor, and all this stuff.
Then he had this one principle which he still is very adamant about, and in fact he mentioned it in the board meeting we had last week, which is that every coder should have a private office with a door that closes. He really feels it makes a huge difference [in] productivity, and that there is a lot of research and evidence that shows having the space is very valuable, which I’m a full believer in.
I had never — before I worked at Fog Creek and now Glitch — had seen a workplace where that was true. I saw that they were able to have these results for a long time, and then when they built Trello they really started to grow the company and they ran out of offices. So they were sort of like, one, we can’t just expanding forever here but, two also, there’s lots of talent, or people that didn’t want to move to New York City. Those two things led to the obvious, in retrospect, conclusion of the company should be open to remote contributors.
It’s interesting because the impulse there was in order to keep doing the structural things they were doing. So remote workers at the company are required to have a private office at home, a room with a door that closes, and that is not your kid’s nursery or where your greenhouse is that you’re growing your plants. It’s its own space.
In exchange, the company, we provide a sit/stand desk, if you want a motorized one, a really nice chair, we’ll pay for the internet connection, we make sure you get broadband, all that stuff is taken care of. But it’s very much this sense of — to create the same atmosphere as what we have in the office. Then there’s a whole set of processes around how everything works to make sure it’s uniform, whether you’re in HQ or not, for coders. And that was something that was probably the biggest culture shock when I joined.
MATT: Having an office with a door is good for developers, why not for everyone?
ANIL: That was the first question I had, as somebody who is not a developer, right? [laughs] Can I get one of those? I can write a couple lines of code, like, what do we gotta do? And actually what we came to really talk about — and this was really catalyzed in particular — because we had shared an office with Trello as one of the other spin-out companies, and they were growing leaps and bounds around and after their acquisition by Atlassian. We were growing our team, and it was really crowded, we had this shared office in downtown Manhattan. People were getting doubled up in offices, and there was a question, “Why are these people over here and these people were there?”
What I came to understand as we built out a new office here is there’s different work modes. I’m someone that definitely falls into this where I have one way of working when I’m writing, which I try to do a lot, one way of working when I’m coding, for sure, and then of course you end up in a meeting, you’ve got to be in a conference room and sit down with people. There’s all these different ways of working.
I found one of the things that I find for a lot of people really consistently is they’ll say I’m most productive on email when I’m on a plane and there’s no WiFi, or I do a lot of writing at a coffee shop, even though I’m not talking to anybody else that’s there, I like that energy.
MATT: Totally, I can totally relate to that.
ANIL: Yeah I have that experience with libraries. Yeah, right? So there were different work modes and that was what really was the epiphany — even though we’re very effective with giving coders private offices, at the time the company was just that. There were really only coders. As recently as in the past decade, Fog Creek was a company where there was nobody who had a title of marketing, there was nobody with the title of sales, there was nobody… [laughs]
ANIL: I mean there’s entire business functions that did not exist. It was a very, very extreme environment. And good in a lot of ways — obviously it had a lot of success — and then bad in a lot of ways. So you couldn’t do normal things like having a couple of people get together to talk about messaging or design because those functions didn’t exist and you didn’t have any place to do them if everybody is warrened off in their own little room.
Finding the balance of how do you have different ways of working was really powerful. In HQ, it translates into conference rooms. We have a space that I literally set up to be like a coffee shop, and we’re doing that with our new headquarters that we’re building — it’s going to feel like a coffee shop. You have the nice couch and you get some caffeine whenever you want.
ANIL: So that’s there. But the biggest thing was the very first day I joined we had a meeting, an all-hands. We still have them every week, and everybody, even in headquarters, even at desks, if they didn’t have a private office, put on their headset and joined the video conference from their desk independently.
I had had this in many other companies that I had worked for or worked with. It wasn’t the scenario where if I was remote, everybody in HQ was all around one table, in one room together, and then I was on the other end on some weird conference phone, like one of those Polycom phones, and they forgot I was there and they couldn’t see me.
Or the weird creepy thing [where] your video shows up when you talk so every time you talk you know you’re on some giant 10-foot screen with everybody staring at all the pores on your face. These were the only two modes of interacting that I had ever had. [laughs]
ANIL: Instead, everybody had the exact same set up, the same camera, the same kind of headset and could see each other, and it didn’t matter if you were in HQ or if you were thousands of miles away, you had the same experience at that meeting. And everything changed.
MATT: It sounds like y’all — it must be a delight to work there. I’m sure some listeners are like “Oh, I need to check out their careers page.”
ANIL: As I said, we have a weird background. The company overall is — in two months, we’re going to be 20 years old. But we really started reinventing the company about three years ago. Some of the things that we changed: one, the top level, one of the very first things I did was around recruiting and hiring.
Very early on, Joel Spolsky, our co-founder and CEO, had said, “A good tool, if you want to assess people’s skills as coders is to use a white board and talk to them about their thinking about code during the interview process.” This is something, as I’m sure you’re aware, that became almost a cargo cult thing. Right? People would do this, what to me felt like a hazing ritual, of demonstrate your memorization of some algorithm by writing it on a white board from scratch, which is not how anybody in the world ever codes, especially not when we’re the company that had co-created Stack Overflow! [laughs]
MATT: It’s all about Googling things, yes.
ANIL: Of all the places. Right, yeah, of all the places you should be able to just copy and paste from Stack Overflow, it should be here. So his impulse had been right, which is like, let’s see how people think, and don’t put them through all the pressure of having to set up a dev environment just for an interview.
But I think in the absence of the steady hand saying, “Here’s the principle behind it,” it became, as I said, almost a hazing ritual. I felt very personal about that because one, I care a lot about inclusion in the industry and anything that’s a barrier is really dangerous. And then for me personally, I had worked as a coder for years and yet I knew I couldn’t pass that test. So what did it say about the utility of that test? Clearly it was testing something else, you know, to be standing at that white board and be put on the spot that way.
So we changed that. But it was the beginning of a change in the culture overall, which was interrogating which assumptions were about us being different for a real reason and which things were rituals that we had held onto from the olden days. Things changed pretty quickly after that.
For example, we had always had a pretty public set of rules around how the company ran, but we formally turned that into a public handbook that people could look at for our HR policies. And that both helped us refine and sharpen what we shared but also, again, helped in recruiting, because people outside could be like “Wow, I know how that company runs and how processes are going to work for me and whether I’ll be taken care of.”
We did that around compensation. Joel and Michael Pryor, his co-founder, had set up an engineering compensation ladder, gosh, probably a dozen years ago. A long time ago, but it hadn’t been kept up to date, and it hadn’t been modernized, and it was only for engineering. We turned that into a full salary transparency document within the company where we had documented what compensation and pay would be for every role, including non-engineering roles, which we introduced and actually defined. [laughs]
That was, again, it’s far from perfect, there is a lot of work to do to improve it, but it built so much more trust where people, especially underrepresented people in the industry, could be like “At least I’m being paid fairly compared to the person next to me. There’s always improvements to make in compensation but I’m not being targeted because of who I am, to be exploited, to be paid less.”
So there was a lot of this stuff where we took the building blocks that were there and then modernized them or updated them. That took a year, a year-and-a-half of work, but it really, really paid off in terms of one, attracting a new wave of incredible talent that were every bit the peer of anybody who’d ever worked here — which is extraordinary — and then we grew this company and this product while hiring at a fast clip, while expanding to a new headquarters, while doing all this other stuff simultaneously.
It’s very hard to walk and chew gum at a startup at the same time, and I think we were only able to do it because we had built that level of trust around the process, and especially around treating people in that basic way that they expected in the workplace. And some of that is about being remote friendly, some of that is about being friendly to people who are underrepresented in the industry, some of that is just about building fair processes for how everything works day to day.
MATT: Why not go fully distributed? Why do you still keep the New York office and why and how is it still important to you?
ANIL: We talked about it. That was a consideration actually before I had joined. There are a couple reasons that I had decided to keep it.
The first is I am, at the personal level, a very unapologetic New York City partisan. I’ve been somebody who cares deeply about the tech community in New York for almost two decades now, and I felt like the company is a symbol to the city of our tech culture and community. So there was a sense of which of the old guard, classic New York City internet startups were still relevant and pertinent and here, unapologetically here.
The other part was that, at a real pragmatic level, we were able to attract talent in certain ways because we’ve got options. There are people that one, like to work around other people. There are people that like to have camaraderie and they like to have a connection to the people around them. Certainly we see that especially in some of the other roles we had brought in, where we have a team doing media and we have a team doing community and all these other things, and they like to be able to commiserate and share ideas. So that was really powerful, and we unlocked that.
Also, there were people that want to live in New York City. [laughs] So we were just able to attract talent of people that were like, “I get to work at a cool tech company and be in New York? Sold.” The wild thing about that is that it’s cheaper than living in San Francisco!
MATT: In-person is so powerful so how do you keep those ties, being so much stronger and that communication being better?
ANIL: It’s something I’m very mindful of, and we are really disciplined in a couple ways. One of the things is no hallway conversations about work. And actually, one of the telling things is that we as a company have always done lunch together every day in HQ. So everyone in New York, really, meetings stop, calls stop, whatever is going on, everybody sits down, actually company-wide because it tends to happen in other locations as well, but really in HQ we provide lunch every day and everybody sits down and has lunch together.
Understandably remote folks worry about — that first day they’re like, “Wait, everybody there is going to be talking about stuff and I’m not going to know about it.” And the great thing is, saying “We’re not going to talk about work, we’re not going to make decisions about work in that lunch,” makes it a social place. It’s actually much more low key.
MATT: Do they really not talk about work? That’s so hard to believe.
ANIL: Yes, well because — it’s not that hard because there is so much stuff that’s interesting, well, one being in New York and two, at a social level. People like to put it away. We have really good boundaries. And also, they feel a real out-loud sense of “I’m not going to have this conversation without her here, without him here, without that person who is in charge of that.” Because we have leads, we have people who are leaders in the company that are remote, so it’s not — you couldn’t have a full conversation. Half the company is not there. So you couldn’t have the conversation without them.
So that’s really powerful, and then we already have a habit of taking notes at every meeting. If three people are having a meeting and they’re all here in New York, we’ll sit in a conference room. There’s not some weird ritual thing where we will refuse to sit together. But even there, there will still be the notes that are shared, just like they are for any other meeting.
So I think those habits, because the habits are deeply ingrained, and at this point for 10 years, there just isn’t that impulse. And it has been very clarifying to us about, one, just real life and what affects people, but also it helps with not being insular to just — to think about what’s happening out in the world to people. I think it can be really powerful.
That, for example, led directly to a policy we had created around paid climate leave. And originally the catalyst came from a worker of ours who is in Florida and was being displaced by a hurricane, and then we simultaneously had somebody that was being potentially displaced by a wildfire at the same time in California. So we made a policy to accommodate them. So it’s paid leave if you’re displaced by extreme weather.
But that core combination was one also informed by — we had had our headquarters displaced by Hurricane Sandy here in New York, when that hit. But that sense of just, we have people in different places, and we build policies that work for all of them. And that rigor, that perspective, I think it makes us run better, it makes us much more disciplined, it makes us have to be intentional about communication.
Those things are superpowers. They make our work better, faster, more efficient, more reliable, more trustworthy, because we have that discipline. And we honestly, I think we would be lazy if we were not distributed, if we were not rigorous around how we communicate, and in ways that would hurt our long-term opportunity.
MATT: Well and I think it’s one of the reasons that listeners will find the example of Glitch so fascinating, because you do put a lot of thought into all of your policies. One of the ones that I’m a little curious about, especially given you have a hybrid structure, is how you approach onboarding. Do people fly to New York and do some of that in person? How does it work?
ANIL: We do onboarding here in New York. The most extreme was actually back in the spring when we were really scaling up, we had nine engineers join in a week. Interestingly, we have a number of leads, whether they are director level or higher in the company, who are educators.
Onboarding, in its fullness of the plan, onboarding is something that we spend months on, so it’s three or four months. But the core of it is about those first two weeks, that’s the experience that we really design most deeply. And the first week is school. People come in, and the first day is like, “Get your laptop, get plugged in, get connected to Slack.”
We time it so that people join during the monthly all-hands. That’s when, in addition to an overview of how the business is doing, we share financials with the whole team, but also we do our bravos. That’s acknowledgement of people, thanking each other for the work they’ve done, when they’ve gone above and beyond on certain work.
They get to see — a new employee on their first day — how we share gratitude to one another within the company, and they get to see a meeting where everybody is dialed in, in the same way, whether they’re remote or not. And they get to see us model the behaviors of what we are in the first day, and then they spend some time getting two-factor all set up and all that other annoying stuff. [laughs]
Typically the rest of that core onboarding process — and this varies by team — we do a big overview of how the business runs, what the mission is, what the history of the company is, what our goals are, what the road map is. I spend a lot of time talking about the values and the vision of the company, and that’s typically in one-on-one sessions with people who have joined so they can have room to ask questions, or sometimes with their lead. That’s the better part of a day, usually right towards the beginning of the week when they have joined.
We have a much more social get-to-know in small groups, people across different teams. Typically we’ll have a dinner with some folks so they can say hi to other people in a little more social way. We will always bring in a person to HQ for onboarding and bring in their lead if their lead is remote. So people do get face time. We do think that has value. There is a way to deploy that as a tool, and especially for social interaction, when people are getting to know each other and have to build a relationship, so we’ll do that for a full week with the person and their lead in for a week.
The two biggest anxieties when people join a company are basically, “Am I going to know where I fit into any of this?” And then variations on what I think of as the problem we all had in junior high of “Which lunch table am I going to sit at?” [laughs]
MATT: Oh totally.
ANIL: We are very intentional about that, because we all eat lunch together every day, we’re like, “This is where you’re going to be, and everybody is going to sit with you, and you don’t get picked for a lunch table, everybody sits together.” And that stuff is — nobody wants to talk about it, but even if we’re adults, even if we’re decades removed from those points, that’s still that lizard brain in the back of your mind — that is always there. And so we do a lot to accommodate that.
Especially because the majority of our team, including management, are women or non-binary folks, and that is very rare in tech. And that’s true whether you look at race or LGBTQ representation, we over-index in a lot of ways there compared to the rest of the tech industry. So I try to be much more mindful of — there is a higher percentage of our team who will not have been in that scenario where somebody accommodated them and somebody asked “What do you want here?” or “What do you need?”
Just that little bit of accommodation or thought I think has been extraordinary for people where they feel overjoyed at “Oh, I wasn’t crazy in my last job when I thought it was wild that they didn’t even tell me who I report to or where I sit.” [laughs] “And that this place will tell me ‘You’re going to sit here and this is the work you’re going to do and this is why it matters.’” That’s all it takes for people to feel like they were welcomed.
MATT: We have talked a bit about benefits and I know equity is really important to you. So if I’m not in New York, how do you think about — I don’t get free lunch every day. Or if I am in New York, do I get a free internet connection?
ANIL: [laughs] No. So that’s where we go to. It’s not going to be exactly the same. We want things to be fair but they don’t have to be identical. There are things that are going to be better or worse, stronger or weaker in HQ versus remote. We try to, again, to document them and to be clear, and I think those are really good examples.
And it’s hard, right? We do cover internet access for people who work from home and we don’t cover it here. And we’re iterating on this stuff constantly. But one of the big things we do is we just talk about it, we are able to raise it, we have people — leads are encouraged to ask their team about it. People want expectations to be clear and they want to know that things are fair, but they don’t demand that they be exactly identical.
MATT: Anil, thank you so, so much. This has been Anil Dash from Glitch on the Distributed Podcast with Matt Mullenweg. We’ll see y’all next time.
MATT: That was Anil Dash. You can find him on Twitter at @anildash. That’s A-N-I-L Dash.
It’s good to know that people like Anil Dash are out there running companies and redefining what the tech industry can look like from the inside.
On the next episode of the Distributed podcast, I’ll be joined by my colleague Mark Armstrong once again. We’ll talk about the things we’ve learned over the last year, discuss some of our favorite takeaways from the podcast, and take a glimpse at what awaits us in 2020. I’ve learned a lot about how other companies are doing distributed work, and I’m excited to collect my thoughts and maybe make a few predictions about where we’re headed.
Thank you for listening, and see you next time.