Library Blog

Post featured image

Growing a Developer Relations Program Heavybit

In this Heavybit presentation, a panel of developer evangelists and advocates describe the tools and community organizing methods that make their company’s developer ecosystem thrive.

CMX Media Founder and CEO David Spinks moderates the discussion around how building strong engagement with developers and community impacts product design, customer loyalty, and brand recognition. He is joined by:


David Spinks: My name is David Spinks, I’m the CEO of CMX. CMX is an organization that provides membership and training for community professionals of all kinds. We have a lot of developer evangelists, community managers, VPs of community. Anybody who’s leading community for any organization. Our goal is to bring together all those people, share their knowledge, create education and training, we host big conferences. Things like that.

I’m really excited to be here on this panel, to really dig deeper into developer evangelism. Which isn’t something we always get to go real deep on. And share some of the learnings of these amazing human beings next to me. Why don’t we start with some quick introductions? First we have Tristan Sokol. Tristan is a developer advocate at Square where he works to drive product improvements via community feedback, and the design of powerful developer toolkits and SDKs.

In the past he’s built community for Intuit’s QuickBooks online APIs and third party developer applications. Can we give a big hand for Tristan? That was like a medium hand. We’re going to need more. We’ll do it on the next one.

Atlee Clark is the director of app and partner platform at Shopify, where she connects developers in the Shopify platform and manages the go-to market for the company’s third party developer ecosystem. In the past, she’s also been the executive director of The C100, a nonprofit community dedicated to supporting Canadian technology, entrepreneurship, and investment. Give it up for Atlee.

Atlee Clark: This is a good way to start.

David: That was like a 6 out of 10. No offense, that wasn’t because they liked her more than you, it’s just because I’m egging them on.

Tristan Sokol: They should.

David: Romain’s clearly going to be the most popular one, with this 10 out of 10 applause he’s about to get. Romain is a head of global developer relations at Stripe, where he works with thousands of developers and entrepreneurs to build thriving internet businesses. Before Stripe he was the first member of Twitter’s international developer relations team and co-founder and CTO of Jolicloud, a pioneer in cloud computing. Give it up for Romain.

All right, Romain. Clearly you’re their favorite. I don’t know what you did there. I don’t know if you guys can sense the tension on this panel right now amongst this group of competitors. Can you feel it? It’s really tense. All of them are in a similar space, working with businesses, many different kinds of businesses, a lot of small businesses. Working with developers who are building on their platforms.

It’s going to be interesting to dig deep into this specific area of developer evangelism. A good way to start to ease the tension on the panel a little bit is I would like for each of you to say one nice thing about someone else’s company on the panel.

Atlee: Ooh.

David: We did not plan this, so they’re shocked.

Tristan: I can go. Stripe has excellent documentation. You can ask anybody who’s used that, great documentation. Beautiful looking and great to use. That’s a nice thing about them.

David: Very nice.

Atlee: I would agree with that. Also your designers, your product design and product marketing design is real nice. You guys just did your earnings call, and Square capital? Damn. It’s good. I love it.

Romain Huet: I guess I’m going to have to talk about both now to make it equal. Shopify is obviously a great company, a great partner of ours at Stripe. They empower a lot of businesses that would otherwise not be served directly by Stripe. And I really love the Square customer experience. Amazing to walk into any shop or restaurant and have this amazing feeling when you go sign your receipt.

David: All right.

Tristan: We all got a lot.

Atlee: So much love.

Developer Relations in the Community

David: I feel a lot better now, I don’t know about you guys. That was more for me than you. Now that we’ve got the important part of the panel out of the way, let’s talk about some developer relations. We’d love to hear from each of you first. Could you just give us a high level idea? What is the business objective of your community program, who are your members, and how are they participating in this program?

Tristan: Sure. With Square I would say that we have somewhat more of a traditional community structure. We have a spectrum from longtail newsletters and content into a more integrated Slack community that we try to do support through, well getting more members into that community for cross-selling and promotion, as well as a few initiatives that are in person. Where we go to events and meet ups in different locations.

The primary objective we’re trying to get through here is all around top of funnel growth, and then the bottom of funnel product feedback.

We try to get people to get on top of this great platform, primarily learn that we aren’t just a credit card swiping device, that we also have a wide variety of developer focused products as well. Now you all know that, my job here is done. We’re trying to get people engaged from the beginning and then meet their needs at the end with customer support and product feedback.

David: Awesome. Thank you. Atlee?

Atlee: For us, we look at developer relations holistically. Between web developers who build apps for web apps for our platform that are sold to merchants through our app store, but as well as front end designers and developers. That whole ecosystem is really to bring and help our customer merchants start and grow their businesses. We look to that entire community to get us new merchants, but then also help them along their journey to increase sales, stay with Shopify longer, and then the flywheel keeps going. Because they pay the partner, and then they stay with Shopify, and we keep going and going and going.

David: Is that a combination of online and offline? What does that look like in practice, as far as engagement?

Atlee: Yeah, it’s definitely all of those things, the offline/online. We do focus a lot more on the online side. Our headquarters is in Ottawa, Canada. Most of our company is in Canada and we have less of the easy-ness of this place, so we really focus on engaging our community online.

Romain: For us at Stripe, it’s pretty unique I would say, in this panel. Because if you think about what you’re trying to build at Stripe, we’re building fundamental infrastructure for all developers out there. We are essentially thinking of that as a problem that used to be a deficiency in the internet infrastructure that we’re trying to fix. As a result we live and breathe developers all the time at Stripe.

The goal for us, for the program, is to think deeply about the challenges that developers and founders meet when they start their businesses, and how we can provide them with all the tools they need to be successful. It doesn’t matter if they’re building a marketplace, an on-demand platform, or maybe a crowdfunding platform or SaaS business.

For all of these business models and use cases, we want to provide them with the best set of tools and the best developer experience possible to be up and running very quickly.

David: Awesome. Cool, thank you. Hopefully that gives you a quick idea of the programs that they’re all running here. Tristan, I’d love to hear from you. When we spoke, you were talking about Slack being a platform that you used a lot so far for your community. But you’re starting to hit that ceiling of how far Slack can take you as far as data, and how you’re managing the program. Can you speak a little bit about where you’re at and how you’re thinking about transitioning from this first platform and tool that you’re on, and how you’re getting past that ceiling?

Tristan: Absolutely. We started our third party developer community Slack about a year ago. At the time we were thinking about creating some kind of forum experience, hosting a lithium instance or dedicating ourselves to being excellent Stack Overflow answerers and trying to build the community there. Reddit was another idea we were thinking about, but we stuck with Slack because it had in our minds the best end user experience at the time.

Somebody could get on to the Slack team that we had set up, talk to somebody in real-time, whether that was an engineer at Square or someone on the support team or a product manager even. We felt that was the best experience for them. So we walked along that way, and at the time was a beta community. It was only people who were doing advanced beta testing, or invite only programs. And then we opened it up and it grew bigger, and bigger, and now we’re at the point where it’d be really nice if we had some more tooling. Our use case for Slack is fairly outside what the designed use case for Slack is.

David: I think that’s most people’s use case of Slack, is outside of what they’re designed for.

Tristan: Yeah. So our particular situation, something like self-service onboarding is really important for us. The free tier of Slack has that great, “You can just have an e-mail link, and do it like that.” But when you think about, “We want people to just ad-hoc and onboard onto our Slack team,” the pricing structure is maybe challenging there. The tooling that’s available is not designed for that in terms of your ability to moderate what people say, in terms of your ability to try to enforce community guidelines. That can be challenging and that’s something we luckily haven’t had any giant issue with so far, but it’s definitely a concern that we have moving forward.

As we’re getting to the point where our community is really taking off and if you can imagine, if you have any active Slack channels at work, we have #SFO which has roughly 2,000 people in it. And you can imagine, it’s pretty busy, and most people leave as soon as they join it. We’re at that point where it’s like, “There’s a lot of conversations going on here. How can we better shove people into channels? How can we make sure that people thread?” We’re hitting the limit there.

David: Do you know what platform you are going to look into next? Or are you in that process right now?

Tristan: We have a lot of the research from our time when we wanted to jump into Slack. A more traditional forum is the direction we will ultimately go. But it’s challenging now, because if you think about the support or even the community aspect, jumping from real-time community where you have almost one on one relationships with some of the people you’re working with on the other side, into a more, “I guess I’ll make this post, and then four to six hours later I’ll get an e-mail about going back to the forum to then check out my post to write a response that will be responded to in four to six hours.” So thinking about that experience, it’s going to be a hard sell to say, “Check out the forums.”

David: Yeah. It’s definitely a shift of expectations. We see a lot of companies in crypto running into that now. Where they’re like, “We have 100,000 people in a telegram group. What the hell do we do now?”

It’s interesting to see a lot of companies that start with chat forum communities because it is very fluid, and easy to get up and running, easier to build engagement. But then you hit that upper ceiling where it doesn’t really work anymore.

Tristan: Definitely.

Roadblocks to Scale

David: Atlee, I’d love to hear it from your perspective. You have 700,000 active shops, 25,000 web apps. Your program has now scaled up to a pretty significant point. It seems you’re pretty far along in that journey. So we would love to hear, were there any roadblocks like this? Or any key things that really looking back or reflecting on it, you could advise others on how to avoid or how to navigate scaling up a community like this?

Atlee: Absolutely. So, tooling. I think at every stage it’s just always the question. Because as soon as you figure out your tool it already doesn’t work for you anymore. You’re constantly trying to figure that out. Actually, on the Slack thing, it’s been interesting. We don’t have a Shopify community Slack, the community started it and then started inviting employees to it. Which was smart and it works better in a lot of ways because the expectation isn’t there that we would necessarily go into it and answer or provide support there. They create the rules themselves on what channels stick, which are shutting down, what are the rules. That has been an interesting thing to watch.

Scale for us came through team structure.

When we started, we started like everyone else. Everyone on developer relations did everything in developer relations. They were going to events, and they were writing blog posts, and answering people’s questions, and all the things. What we had to figure out was what was the first thing that we were going to take from this group of people and specialize, and go and have people who just do that thing. And for us, it was API support for partners. Because Shopify is a platform, we have an amazing support team, but they support our merchants who are building Shopify stores. They’re not equipped to answer complicated API questions. So that was the first thing, it was an experiment with a self-taught developer on our team.

I was like, “All day, just answer API questions. Let’s see what happens.” And it just exploded. And then of course we hired a bunch of people, and then we hit another roadblock with that because then the expectation was, “You’ll always get your API questions answered even if you don’t have an app in our app store.” So we had to start making rules around how that works or doesn’t work, and then how do you scale that. We started peeling away different things like this, and now no one has the title of developer advocate at Shopify. Because all the jobs got given away and everyone has a different part of the job, and we really try to focus on making sure that the soul of the thing doesn’t get lost by how we communicate. But really, it was scaling the people and understanding where we needed them to be.

David: And you have a really interesting program with the go-to market team, I would love to hear a little bit more about how does that team specifically function within the larger developer relations ecosystem?

Atlee: Totally. That was the next step from that API support team that we started. What happened was not only third party developers were calling, but internal developers were like, “How does this work? And why is it not working?” That was also the team that when a well-meaning team somewhere else in the company made a change and then everything broke, they were the people who heard about it first. So we spun out another team from there that we call API go-to market and it’s a SWAT team that works internally with other product teams in the company all the way from ideation through development, launch, and then maintenance.

Actually, all the way through maintenance and deprecation. They really build the template on how to do it, they’ll set up data groups for you, they will do tutorials. They will take your idea and tear it to shreds on why no one in their party community is going to even want this thing. It’s proved to be really helpful internally because everybody is committed to building a great platform, but sometimes as most people in this room probably know, product teams get so focused on building these amazing features and amazing parts of the platform, that they are not always considering what other people would do with it.

We brought this team in to just be those people internally, and they are a very small team and they are amazing. I always say, “Their biggest problem is I talk about them too much.” Because they’re like, “Atlee, you have to stop telling people that we exist. Because there’s only four of us and there is only so much we can do. But it’s something we’re going to invest in more because it’s proven to be really helpful.

David: I’m hearing that you had a general developer evangelism and relations team, and then you started seeing very specific needs. Maybe it was based on a specific group within your community that had specific questions that not everyone could answer, or just different kinds of programs or needs that your organization had. And as you saw those opportunities come up you started to specialize a team in more specific ways.

Atlee: Yeah. Then get them to choose the tools to best serve those specific needs, because we deemed them the most important needs of the group.

Selling to Developers

David: Awesome. So Romain, your community is unique in that everyone in your company is developer-facing. All of your customers are developers so in some ways the whole company is involved in developer relations. I would love to hear, how has developer relations specifically developed at your organization? And how were you able to scale that over time?

Romain: Yeah sure. What’s interesting, as you pointed out, at Stripe all of us think about developers all the time. We want to have all the founders, builders and engineers of the world to basically be very successful when they get onto our platform. As a result you can take a step back and think about developer community as a function. At Stripe it sits throughout the whole organization. From marketing to communications to partnerships, to our group and product management.

We really think about developers all the time. We have a developer experience group, that’s part of product and engineering. That’s why we spend so much time refining the product, talking to users and trying to understand their challenges and how we can meet them. We’re taking a step back.

The entire company thinks about developers all the time, and how we can help them to always be the most successful with great APIs and great tooling to build their products and services.

David: Has it been really unique from what you were doing at Twitter? Being a consumer-facing company that just had a developer evangelism team?

Romain: Yeah. At Twitter we were focused on driving growth and adoption for our APIs, and our mobile SDK called Fabric which is now part of Firebase at Google. But it was very focused on that. Driving the adoption of one specific suite of platform and API, and API support developers. Which was fairly separate from the consumer product that we all know and use every day. But at Stripe, when you think about our suite of products, they are products for builders. Whether you’re a founder of a company who wants to use Atlas, or whether you want to start a business of any kind, that you’re going to use one of our API, it’s ready to be up and running.

So, it’s a very different approach. And what’s interesting is again, to the point of having multiple teams thinking about developers, we tend to work at scale on my group. Thinking one too many and also engaging with developers and users to really refine the product and think about how to build the best developer experience. But we also have teams that will be more user-facing, and building those relationships one on one. For instance, we have a team called field engineering. And you can think of them as the experts of all the details of the product that will help developers, founders, and large companies and engineers to be successful when they integrate the product. Essentially, yes, all of us are really thinking deeply about developers.

David: That’s an interesting question too, of distinguishing between one to many engagements. How do you as a small team manage this large community of developers? And when are the opportunities or the moments where you need to be more hands on and one to one? Do either of you have an opinion on specific areas that is more important to be hands on and one on one, versus areas where you can be a little bit more general and hands off?


In some ways, having to do API support indicates that you’re not doing something right.

You should have amazing documentation that explains all these things, regardless. But that’s a constant. Every day that changes. I would love to say, “Yes. I’d love to have amazing docs and no one has any questions about our APIs ever.” That is a goal. That’s why that team is only four people.

David: We joke around that a community manager’s job is to put themselves out of a job.

Atlee: Right, exactly.

Romain: But we’re never successful, thankfully.

Tristan: A challenge that we, in particular, share is that our user base is on a spectrum of, “I am a bootstrapped entrepreneur who has never worked on a computer before, but I’ve gotten this far by myself so I’m going to get whatever website thing out the door,” and then we have people who spend all day in front of a computer developing solutions for a merchant. So one of the challenges, particularly in the support context, is that we have people from a wide variety of backgrounds and skill levels and interest levels, even. Some people who are coming around like, “Who can I pay to do this for me, because I don’t want to do it.” Those are all opportunities to find, “Maybe we should get a consultant marketplace going, because this is a big need for our community.”

David: So there’s always the opportunity, if you need more hands on assistance for people, to let them be hands on with each other as much as possible.

Tristan: Yeah.

Romain: It even goes beyond just support and assistance. It’s not even just about fixing problems or places where the documentation could be clearer, we also see it as a way to build relationships with founders and entrepreneurs and see how we can help them. In many ways when they reach out to us it’s mostly about building that partnership, which is interesting. For instance we have a product called Connect which is the product that our marketplaces use.

What’s interesting is, a few years back before that product existed, we had a ton of people asking questions, “Could I potentially have thousands of bank accounts in my Stripe account?” And then we started to work with them and alongside them and realized, “Wait. What you’re trying to build is actually a marketplace. There should be a product for this. What you’re trying to do is very legitimate but the product does not do it yet.”

It’s not only about the support side, it’s also about building those close relationships and empowering those founders and developers over time.

Atlee: I completely agree. One of those teams that has emerged from this is a team that exclusively does that, and just talks about product all day long with developers that we know are committed to the platform and are committed to evolving their product over time.

Romain: That’s great.

Fostering Community Amongst Developers

David: I’m curious. Something that just came up for me is hearing there is a lot of programming that goes into providing support, answering questions, or collecting insight and feedback. Which is great. It’s all the things that’s driving value for the business. To what extent are you focusing on just purely creating a sense of community amongst those developers? Just the social capital, regardless of what that does for your organization. Whether that’s relationships between you and the developers, or creating those relationships amongst them.

Tristan: We’ve started doing that a little bit more as our community has gotten to a critical point where we have people in it all the time. We’ve started doing more individualized community-focused events, where we have maybe some of our product managers do and “Ask Me Anything,” an AMA inside the community. And that’s been really interesting because everything that somebody says in one of those AMA sessions, they’ve probably already said in the feedback channel or delivered it in a support ticket somewhere.

But this gives them the opportunity to feel heard about what they’re talking about and then also get our perspective on it. A lot of times it’s a product manager just typing away like, “We’re considering that in the future. We’ll get to that someday,” in this small use case. But that opportunity to talk to them, and because of the real-time chat aspect you can feel like you’re talking to somebody on the other side.

David: Just to clarify too, that still sounds like it’s on product and connecting around the product and what you’re building. I’m curious to what extent you focus on just, “Let’s get these people together, and develop relationships amongst them.” Or create almost more recreational experiences that develop those bonds, that will probably lead to a whole lot of business value but are just purely, “How do we build community?”

Tristan: Going on that we’ve also done some events with community members as the focus. So trying to organize people in different geo-locations. Like, “You guys all work with Square together. Want to get a coffee sometime? Swap expertise on consulting? That kind of thing is another avenue.

David: And you’re organizing those, or are the members self-organizing?

Tristan: We try to get them to self-organize, because there’s a lot of people out there. We’ve done some events that we host ourselves when we’re on the road. But primarily that’s taken a backseat.

David: Any insights on that from either of you?

Atlee: Yeah. We have a meet up strategy, I guess it’s a strategy. Where we say, “If people in the community want to hold a meet up, we will support them to do that.” And that is self-run by people in the community. One of the investments that we made starting last year was something that we now call Pursuit which is a six city stop including North America, where the content is 90% partner-delivered, but we produce it. It’s a high-production event and we have a different person from Shopify keynoting each, but all the other content is all partner-delivered.

That has been really interesting because it’s enabled us to reward great partners by putting them on a platform that gets them in front of other people in a highly produced situation.

It’s a reward mechanism, but it’s also been a great engagement opportunity.

And we have seen some of the cool stuff that we’ve seen like people starting partner programs, like Shopify partner programs amongst themselves out of that event.

That is a community person’s dream and we didn’t even touch that, that was just them figuring that out. So yes, providing space. But what we’re seeing is that there’s lots of spaces where people can just get together, if you want to create a space for them to get together around your thing, you have to provide more of that packaging or that thing around them.

David: Got it. Romain, anything to add there?

Romain: Just to make it short, we’ve also been running some events at Stripe. Mostly focusing on user education events, trying to help them understand the latest products we’ve built, the latest features and how they could benefit from those. And also to your points, trying to make users meet with each other. We’ve also focused a lot on the online communities. For instance we have this product called Atlas which is specifically designed for founders who want to start their companies and get them off the ground.

We have an online community for Atlas founders where they really help each other at various steps of their journeys. So that’s been very interesting and when it comes to the offline community, we’ve not been in too much of a rush to organize those. Because we’ve seen a lot of appetite already and people are kind of self-organizing and helping each other get started on Stripe.

Key Metrics for Developer Relations

David: It sounds pretty consistent with what we see. You just build a lot of value and you start to do your own things, and then the community starts to organically want to do these things. Then you slowly build structure around it to support them. Awesome. I guess a good segue from there is, there are the community programming and events and all these nice things that you do to just create more engagement, and then there’s very specific things that you’re trying to do to extract value.

Whether it’s product feedback, or marketing or support, or whatever. What are the key metrics that each of you are looking at? I’d love to hear both on that community engagement level, what are the metrics that you look at? And then are you also tracking the business objectives, and the way community is driving that? Tristan, do you want to start?

Tristan: Sure. From a business perspective the benefit of being a payment-oriented company is that you can measure money in all sorts of different ways. So we definitely keep track of what kind of value our–

David: Yeah, you have it easy.

Tristan: We definitely keep track of what kind of value our developers are bringing in to not only us but also the merchants themselves.

Even small changes in retention can make big impacts for our bottom line.

We do keep track of that, we keep track of something like your weekly monthly active developers and how many applications are in the pipeline. But from the community point of view we’re really trying to figure out a good metric for health and engagement in a small period. So one thing we look at quite a bit is the weekly number of people who are active in their posting method. They’re posting something somewhere, or they’re making a post on Slack, or responding to something online. That has been our proxy method to see, “Are we trending up or trending down?”

David: So is that like total number of people who’ve taken an action in the last seven days?

Tristan: Yeah, more or less. We break down into different groups. I focus on our Slack community quite a bit, as you can probably tell. We take a look at like, “We did an AMA with our head of product. How did that change things for the next seven days?” And then then onwards. We break it down by different sub-communities where social is its own weird thing. But those are things that we try to keep track of on a regular basis.

David: Awesome, super helpful.

Atlee: On the community side, we’re focused on what everyone here is focused on. Engagement on the various platforms, social platforms, attendance at events, etc.

David: To clarify real quick, on social platforms, you mean the social media platforms. Public platforms.

Atlee: Yeah. Our biggest online community is actually our Facebook community.

David: The page?

Atlee: Yeah. So we look at that. And it’s very engaging. It’s interesting because there’s both a place on Facebook where just partners are, and we watch that engagement. There’s also huge Shopify merchant communities on Facebook. So we watch, and it’s hard to measure because we really don’t have this all figured out yet. But the engagement of partners on Shopify merchant pages, because that is where that ecosystem gets really interesting. And we also want to try to figure out how to reward people that are actually helping merchants, and not by just saying, “You want to do marketing? You should download my app.” We’re looking for that. I would say though, right now is a little bit more qualitative and then when we catch it. So that would be on that community side.

On the business side, yes money, that we get to watch the money. We definitely look at that. We always look at the average amount spent on apps by merchants, and watch that over time. We look at that in different bands of merchants in terms of how much GMV they have to understand what’s going on. And actually how big the “economy” is.

We also watch referrals because lots of our partners bring us new customers.

I think though where we’re really trying to zone in right now is that place in the middle, which is the activation of merchant value.

The place where a partner provides value for a merchant. And sometimes that’s in installing an app, but it could also be, we have a services marketplace where people can come and offer to do a new banner for your store. And we have ways to track that, but we’re really trying to figure out how we identify those moments where value is produced by a partner and given to a merchant. Because we believe that that is driving the business forward in terms of keeping people on the platform, and making them more successful, which ultimately makes us all more successful.

David: That’s cool. It sounds like you have a few different objectives that you’re looking at and metrics assigned to each one.

Atlee: It’s like a funnel, that I explained in a not cohesive way.

David: That’s probably why.

Atlee: But that middle piece of the funnel is not just about activating partners, it’s not saying, “Look. They submitted an app to the app store.” It’s when they hit their 1,000th API call and money is being exchanged for that activity. And trying to refine that and really figure out when that’s happening and why, and what the value is that’s being created there.

David: Awesome. Romain?

Romain: In terms of metrics I would say all of us at Stripe produce a lot of content, whether it’s for founders or entrepreneurs who will take a look at those metrics. But to be honest I don’t think we obsess too much about those metrics. We don’t obsess over the reach of those goals. What really matters the most to us is the metrics around the health of the integrations. How easy it is for a new developer or a new founder to get onboarded onto the platform. How long does it take them to make their very first live payment? Are they really set up for success and are they ready for scale? Is their integration healthy enough? Are they making any API errors, or should we guide them in a particular different way? And all of those things, and developer satisfaction and the importance of delight.

It’s really hard to measure delight.

But for us that’s the things we obsess about, and taking a holistic approach on the developer experience and making that as best as we can. Because we’ve seen the better the product the better your engagement. If we make the best product and the best API, developers will just stick onto the platform because it’s so easy for them. And I would also point out, when we think about developer communities it’s really easy to fall in the trap of looking at big metrics and your reach and so on. But it’s important to also keep in mind the things that don’t scale.

We care deeply about those. I don’t know if some of you heard the story, but in the very early days of Stripe Patrick and John Collison our founders, would actually proceed and do what is now called the “Collison Installation.” It’s now referred to that at YC and it’s part of the program essay, if you’ve seen it do things that don’t scale, if you wanted to read more details. But if they had an interest from a founder willing to use Stripe, instead of sending them an email with a link, they would just tell them, “Give me your laptop and I’m just going to get it done for you in a few minutes.” And that was it.

So this importance of doing things that don’t scale still transpires to this day. I spend a lot of time and my team also spends a lot of time talking to users one on one, really trying to understand, “How can we do better? How can we help them solve more challenges in the future?” There’s clearly the lens of getting the product better and better. But also the lens of not forgetting users first, and doing things that don’t scale at times.

David: Got it. So you heard it here, the best way to build community is to grab their computer from them and just install shit on there.

Tristan: Just get them in the community.

DevRel Wisdom

David: That’s right. So this is going to be rapid fire and then we’re going to open up to questions from the audience. So, last question here. It’s a small one. I don’t know if you know this, but you’re all retiring tomorrow and you have to sum up all the things that you’ve learned in your career as a developer relations expert and developer evangelist into one concise, brilliant, memorable, message that anyone here can walk away with and apply to their career. What would you say?

Tristan: I’m ready.

David: Tristan, you’re first.

Tristan: The concise thing is empathy over everything. One thing that we do quite a bit is try to walk in our user’s shoes, particularly around the community. And see how would somebody discover it. Like, nobody is going to walk off the street and say, “I really want to join Square’s developer community.” So we think about the touchpoint they might have. A lot of times they’re connected to a merchant or something like that. We think about how they’re going to discover our community, how they’re going to interact with our products and try to figure out the different paths and channels where they might get mixed up. It’s particularly important because of our merchant developer spectrum.

David: It started out really concise.

Tristan: “Empathy over everything.” That’s the tweet.

David: Awesome.

Atlee: Trust is built on a foundation of managing expectations.

David: Can you say that again?

Atlee: Trust is built on a foundation of managing expectations. That sounds sort of disheartening, it’s not supposed to be at all. It’s because in some ways, managing expectations might sound like, “Limit your ambition.” That’s not it at all. It’s to the empathy piece.

Really understand why you need a developer community, how you’re going to create business together, and why.

And then be honest about where you are and how they can participate. Because that’s the only way you’re going to keep them forever. Because we all have a million examples of people going out and selling a dream, or an API that’s just never going to exist. And that can ruin it for forever. So I think it’s embodying empathy and honesty and then communicating it.

David: No pressure, Romain.

Romain: Yeah, no pressure. My answer will be twofold. If you’re just about to get started to build a new developer product from scratch, my advice there would be to find a real problem. Try to find that problem and provide a solution for it. At that time you can do the things that don’t scale. Find some users. Doesn’t matter if the size of the audience is super tiny. It might be a small niche, but if you have a real problem and a real solution for that problem, I think you’re onto something great. And the second part of my answer would be if you’re already having a product, you’ve built a developer product or an API, and you want to grow your audience and take it to scale.

At that time my advice would be to obsess about product. Make the best developer experience you can build. It doesn’t matter, you can build a community on the side and do all of those things and those meet ups, but ultimately the best product will win among developers. There is a saying which says, “If you build it, they will come.” That’s not always true for every product. But we are very lucky in this room to think about developers all the time. Developers are pretty unique when you think about it. Developers are passionate about their tools.

It’s funny when you think about it, if a developer is passionate about API, developer tools, their libraries, their frameworks, they tell their friends. They tweet about them, they share them on Hacker News, they write blog post about those. That does not translate to other industries. So we are very lucky to have that fundamental piece that the developers are passionate about what they do. So build the best product you can, build the best developer experience, and ultimately, they will come to you.

David: Awesome. Thank you. We’re going to open up to some questions from the audience before wrapping up here.


VIP Components

Tristan: In regards to a VIP program, it’s something we’re definitely working on. We’re still trying to figure out the best way to do it. We are starting internally so we’re trying to do a little more work in getting people who aren’t really related to our developer platform to work as VIPs in our developer community, giving talks at meet ups and that kind of thing. And then for in-person, primarily our in-person events now are all top of funnel awareness building. We’ve done some meet ups and other in-person events but that’s by the wayside compared to our top funnel building.

David: Do either of you have a VIP program?

Romain: I wouldn’t qualify it really as VIP, but what’s very interesting and top of mind for us, because it was just last week we launched a new partner program and we are very lucky to have Shopify here as part of it. With the partner program we are thinking about it as a subset of the Stripe developer audience, but a very critical subset. Because these are the partners that are building on top of Stripe’s infrastructure. And in the case of Shopify they provide great tooling for businesses to accept money online and run their businesses online, but you also have other types of companies. For instance Segment and Avalara who are extending what Stripe is capable of offering. This partner program already has thousands of companies and that’s probably the closest of what you have in mind for VIPs.

Atlee: Can I ask you a question?

Romain: Sure.

Atlee: Because I saw that you guys decided to have a fee for verified partners. How did you guys make that decision?

Romain: I was not very involved in that, to be honest.

Atlee: That’s a good answer.

Romain: I could find the answer for you.

Atlee: OK. I’m curious.

Romain: We’re also experimenting. It’s pretty new for us.

Atlee: It’s been very successful for Apple, but they give tools in exchange for their fees.

Romain: For us, to be honest all of us are in the boat at Stripe of, “We’re just getting started.” This is the very beginning of trying to organize what has been so far very organic. We’ve been lucky over the years to have thousands of companies already extending Stripe in some ways without us really doing the work of organizing that effort. So we’re just getting started and we’ll learn a lot along the way.

Atlee: We have what we call experts, but traditionally they have been in that web design group of partners and we’re currently trying to figure out how to extend that to our app developer partners. There’ll be more information on that soon.

Nascent Community Recommendations

Tristan: I have a talk I did last year about Stack Overflow. It’s great. But it’s really hard to ask non-technical questions there. So if those are your options, depending on your growth Google groups is going to really hurt you, Slack might be the best group there but it also has its challenges.

Atlee: You just have to figure out a way to get whatever is talked about on Slack in a living document somewhere, because that’s the biggest problem. Especially at the beginning because you have to onboard people as quickly as possible and you don’t have the time to do that. So how do you extract that information so that the same questions just aren’t asked a million times? Slack didn’t exist when we started, so ours was mostly our own forums. Mixed. That’s a crappy answer.

Tristan: It’s a tough situation.

Romain: It depends on the nature of the product and what you’re trying to get from that community. For instance, if it’s open source or not, if it’s an API you’re going to have different answers I guess. And Stack Overflow is more this place where you’re going to have questions and answers. You want to have this answer being the canonical reference for the foreseeable future. If you want to have more chatter and more trying to build relationships with users and understanding how they could use your product and so on, that’s what’s not going to be taking place on Stack Overflow. So it really depends on your goals and the type of engagement you’re looking for.

Marketing Budget & Conference Investments

Romain: Conferences are very interesting. For us we’re mostly seeing them as opportunistic. Every time there is a conference that would make sense for us to speak at or potentially we can provide value to the attendees, then we’ll happily go. Hackathon, we’ve done them a bit less lately. When you think about the nature of the product at Stripe it’s usually not where you would start. If you’re building something very new from scratch for instance, maybe a product like Twitter is more suited to stop to hack around versus Stripe where you probably will use Stripe for payments slightly later down the road. Of course it depends on the nature of the product. But yeah, conferences are mostly opportunistically I would say, and a bit less Hackathons lately.

Atlee: We use sponsoring conferences for very specific things. We’re really trying to do a better job of bringing diversity and inclusion into our community. So we’ve spent money sponsoring those types of conferences. But we have found that our best way to spend our money was to hold our own conferences. We hold an annual developer conference called Unite. And we have this pursuit series that continues off of that, and that is where from an event perspective we’re spending most of our money.

Tristan: For us, we’re in a marketing function so all of our money gets spent on marketing. We’ve actually really ramped up our conferences and speaking recently. But our developer evangelism in time, very little of it is spent on that kind of stuff because we’re mostly building content and things like that. On Hackathons in general, for the college Hackathon scene, the market has shifted towards people getting the most value out of that are the people organizing college Hackathons. Particularly, like you said, it’s hard to say, “Use a payment processor for your college Hackathon weekend.” Some kind of cool social media would be way more compelling of a use case.

Finding a Developer Evangelist

Atlee: You should go to David’s conferences, they talk about this all the time.

David: We actually help a lot of companies hire these kinds of roles as well. I said, “A lot of people end up hiring from their community,” so it’s just somebody that organically becomes really active and takes that role, and then you just make it more official. But even in this room there’s a ton of people who are experienced in being able to communicate with developers and speak in more technical language, while still enjoying the practice of organizing events and interacting with people and driving that engagement. Ask everyone here how they got their jobs and you’ll probably find out how to find them.

Atlee: I would say it obviously depends on what you’re doing. But as a platform company and not an API, we don’t sell our API. We could be a lot more liberal on who our developer evangelists are, and that served us well. Because I hired people who were not developers. And in some circles that is viewed as the worst thing that’s ever happened to developer evangelism. However, if you have a mix of people, you can leverage depending on the event or the partner or whoever it is. You can pick and choose. The other thing is this is where boot camps are really helpful. Because you have people who have demonstrated that they want to learn something new, that they’re willing to give up their lives for 12 weeks to do it, and they get enough out of that if you meet them and you’re like, “You seem like you’re desperate to learn about our technology next,” those are great junior dev relations people in my mind, again, from a platform perspective.

Tristan: As a shameless plug, you should definitely join the SF DevRel meet up. There’s a lot of people who are looking for jobs there or who might be convinced, and I run it. So it’s great.

David: I think that’s it. Please give a huge round of applause, 12 out of 10, for our panelists.