Everyone is an evangelist

Justyn Kestelyn: Morning, you know.

Speaker 2: Some equations, some history, some future, Star Wars, super fun. The next ninety minutes, you will hear all about how do we scale developer relations. That particular topic is very close and near to my heart. I've worked at bigger companies and smaller companies. It's a lot more important if you're a small team, how do you really scale your developer relations?

What sort of things you need to do to scale it? So our very first speaker in that fragment is Justin Castellan. I've known Justin for quite a few years through my working at Oracle, and then we have kept in touch. He's of course local to the Bay Area. Let's hear what a law school dropout has to talk about scaling developer relations.

Stand up, don't be shy, stand up. Everybody stand up. So thank you

Justyn Kestelyn: Arun for that introduction. I guess that's sort of interesting. One thing I guess that is kind of interesting about myself is I went to high school in a place called Laguna Beach. I don't know if you guys have ever been there. Yes.

You have? Yeah. Nice place. Nice place. You know, I'm not from there but I went to high school there.

And I was thinking about who I went to high school with and it's kind of amazing. So at my high school we have the co founder of the Apache Software Foundation, an Olympic gold medalist, somebody who's on the FBI's top 10 most wanted list, and the managing editor of Rolling Stone. That's pretty amazing, and me, you know, for what it's worth. So today I'm going to talk to you about evangelism at doing evangelism at scale. Which is something that hopefully all of you will have to experience at some point someday.

If you're working at a company large enough and with a community that's large enough where this actually is something that's useful to you. Anne, thank you very much. So to tell you a little bit about myself in terms of my background and why I can talk about this subject. So I am today head of technical evangelism developer relations at Cloudera. Which is a company that's heavily involved in the Apache Hadoop ecosystem, if you heard that.

And I use the word head loosely. There's only one other person doing this function besides me at the company. But I'm gonna use that title anyway. So I held a similar role at Oracle for a long time, for probably, I don't know, nine years. Which is where I met Arun and some other people in this room.

And that was a really, really interesting experience because during that time, as you guys probably remember, Oracle acquired I don't know how many companies, for better or worse. I think the count was 40 or 50 or something. And a lot of those companies had developer programs and community outreach programs of one kind or another. So during that process, as you might imagine, those programs all sort of went through an x-ray machine. And what we ended up with is what I would consider really a Star Wars bar of communities.

I mean I've seen absolutely everything ranging from proprietary to full open source to everything in between, small, large tools, platforms, you name it. Pretty much everything you can think of. So it's kind of a nice range of experience. And even before that I was actually a technology journalist. I worked on trade publications which no longer exist because the publishing industry, as you guys know, has died especially in the technology area.

Close. Close to died or at least transformed. And Josh's presentation was really interesting to me in terms of looking at the roots of our whole function. Because it got me to thinking about how magazines and publications, at least back in those days, were actually proto communities. They were like the original communities.

And the reason I say that is because they were premised number one completely on contributions from the people in the community. And number two, on the fact that their mission was to make people better at their craft, whatever it was. And I think the consummate example of that was something called Doctor. Dobbs Journal which no longer exists sadly. But I think that was a great example of what I'd call a proto community.

So let's get into the meat of things here. So I don't think I need to talk on this slide too much. You guys are all in this room so you know the difference between marketing and evangelism. So I'm not gonna belabor this slide. Suffice it to say, in summary, marketing is about push and evangelism is about pull.

It's about attracting people to your community. Again, because they see a way to make themselves better at whatever they're doing. So that's a huge difference and it's a lens that you basically need to look at everything through. So we have this thing in evangelism called the professional evangelist. I think somebody already asked this question in this room.

How many people in this room would consider themselves professional evangelists? Okay, there's a few of you and I would expect to find you in this room. Outside of this room, there are not that many of you. So here's how I would characterize people like you. First of all, if you're gonna be good at your job, you are certainly technically proficient at whatever technology you're involved in, whatever the case may be.

You're prolific. So you're not somebody that just produces two blog posts a year and a couple code samples and does two or three talks. You probably do a lot more than that. You're creative. You find new and interesting ways to talk about things that you know that your audience and your community is interested in.

You're connected. You know a lot of people. You're always meeting new people. And probably most importantly, you're a great communicator because you can have all of this stuff in your head but if you're not good at communicating, it's going to stay there. And actually Arun to me is the consummate example of one of these people who can do all of those things.

There are people who can do one or two of these things very well. It is very, very, very rare to find a single individual who is that complete package. And across my career in this area which is probably going on a dozen years now, I can only think of a handful of people that I would put on that level. Those people are really rare. So if you're premising the success of your company or your organization on these people, on finding these people to carry your message to your community.

And I'm not gonna go again to the details about why that's important because you guys understand that and some of our previous speakers have covered it already. But good luck. I mean we're talking about unicorns here. This is our second unicorn reference today I think. I'm sure there's gonna be more.

Now if any of you are in the big data area you're probably aware of conversations lately about the data science scientist role. You know the data scientist is this ultimate example of a unicorn today in this area. I would argue that the person I just described is an even more rare unicorn. It's probably a double horned unicorn. Even harder to find.

So let's assume that you can even find one or two of these people. They can't be everywhere.

Speaker 2: Except of Josh.

Justyn Kestelyn: Except and Arun. And they have to sleep. Yeah, you can put them on a plane for nine months a year and expect some incredibly crazy level of commitment and output. But eventually they're gonna get burned out. And it's just not scalable.

It's not efficient. So instead of looking for these people and trying to hire one or two of them, why don't you create a herd of unicorns? Wouldn't that be a better way to approach things? And to scale out this entire process? And yes, maybe you're not gonna create a bunch of the people I just described but I bet you're gonna find people who can do one or two of those things really, really well.

And when you combine forces you end up with something that, a program, a full program that actually is just as efficient and just as prolific and productive as hiring one or two people to do this thing that I just described. So let's talk a little bit about building a culture of evangelism at your company. And when I say building a culture of evangelism I don't mean just internally, I mean externally as well. So we're talking about not only people who work at your company or your organization but also people who are your community in other words. So in other words, don't turn your own community from consumers into contributors.

So the point here is to empower and enable these people again to become evangelists. To do the things that they're good at doing. To carry that message that is so important. Whatever that message may be for your particular company. Whatever leads to that conversion that Adam just talked about.

And in order to do that, the most important skill you can have is how to identify talent, evangelistic or evangelism talent, however you want to call it and then nurture it. Whether those people are employees or they're in your community and they're your customers and your users. It's kind of irrelevant in fact. I would argue that it's the most important job of a professional evangelist to do that. So in other words, stop thinking of yourself as a player and start thinking of yourself as a player coach.

So not only are you somebody who practices evangelism but you're also somebody who can create other evangelists and can inspire other people to contribute and to serve that function in whatever means they have to do it. Now, before we get into a little bit of detail about that, you need buy in from your wider organization, of course. And I can't talk too much about what that means because everybody works at a different place and they have a different structure with different priorities and different definitions of success and different definitions of a community, different definitions of a user, what have you. There are lots of variables there. But the point is is that this is something that you kind of need buy in from either your superiors or your coworkers or your colleagues or what have you.

It's not something I recommend doing as a rogue effort because it is something that requires again a group effort. So, what does that actually mean? As I mentioned before, you know, I described what an effective professional evangelist is and the kinds of skills that they have. You're not gonna find a single person who can do all those things but you can find people who are good at something or a combination of things. And in this effort it's really, really important to be able to identify who those people are.

To identify their talents. And to map those talents to roles. So for example, there are people who are really good at writing blog posts. You know that's their thing. They are just prolific blog post writers.

They're tough people to meet in person. And they may not be incredibly technically proficient. But they're very good communicators. They can communicate in writing. They're good at that.

There are people who are really good speakers. Great presenters. Again, maybe they're not the best writers in the world but they know how to get in front of an audience and to get people's attention and get a message across and to teach them something. You know, these are often people who work in a trainer role, for example. You know, all these people, they literally just fly across the country and train people to do one thing or another.

Some people are obviously great coders. But they might not be the best communicators. As we know. Those two things often do not go together. Some people are just great connectors.

Know, these people who run meet up groups and user groups and run conferences and they're just prolific in terms of bringing people together and getting people to meet, to create connections and to meet each other and to share things with each other. And they're creative about how to do that. The organizer of this conference, obviously, I put them in that category among other categories. They find creative ways to bring communities together and to get them to share stuff. That's a talent.

That's not something you just are born with. Or, well, maybe it is something you're born with. It's a talent. And then of course there are helpers. There are people who maybe don't do a lot of these things very well but they are good about persuading people to contribute and to become part of the community and to teach them to understand why that's important and why that's in their best interest.

So I think the responsibility here is to be really good at identifying these people either in your company or in your community. And you know, there are some tools for doing that. It's not something that happens by itself, obviously. I know that in my past, for example, at Oracle and then working with the Java community that was onboarded at Oracle, You know, one way that we recognized people in the community who we felt were potentially really, really good evangelists was through discussion forums and mailing lists. Because there are people who are really into that, into participating in those kinds of communities.

They might not write blog posts. They might not be presenters. But there are people who just love answering questions. You know, those people are really important potential evangelists for you. I mean, I'll just give you an example from Cloudera.

We have an employee who works in support organization. And the reason he's an employee was because he was literally answering every question he saw on our original mailing list. I mean, there was actually a theory that there was some conspiracy that there was more than one person using his account because he literally did, it seemed like he didn't sleep. He was constantly answering questions. And he, that is somebody who literally by doing that in a way was just raising his hand and identifying himself as somebody who was really, really good at that kind of sharing and that type, that brand of evangelism.

There are people like that out there. There are obviously also people who are, you can see on the conference circuit and the meet up talk circuit who are really good at presenting one thing or another. You know, you guys can all think of examples from your own communities of people that you've seen give really, really good talks. You know, in the Hadoop community, there are several people like that. And I'm not going to go through the whole list but my point here is that there are ways for you to identify these people and generally they will self identify.

You can't find these people by placing a help wanted ad. I mean, you can but it won't be as efficient for you in terms of finding these people and who they are. They're good at what they do so they self identify. So you want to take a talent first approach from that perspective. So in terms of measurement, I think there's going to be some presentations here later today about measurement specifically and I don't want to get too much into that.

Personally, it's an incredibly hard challenge. And again, I've seen it from the proprietary software world and I've seen it from the open source world. And emphasize some of the things that Adam talked about. In the open source area, measurement is really, really hard. And if anybody in this room has figured that out, please I'd like to talk to you because I'm still figuring it out myself.

It's really, really hard. You know, one of the, historically of course, one of the most important ways to do measurement and effectiveness of your evangelism program is through downloads. And in the open source world, can't do that. You can't gate downloads. So that's taken away right there.

But there are things you can do like measuring visitors and eyeballs in terms of whatever your web presence is for your program. Registrations of one kind or another. Maybe you're asking people not to register for a download but maybe it's for a developer's guide or a white paper or what have you. What we've found and actually I've found in my experience generally is that newsletter subscriptions are actually a really good proxy. Because there are not a lot of things that people will register for.

But newsletters tend to be one of them for some reason. I think it's I think people think generally that giving up some personal information to get a newsletter is not a really, really high bar for some reason. So I found that people who subscribe to email newsletters and email is not dead. I've been waiting for it to die for years and years and years. But it's still not dead.

It's actually a good proxy for measuring effectiveness of these types of programs. And at least, again, in my world, if your outcome here is leads, in my world, in my company, that's not really an effective measurement. Because the people who are using our software are not the same people who are buying subscriptions. So I think it's important, some people get caught up on leads being sort of the end all and be all of measurement in this area without thinking about it a little more deeply because you really need to understand who your users are, who your buyers are and how you measure success. Don't automatically assume it's about counting leads.

Although maybe it will be for your particular business. Boy, this presentation's really short. I think generally speaking evangelism is like a garden in that again, through some of the things I've described, your job is really to encourage blooms. So you want people to become evangelists. You want to again identify and nurture talent.

Which maybe I'll talk about a little bit more. You want that to happen. But sometimes there are also weeds that need to be pulled. There are sometimes things that are not necessarily good for your community or good for your message or what have you. And that's activity that although it's out there and it's great that it's organic, you don't necessarily want to encourage it too much actually.

So you have to recognize that. I guess my point here is that not all activity is good. Sometimes it's not particularly healthy for the community. So you have to recognize the difference between those two things. So in terms of encouraging those blooms, let me give you some examples.

At Cloudera, we have some people who work in engineering and in our technical services team, which is like professional services, who are really, really good at these things. At one of these things or another. Know, they're really good writers or they're really good presenters or they're great at writing patterns and sample apps and things like that or videos or what have you. So how do we nurture that or how do I nurture that? Really, I sort of almost think of myself as like their agent.

I think that's a good way to think of it. So there's one particular guy named Ted Malaska who I was just talking about earlier with somebody in the audience with Pat here. This guy is just a genius presenter. He doesn't write very much. But he is an absolute genius in front of an audience in terms of carrying a technical message.

And I sort of think of myself as Ted's agent. So when I see a conference CFP come over the wire for something that I think is relevant to what Ted likes to do, I will literally submit on his behalf. I'll ask him first, of course. But I will work with him and encourage him to present at that conference and be very, very proactive about it. So, it's not enough for me just to tell Ted, hey Ted, I think it'll be a great idea if you do some more conference talks or do more meetup talks and then leave it at that.

I am actually really pushy about it to the point where Pat told me that Ted told him at a conference that he considers me his second boss. Now, don't recommend that. That's not something I want to be. But again, I feel like I'm Ted's agent. So it's my job to get Ted in front of as many audiences as possible and to facilitate that whether it's through helping him write an abstract, getting permission to travel, what have you.

And that's just one example. Again, if somebody's a really good writer or a really good coder or what have you, there are other ways you can encourage their activity. But the point is, is that you can't be passive about it. You really gotta play the role of somebody whose job it is to facilitate that activity. And again, I would not draw that line just at employees.

There are people who are non employees in our community who I think are really, really good at doing one of these things or another. And if I see an opportunity for them or we have an opportunity to publish a guest blog post for them or what have you, again, I will be very, very proactive about it and go to them and again sort of play that agent role and get them excited about contributing in ways that I know that they're good at contributing. And this is sort of the seed of all good external evangelism programs like the Microsoft MVP program or the Java Champion program, the Oracle Ace program. These are all external evangelist programs that again, that were designed to manufacture evangelists outside of a company. And some of those programs were historically successful.

I mean you could argue that the Microsoft MVP program is probably maybe one of the most successful examples of that and I'd put the Java Champions not too far behind. So, would like to leave some time for questions, which I've just done. I have about five minutes left. I know I kind of rambled on there. It's the first time I've actually given this presentation.

But I would like to answer any questions that you might have about anything I said or anything else. Does anybody have any questions? Yes. Yeah. So, the question was about what do I do in terms of measuring metrics personally in my area?

So, there's not a lot I can do. I do kind of the basic stuff. Like, I will track tech talks around the world that are done either by our own employees or people who I believe are talking about important things. So that's literally counting. That's like A, B and C spoke at DevRelConf or what have you in such and such a city and such and such a date.

And I will just maintain a spreadsheet and I will count those. I will also try to track attendees to those things as much as I can. Sometimes it's not very efficient. I will track website visitors and views to the developer facing portions of our website. I will track upstream contributions.

I pay attention to who's contributing to the upstream open source code base. I will count newsletter subscriptions. None of these things give you a particularly informed picture. But if you put them all together and sort of look at them in a sort of coherent way, you can get at least an up down arrow or a sideways arrow. And sometimes that's good enough.

Again, I will leave it to future speakers today to talk in more detail about measurement. But you know, in my experience it's really, really hard. Yes? Yeah, well I think Yeah, the question was can I provide some examples of behavior that I don't want to encourage? Well, at Cloudera, just two or three years ago, we created a new discussion forums community to complement the upstream mailing list.

And there was a guy there who was doing his best to, again, answer every question he could but unfortunately his answers weren't very useful or technically accurate or really helpful to anybody. He just apparently, he just felt like having his he was gaming the system. I think the danger of adding gamification to some of these platforms is you encourage people to contribute in unhealthy ways. And that's one example where I would say that, yeah, he was super, super active and his ticker went way, way, way up in terms of answering questions. But he wasn't actually contributing anything.

He was just creating noise. That's one example. What did you do about it? I actually sent him a private message. And I told him that we appreciated his contributions and that we valued his activity in the community.

But we felt like when he is When he does feel the need to answer questions, to just be thoughtful about his answers instead of providing what I would call drive by answers. You know, like incomplete sentences. Stuff like that. I wouldn't call it spam but it certainly wasn't helpful. Did it work?

Sorry, did it work? Yes, it did work. Well, it worked in that we never heard from him again. But honestly, I didn't feel like he was an authentic part of the community. I think he was somebody, again, who was just trying to game the system a little bit.

And if you, again, looked at the numbers on their face it looked good but a little more deeply it wasn't very helpful for anybody. Yes? What was he gaining from the gamification? Just status on leaderboard or swag? Think the question was what was he getting?

I think he was just getting, gaining an online reputation. I think that's a whole another subject. Is the advantages and disadvantages of gamification in communities. Personally, I'm not a huge fan of them. Because I feel like contribution is its own reward.

That's the kind of attitude I think you want to encourage as opposed to hey, you get stuff for being a contributor. That's just my personal philosophy. There's time for one more question if there is any. No? Okay.

Well, we're done. Thank you.