The Accidental Rise of Live Coding on Twitch

Matthew: Hello and welcome to Developer Marketing Stories. My name is Matthew Revell.

Adam: I'm Adam DuVander. And today we are joined by Katie Penn, who recently left Twitter as the global head of marketing and previously held roles in developer platform at Twitter and Twitch.

Matthew: And before we get into the conversation with Katie, one of the reasons we're doing this podcast is to share with you about our training and coaching programme for developer marketing professionals. It's called Developer Marketing in Practise, and you get a mixture of coaching, learn at your own pace training, and also live sessions with experts in different areas of developer marketing. If you want to find out more, go to developer marketing.

Adam: Great. And now as we get into this conversation with Katie, I think listen for if you are a marketplace or if you are looking to engage developers before you have a product ready, there are stories within her time at both Twitter and Twitch where they really did that in some great ways. So listen for that and let's get to that conversation with Katie. Alright, welcome Katie. Can you take us to that moment where at Twitch you realised that a developer audience might be one to focus on?

Katie: Well, I was brought in to focus on a developer audience, so it was my remit when I was brought in, I had been at Twitter running developer marketing at the time, and they wanted to build out a platform ecosystem. I think that it evolved while I was there. So originally it was gaming integrations, so we were working with gaming companies on integrating their games into Twitch. So there were unique experiences within the stream wall. The streamers were playing the games, then we started investing in the platform in a way that would require us to bring in a wider audience that developers. And as we were making that product investment, we realised that there was a big opportunity and challenge because developers probably didn't think of us as a platform for them to build upon.

Matthew: So was it an organic thing of there were already developers who were live streaming, coding and similar things?

Katie: Not that we knew of, no. But as I built my marketing team, so I was the first hire and then they wanted me to build out the team and I brought over someone from Twitter. I usually in the developer marketing space and just for transparency, I haven't worked in that for a while. But in the developer marketing space, usually I'll start with, I brought in an advocate to be working in developer marketing. So I brought someone in that had marketing shops that ended up in my team, but had started in the developer advocacy team at Twitter because a lot of developer marketing is content. And so he was my first hire. And yeah, my remit was to grow that space as we're investing in a platform play, which we eventually launched, which was essentially like an ISV app store where developers could build, we called them extensions, but they're essentially apps for streamers to use in their streams.

And there was a monetization share play around that app store. So we had to actually start bringing the developer audience in and making them aware of the API so that when we launched the product, we had some builders that were already familiar enough with the platform that they'd be interested in participating. And in that process, my first marketer just went through the process of setting up, which is a heavy lift setting up to be able to stream to understand the product, not to stream on it. But then as we started to try to pull in a community for the launch, we didn't have a lot of developers that thought of Twitch, and so he and I thought it would be a good idea, maybe he just streamed a little bit to start coding just one to see what it was like, but two, to start using it as a platform When we launched, and we are very surprised where we wanted everyone to come in as an API developer, we were surprised how many actual advocates started coming in for their own platforms to start streaming on there. But it didn't hurt us. It was just a surprise community. We weren't actually intentionally going after, which is the developer relations community.

Adam: But it wasn't intentional way to be able to understand the Twitch audience and recognise that you had to find some way to get developers excited about the Twitch platform.

Katie: Absolutely, and the thing that happens is when people start using your product as a consumer, because essentially that's what they were, they become a lot more sticky in terms of wanting to use it and integrate with it. So the good news was we had developer advocates now using it to teach well who is their audience? Their audience was developers. So it was very easy for us to either plug in so the person on my team would co-host with them and then be able to do a plug for what we are doing in the developer space. And now we had customers or consumers, not really customers, but consumers listening in on developer advocate channels that also now were like, Hey, this is a cool product and what are they their builders? And so then they were probably going to look and be like, oh, there's an API, oh, I can fool around with it. So it was more of an organic thing that happened, but it definitely served our purpose of one, exposing them to the platform and two, exciting them about the possibility that it was a new platform for them to integrate and build upon. And three, having a monetization play in the beginning where their investment may have returns, which a lot of times it's just build with us, but there's really nothing that we're actually offering developers.

This was a pretty cool model where they were going to be able to have a revenue share with streamers.

Matthew: Looking back on it, it was almost a happy accident. Would you say that you learned, let me put it a different way. Are you used to having happy accidents in your career or was this something new that you had to take stock and think about how to change course to take advantage of it?

Katie: I'm used to having accidents in my career. I wouldn't dismiss the fact that we had people on the team that had relationships with the developer community. We had all worked in the developer community, so we understood personas and behaviours and we'd had big launches before. That's why we are hired. So I had been at Twitter before we had launched Fabric, I had done acquisitions for the developer community. I understood I had started at Salesforce when the developer relations team was, we were a team of four or five. We're so small. And so I'd seen it kind of grow up and understood about, it's funny in the creator space now everyone talks about authenticity, but that really started in the developer space.

There was no bake marketing to developers that was going to land. And I had mentioned this before, part of the reason that we knew that once we started live streaming and we saw developers engage that we knew that we had something there is because when we had launched this product called Fabric, which was a mobile SDK at Twitter, we had a big launch event, which is always hard and I may not recommend it in the future because at which it ended up quite problematic.

There were a lot of lessons learned with launching the fabric at one event, sorry, launching the product to one event. But at Twitter it worked out well. But it was a difficult event to get developers to sign up for because we didn't have a mobile app developer play at the time. We had an API that was quite restrictive.

But what we ended up leaning into based on our experience was at least programming content at that event. They knew there was going to be a reveal, but programming content at that event that would resonate with developers. So we hadn't launched our products, so we couldn't actually reveal the content around building with the mobile SDK, but we did lean into the open source investment. We had made it Twitter and had those individuals that had built that come on stage and were able to publish that part of the content. We're able to lean into how Twitter managed the size and growth and the data sets. And what I noticed at the event is those sessions were packed even though they had nothing to do with our product and they were packed because it's a community and they want to learn how people build their own products.

And so I always weave that into any strategy we had. So whether it's just engineers like writing blog posts about the challenges they have, it may not be the same product, but it's the same language and the same experiences. And so revealing under the hood helps a lot when you're trying to win the hearts and minds of developers. They want to hear from the developers inside the company that are building that product as well. So it's kind of a strategy that helps where we're like, of course, maybe we should be showing them how we do a Twitch setup and how I stream on Twitch so that people can, I just find it more appealing.

Adam: And so the common thread I'm seeing there is the product isn't quite ready, but you still need to engage the audience.

Katie: Yes, yes. I mean you want, from a marketing strategy, you kind of want to not start from zero when you do a launch, right? You don't don't want to start at a thousand either, but you kind of want to build that community a bit. At Twitter, we knew we were going to launch that product. We knew that we had had a perception problem because of the API. But we're going after a totally different audience. We had bought Crash Lytics, we did mini meetups before we ever launched that project. I did one in Japan, I did one in London.

I mean we were scrappy, but we are doing them so that we could build a bit of an audience for when you launch because who are you going to share it with when you launch? You need people that are somewhat already showing intention and engagement. It's a little bit easier probably just based on the way people communicate, but we kind of invested in the developers before we revealed what we were doing so that we had somewhat of a base.

Adam: I want to zero in on one little thing that you said there. You mentioned that there was this a I perception at Twitter, but that was okay because it was a different audience. I think a lot of marketers listening to us right now would say, wait, that's a developer audience. How is that a different audience? Developer is a developer.

Katie: But at the same time, it was a different audience because we're offering, you weren't building off of Twitter data, what we're offering as a product. It was really an SDK, the foundation of, they ended up selling it to Google and that audience was able to be portable with Google because what they weren't building were apps on top of the Twitter platform. It was kind of an independent product now perception and trust to use a product that they wouldn't end up. I don't know if I can curse on this podcast, but am I allowed to screwed by Twitter? Yeah, we still had that problem and in the end, I think the second flight conference I ran, we had Jack at that point and he apologised developers and then they still sold the product. So I don't know if that reputation was, I think they maybe forgave us once and was all about fabric.

But once that sold, I dunno if you could win the trust back twice. I think that what I would say about the fact that it didn't matter, it didn't matter, but it was a different investment from the company. They weren't investing in like, Hey, all developers come and build on us. We're going to let you build your own types of Twitter apps on top of our API. It was very much like, Hey, we believe in the mobile app ecosystem. We believe we have some tools that would benefit you. We had mub, which was sold as well, but like, Hey, we're building this SCK because we believe in the mobile app ecosystem and we also see mutual benefits. For Twitter, it was different than, Hey, you can build your business with the Twitter data. It was a very different,

Adam: I agree with you that it's a different type of developer audience and I think that's one of the keys that we see, especially marketers who are coming to a technical audience for the first time missing is that there's just this one audience. But really there's so many ways to, you have to

Katie: Remember, I came from Salesforce developer audience completely different than going into mobile app. It was just so different in the API audience as well with something where when you're building on just one data set, you're very dependent on the health of that company where it's a mobile app audience. So really about the health of their own mobile application that was independent of a platform. So it was more like a product. When I went back to Moab, it was the same. It was, I ran marketing for Moab and we brought in publishers, a lot of game developers, and that was independent too. That was, they used that product for monetization of their own apps. So you just have to know the intention of your audience and how they would be needing your product and interacting with your product.

But there's different, as when I work in, I've worked in other marketing spaces, the different customers and audiences you bring in.

Matthew: What advice would you give to a developer marketer today who's looking to segment a developer audience so that they can target those developers who are most likely to be of strategic value to them?

Katie: I mean, there's some basics and one-on-ones of marketing you can apply to the developer audience as well. So if you were going to build out, and I know this is, people use Swir all the time, but the personas of who they are, just understanding there's a difference between an enterprise developer, a mobile app developer, a game developer. So those are three that I've worked in. Even within that you can segment out, say we are doing mobile app developers, hobbyists, where do you find your hobbyists who are, they may be a mobile app developer that is trying to build a business but has a full-time job. So they may be mid-tier, and then you have the full-time mobile app developer that is working for a company and is spending all of their time. So you can segment out even in the personas who they are, where they spend their time, what their motivation is, and then which one you want to invest in.

Some of the strategies we've always had is draw a line in the sand and who you're really going after within those segments and then find other activities that can serve the other one. So say you're going after the mid-tier mobile app developer, maybe you're going after the full-time paid developer. Okay, well maybe you have to do events and activations that find them and respect their time. Maybe that time is one of their priorities, like immediate very clear documentation or whatever it is that you can identify as their priority and then know that you can reach beyond that through other activations. So maybe we probably weren't going to get the game audience interested developers on Twitch by only hosting the live stream. So this might be the hobbyists, some people listening into other advocates, but that then we had to be at the certain events and industry moments where we would be able to reach those that are quite busier but are able to focus their time or give their time in certain moments in time. So it's kind of like you build your strategy based on the behaviour, the audience that you're after.

Matthew: Different types of developer have a different role in your strategy. So what would that look like if you've identified some developers, you want to target some different types, knowing that someone's a hobbyist and some other things about them, what does that do to influence your strategy?

Katie: So knowing that someone's a hobbyist, if we were going after the wider, larger community, and the way that that would influence our strategy for hobbyists is probably one, meet them where they're spending their time. So invest in forums and pushing out a lot of content that can have a wider reach, ensuring that who they follow our partner. So having a partner or strategy so that whoever they're following, you're doing more partnerships around them so that your content can also be supported through the partners that might have a reach with those hobbyists. Yeah, probably working with analysts if you're on the enterprise space so that you're actually, the hobbyists are going to be able to surface that information and your play through publications and things such as that. So it's really just depends on who you're going after. Whereas I Slack, it was a lot about partnerships and integration, so they had to go through more of a business slant and going in with their BD teams and getting those really top integrations first with the trust that the hobbyists would follow based on what they were seeing. So a lot of times it's like, okay, showing those bigger wins and knowing the hobbyists are going to pay attention and then come on board.

Matthew: So knowing who they are tells you what they want to get out of the relationship or what they want to do generally so they can insert yourself into their goals. I guess

Katie: The OG days of DevRel was just like, come build, it'll be so much fun. That's kind of how it started and I think it worked out because there weren't a lot of players in the space, and so when you think of how many dev meetups there were 15 years ago, it just wasn't the same. And so it was easier to just say, come build developers, developers, let's build together. And now the mind share and the opportunity, it's much more difficult, one to penetrate that and two, to realise that they're not just going. I think that that audience isn't just going to build out of goodwill and intention anymore. So the more you can have a reason for them to a benefit for them, the more likely you're going to have their attention and interest in trying. And that's where I thought that Twitch was very smart at having weather. The intention for developer is monetization to be able to make, and there were some developers that actually made some good money off of the extension launch that we had.

If that is their intention, then now you're having an offering there. There were some that were also just big fans of the streamers, so if it's maybe a community attention intention, then offer that play, but make sure that there is a reason for them as well as for you. If it's just to benefit your own platform, I think you're going to fail at this point.

Adam: What were some of the fun moments six months into Twitch when you had been engaging the developer audience and you realised that there was this community aspect there?

Katie: Well, I think that especially since the developer audience, a learning audience, it was just really fun to have advocates that maybe had nothing to do with the space we are in reaching out and wanting to adopt a platform. It was just kind of a convergence of, my gosh, my old Salesforce people are writing us asking how to use the platform for their developer advocacy. So it kind of legitimised just the work in DevRel and that was really fun being invited to the DevRel conference, just speak sometimes, especially as marketers, we're a little less respected in the DevRel space than maybe in some other spaces. So it was kind of just this beautiful emergence of a community and I think that that is always quite special when you can help build that on a platform. And it also just doubled down on what made Twitch unique.

So it was just, I ended up going into a new role at Twitch because I wanted to spam my remit outside of I had done it for so long, and so I went into expanding audiences on Twitch and went after music and sports. But we also had this already had an example, a proof point with the developer space, and it kind of helped support the business case for going after larger audiences. The developer space was going to be so big, but we felt like, hey, if we can bring in an audience outside of gaming, and I think the developers came first, we think we can make a big bet in a few audiences that we also think would stick. So the idea that developers were able to stick was because it had the recipe for their own success. They could do live coding, they could interact with the community, they could make clips and then share them on other platforms. They were there to teach. And so we ended up going after music and then we had this convergence where the pandemic hit and it was like, Hey, let's launch the music space, but let's do a fundraiser. So we did stream aid and we were able to build an experience, an audience for a moment in time, but could also then once again be a proof point for musicians to come onto the platform.

I think that today there's 15,000 DJs at livestream on Twitch, so it's a new audience we could go after again.

Matthew: So you mentioned there was extensions or apps that people could build for Twitch streamers use. Then obviously there's the monetization side of things for people doing the actual live code streams themselves, but what were some of the metrics that you used to track whether this new audience was healthy and growing for Twitch?

Katie: Think there's a few metrics. I mean, there were two different audiences at that point. There was the Dev Rel community on Twitch, and that was just, that wasn't part of our business goals, so was something that we're able to watch and follow, but we actually weren't tracking. It may have ended up in the DevRel teams goals, but I don't think it definitely didn't start there. And that was just like you could just track the health of how many live streams about developer topics were happening on the platform, how many new users were coming on that identified themselves as developers. The other magic was that we already had developers that were using the platform to watch gamers because you have to think of the audience demographic there, like Twitch resonated and the majority of their audience were men between the ages of 18 and 26. And so often a lot of, I don't know what the makeup is today of developers, but we knew we already had a community of developers.

They were just watching different content, so it was easier for them to move over to start watching developer content on a live stream as well. But in terms of the metrics around the launch of the extension, the ISV platform, it was really about how many extensions were being built. And then of course, API calls as well, starting from the basics, people coming to the site, people using people reading the docs, and then people integrating with the API and then how many actual extensions were being shipped. And then the final metric was really around how many of those extensions were being used by the streamers. So you had two audiences. We also needed adoption from the creators on the platform that would integrate some of those extensions. And we had a couple extensions that were quite viral and popular, and then some that didn't work. So there was a trust and investment from the developers that wanted to play with it and build on it.

But there were a few of that made, I mean, gosh, I wish I should have done the research before this, but there were a couple that made them overnight heroes and we had them on stage. We ended up doing developer conferences, which Twitch had never done in conjunction with Twitch Cons. The CEO led them, he was there. She, I think that that was a moment where, wow, we really built something here that was unique and different to an audience that was solely focused on gamers. Now we're having our own developer Derell conference.

We did some international trips around it, bringing in more audiences. It was really fun. Gosh, you guys, you're making me Miss DevRel. I'll go back into it

Matthew: Mean. Is there anything you look back on now and wish you'd done it differently?

Katie: I would say that the Twitter experience was probably the hardest because I don't know if I would've had such a conviction on the fact we write the scripts and everything around these big, big events, and there was just such a conviction that we were back and that we were sorry, and that we were investing in developers again. And the one thing that I've learned within the developer community and outside, I just gave a talk for creators, and you see the platforms really changing their, and developers are creators. I really do believe that as well. You see platforms changing their algorithms and their strategies all the time. Is that a relationship that a developer will have with a platform and a company has to benefit them? It can't be all about trust because companies will always move in the direction that supports their business. So we kind of maybe went a little too far and we're all in on the community again, believe us, we believed it at the time, but a lot of us left and saw that the product was being sold, and there's sometimes there was a naivety around like, okay, that's the old days and now we're going to be different.

And then maybe it doesn't launch the way it's supposed to, or maybe it doesn't grow the way it's supposed to. Or if Platform and DevRel is kind of a side afterthought, yeah, we know that we should do it, but there's no proof that it's going to be helping or measurable goals around how it's going to help the business. It can often be dropped. And so I always advise marketers that are going into DevRel, which is a really unique and challenging role at times because you might find yourself not on a marketing team, but on a product team, and you don't want to end up being considered service versus a strategist, which can happen quite often if you're not technical enough. I always advise them when they're thinking about going into Derale jobs is make sure it's core to the business. It's very easy to become an afterthought if you are not being considered a core piece of the business. So usually in marketing, I'm like, it should be the marketing for the company. If it's just a side piece, make sure it's a really big side.

And that was a perspective I just had to learn when I saw it go up, it was all about it and then all at the ground again.

Adam: So it sounds like maybe the message for marketers of developer products is to really zero in and make sure that there's something there for the developers. It's not just at the whim of the company that I just looked up, I'd forgotten what he called it, but Jeff Lawson from Twilio, and this is now eight years ago, wrote your strategies', last message to developers, and it was kind of about that idea that it can't just be this onetime thing. There's got to be something there for both parties, right?

Katie: Absolutely. Absolutely. And it has to serve both parties in a way that benefits them both. So we talked about just getting the hobby to come play, but not actually having a benefit for them except for learning cool new things. They're just not going to be that sticky. But yeah, I mean that's why those companies make sense. Like Twilio or Go work in marketing for GitHub, those types of companies that the longevity of your own work and career serve the purpose of the company and the developers that you bring on because they're just a core customer versus a tangential customer. And at the same time, fabric, I mean, fabric was an amazing product.

It did sell, but in the end, it I believe got broken apart and it was still the real product that they were interested in was so we built Kingdoms and seen them get taken apart few times, but that's not just independent of developers and dele. It can happen with lots of companies.

Matthew: What are the frameworks or the tools, not necessarily software tools, but kind of thought tools that you've been using throughout your career that you would recommend to someone relatively new to marketing who's about to go into either a Dev Rel team or a developer marketing team to be a developer marketer,

Katie: Specifically to a developer marketing career, invest in the product marketing aspect of it. So I think when you go into a new role, especially in that space where you're the marketer and sometimes don't have an equal seat at the table, understanding your product mechanics and go to market and product positioning helps quite a bit. So I wouldn't go in there with just the generic playbook of marketing. I would go in there from the perspective of product marketing and product market fit, because that's something unique you can bring to the table. It establishes you as part of the PM team, and bringing your product to market in a viable way supports your entire business. So I always invest quite heavily in product marketing and understand those are the topics we talked about, understanding your personas, behaviour, pricing, positioning, all those things fall under the umbrella of product marketing.

I think a lot of teams will bring in marketing for dev events, and I love events, don't get me wrong, it's kind of my love language in marketing, but if you don't understand the product positioning features and the way that you'll bring it to market first, you'll be considered just a service arm of the business versus a strategic partner. So I want all marketers to be respected in teams and to bring value to the teams. And I would say lean into the product marketing function, especially in the developer ecosystem space. Often something that I fight for quite a bit for my team is to have a seat at the table. But if you're in marketing and you come into, not a marketing org, but a dev re org, it's a little bit harder. So a marketing org will have different functions. Everyone's a marketer and the leaders of each of those functions have a seat at the table, but their executive is quite often A CMO or CEO and the derell org engineers, advocates, marketing, PMs, data scientists.

And quite often because your peers don't understand your work or maybe don't consider it equal to theirs, you often will not have that seat in a voice at the table when they're making decisions on the product. And so understanding the product, you don't have to be the developer in the room, but understanding both the product, but also our role as marketers is really understanding customer needs and what's going to resonate with customers. And really arming yourself with that data will give you more respect within your organisation to make recommendations. So if you understand your customer needs and your audience, you can make recommendations on launches, you make recommendations on your digital strategy, all of those things versus being told what to do or being an afterthought.

Matthew: Where can people find you online?

Katie: I'm still on Twitter X. You can always find me there, but there's so many. I was on Blue Sky and now I can't log in right now, so I have to figure that out. I'm on LinkedIn and I'll probably start being more active on Pres because a lot of my communities have left. I'm still on Twitter, but a lot of my older communities have left.

Adam: And then you're taking a professional break right now.

Katie: I thought I was, but I left the company in September, and yes, it's been very nice, but I've been doing some consulting and speaking engagements.

Adam: Actually, you probably don't know whether you're looking to go full-time again or even if that would be a developer audience, but maybe you could in closing here, give some advice to someone who is looking to have their first role in developer marketing.

Katie: Yeah, I mean, I think it's an amazing space to go in, and I still am a huge advocate for people going into developer marketing. And the reason is, one, it's an incredible community. So it's a really fun, and then you're surrounded by builders, and I think that it's still an audience that is an equilibrium. I just love that anyone could change their life or career becoming a developer. That's just one of my passions is that I remember being in cars travelling. I mean, you should become a developer. I would say that because I really believe that it could change people's lives and build careers for anyone. You didn't have to come from an Ivy League school to become a developer.

So I think it's an amazing space for someone to go into it. I would recommend that it's an open door that could start going to meetups and meeting people, start understanding this space more.

You don't have to do traditional prep to go into developer marketing. I think you would have to have an appetite for community and understanding community marketing quite a bit, because I think that there's just an overlap there. And then I would just start joining. The opportunities are in front of you. Go to the free days at the Salesforce Trailblazer. If you wanted to get into the enterprise space, host your own meetup, connect with some advocates, and start talking to them on how you could get into this space, start contributing. There's just so many ways versus just applying for jobs on LinkedIn. The more you can dive in and build around the community that's already exists and ideally in a product that you're passionate about.

So it was very easy for me to go into developer marketing at Twitter because I was a super user of the product and I love the product. And so if there's also a product you love and they have, or even a product space you love, maybe you won't get that product, but a space you love and there's developer marketing roles in that space. It's just like you're always a more appealing candidate if you love the products. It's just easier for them to see, hey, this person may not know everything about what we're doing, but they're going to give it their all, and they're going to be super passionate about it because they're a super user themselves.

Matthew: Awesome. Well, thank you very much. Yeah, thanks Katie.

Katie: Yeah, so nice to meet you both.