May 25, 2016
The Evolution of Dark Launching on Snap CI
Badri Janakiraman shares the story of the evolution of dark launching on Snap CI
Andrea Echstenkamper: Hi,
everyone. I'm thrilled to be speaking to you all. I'm
Andrea Echstenkamper, Director of
Corporate Marketing at LaunchDarkly, and I'm going to be talking about
product launches, what makes them high
leverage, and how the needs of the
company change over time, in the context
of the LaunchDarkly journey. So, if
you have a tech journalist covering you as a beat, this presentation
is not for you.
I'm going to cover what launches look like when you're starting out, and what's actually working in the early days, which will be the foundation for launches in the feature. Just some context as well, this is through the lens of category creation, as LaunchDarkly was the very first ' LaunchDarkly' to come to market.
Real quick, "What is LaunchDarkly?" you might ask. "It's got 'Launch' in the title. Is this a product pitch?" No, but maybe it's not completely irrelevant. LaunchDarkly helps with feature management, and one use case is a dark launch. Which is introducing a feature and slowly exposing it to a subset of users to get feedback and reduce risk, so we help companies meet the digital demands of their customers by empowering all teams to deliver and control their software. Meaning, that we allow more than just developers to play a role in the software delivery process. But enough of that.
Why am I giving this talk? I've been at LaunchDarkly for almost five years now pioneering this creation of the feature management category, and I've seen all that we've done in marketing since those early days. It's been a pretty amazing ride. We've gone from a team of five when I first started to over 200, basically from no one knowing us at all to well over a thousand customers across truly diverse industries.
So since we're setting the stage, let's ask
the question, "In a world where
software doesn't come on a disk, what
is a launch?" We're all making
smaller iterations and launching them more frequently,
partly due to feature management actually making
this a safe endeavor.
But let's not get caught up on a single point in time that's implied in the word "Launch." It doesn't happen on one specific day. It's not something that is over until you want it to be. It's really just any opportunity you have to create awareness and relay your message, so I'm highlighting these three "Launch levers" as I'm calling them.
They're some of the most fundamental components of your launch. As I walk you through this journey I'll refer to how we've evolved or bolstered each of these over time. Channels, what are the primary assets? Amplification, how do you allow for more pickup? Cadence, how do you stay top of mind? Not only how often do you launch, but how do you keep folks engaged with air cover for your brand? You might argue that messaging is a lever as well, but I say that's one for a whole other talk.
So first, I'm going to take you back to the
early days. Seed round to Series B,
specifically. What did launches
look like back then? We had a tiny
team for more than a year and a half, it was just me in
marketing working closely with our CEO, Edith. What's
important to keep in mind during this time? At
this point, you're totally unknown. Remember
that it takes 7-8 marketing touches before an unknown
becomes a lead. It's all about a
frequent, persistent drumbeat. We
have to try to sustain attention.
You can't simply make enough noise in one day, you have to launch to get the feedback you need to do better next time. There are very few developer tools that benefit from being stealth. Take your learnings and move on. We did our very first Show Hacker News post, and we were a little bit cheeky with the messaging. The response was crickets.
Then given the opportunity to try again, we learned the exact recipe for success here and even proved out a new direction in our messaging.
So here's some of the stuff we did. Regarding
channels, we didn't have much of a following at the time. We
did the simple stuff, like blog posts and a few press releases here
and there, as well as posting to third party
social sites like Hacker News and
Product Hunt. Quick story time on
Product Hunt, we got exactly one customer
that we know of from that launch, Invision. But
they were huge for us at the time and
their co-founder is still an important customer advocate
for us. They even sponsored our annual
event this year. It seemed like
small results at the time when
you're expecting a flood, but it
really did get the ball rolling.
So anyway, back to this. We relied on the amplification of our VCs, our friends, and a few early enthusiasts. Literally, "Please retweet our tweet" at the very beginning. But you'll see at the bottom here that we started developing a strong cadence for other activities that were the support for this, and these were crucial. When someone learned about our recent funding announcement, they'd already been listening to Edith's podcast, produced by Heavybit, for instance. That's what gave them enough of the push to actually check us out.
As far as timeline, here's a shell of a
workback plan on what you'll need
to consider. First of all, where
are you choosing to launch and what
are the big limiting factors to this? To
post on Product Hunt you need someone with the credentials to officially
"Hunt," or "Post" you. Is
it going to be a press release that you shop to the press? What
does that timeline look like? Look
back from your embargo date so that you have a week to shop the
Make sure you submit on the wire with enough review time for the release to be approved. Getting quotes and permissions from customers can take weeks, so don't forget to build that time in. This won't take as long, but go ahead and plan what you're going to do on social in the early days so you're not scrambling. In those early days it can be one of the most important ways to get amplification. We made a literal list of everyone to ping once those first couple of launches went live.
Here's what it looked like in the early days. Funding announcements, Edith speaking at Heavybit events before she started getting asked to talk at a lot of conferences like GlueCon and DevOps West and Microsoft Build, there's our humble little meetup group. Good times. We were just trying to dig out from obscurity any way that we could. Edith wrote and submitted a lot of articles too, and here's one from ReadWrite.
Before we go to the next section, let's stop and talk about press releases for a second. When is it worth it? I say "Do them for funding announcements? Yes." Your story needs that rigor, especially in the early days when you're not sure how you're talking about yourself. But until then, until you're more than one or two people in marketing, don't spend too much time here. Sometimes it's just a blog post but later it can be proof point to enterprise customers, and then after that when you have someone dedicated to this more full time it'll help you build your PR story.
On to the
next stage. We're a little bit
bigger, what's top of mind? Seed
the market. Start thinking about
where you're going as a company and start to get people ready
to be receptive to that narrative. You
have a little bit more foresight now, so make your content marketing
strategy part of your launch strategy.
Write the thought leadership blog
posts and do conference talks that warm up the audience
so your announcements are more well received.
Another guiding principle is to appear bigger than you are, and it was a factor in how we thought about our annual event Trajectory. We were courting and bringing on big name clients and wanted our marketing to really match the reality of our infrastructure for enterprise scale. Yes, you have more resources to launch on and more channels essentially, so doing a press release for something that enterprise really cares about, like an analyst accolade might make sense.
More people are watching you so give them a
reason to talk about you. We
started sharing how many feature flags we serve in a day,
and as I mentioned before, promoting these milestones was a way to remind and re-engage
promoting these milestones was a way to remind and re-engage
Yes, you're on the map. But you're not big enough to be a full blown machine, so don't worry about being too fancy. Don't worry about going live with five pieces of your campaign in one day. Pull the amplification levers one by one. Get the landing page live, then put out the blog, then post to social, then send the email.
You'll honestly reach different people on different days. So things are going to take longer now, plans are going to be more intricate, especially when partners are involved. Like I said, you have more resources. It's not just asking your buddies for help anymore. Your old buddies are still rooting for you, but who are your new buddies? Who are your new partners? And I do mean "Partners" literally. Start leveraging more co-marketing, and here's what that looks like on the next slide.
Our annual event Trajectory became an important new megaphone for us, and here's an integration launched with Atlassian and partner activities with Microsoft. Heidi Waterhouse, our developer advocate, was speaking everywhere all the time. This one is at an O'Reilly conference-- RIP. We also got Forbes 100 Rising Star and our funding announcement on CNBC.
OK, so moving into present day and post-Series
C. We're starting to get big, we're
just over 200 people today and here's what's
guiding us. Launches are super cross-functional and planning
is more of a yearly endeavor driven by product marketing.
How does your annual event line up
with your large product announcements, and how do you handle
that? How do you make sure that smaller and
medium sized launches are running like clockwork? They
can be that constant drumbeat that is not allowing anyone to
take their eyes off of you.
Like I said, the big discussions around these major launches and how to package narratives that support these, but you can start to do things like product webinars or even special product announcement webinars. This is also when there starts to be more analyst assets. And again, this is through the lens of category creation, but analysts became a much more vital part of this process. We were starting to have more company-related announcements, like Best Places to Work.
Here's what it looks like today. Our Test in Production meetup has over 1,400 members. Wall Street Journal cares about our funding, we're actually on the Forbes 100 and Edith is keynoting or speaking at gigantic conferences. So we're coming to the end of the presentation, and it was great reminiscing with all of you, but I just want to leave you with the idea that your launch is going to be more successful if you're thinking about it as part of a larger strategy to consistently create awareness and fine tune your message. That's it, thanks so much for your time. And thanks for hosting me, Heavybit.
Dana Oshiro: Hi, I'm Dana. I work at Heavybit and I feel like Andrea and I have grown up a little bit together, in that her company has grown from something very small to something very big in a very short period of time, and very much to her credit and frankly a ton of effort in those early days. I actually have a question about those early days, because people go back and forth. They're doing their discovery, they're trying to figure out what their persona is. How many personas did you have at that time, or when did you actually decide how you were going to isolate down?
Andrea: Good question. I think for us, we were really thinking about one person for a little bit. Like, this engineering manager, because that's about all you can handle in the early days. We thought a little bit about product managers, and that was a slightly different idea, but really stayed core to that one group. Just because it was really the drumbeat and the best practices in the building of the category that was most important for us, and if we tried to start splintering it too much I think it would have taken away from that overall needing to have a drumbeat.
Dana: You talk about category creation a lot, and I know that there's a big educational piece to it. Can you offer a little bit more context on how far you had to go back? Salesforce had to basically explain cloud. How far back did you have to explain before you actually got to talk about the product?
Andrea: For sure. A really
big part of this was coming up with
naming conventions. People were
talking about feature flagging and not knowing that they
were talking about it, there was feature toggling
and switches and experimentation and all these
different things. What we decided is that we had to take
one and really make sure that we were running with that, and saying
that over and over and over again. So
that if someone were learning about this category, when they would
search for it and try to understand more, they would
be searching for something where they could find us.
That was really an important turning point for us, I think realizing that we needed to rally around one name like that. Then from there, it was really once people understood the idea of feature flags, we did a lot of posts like "What is it? Why is it called this, and not something else? What are best practices?" People were so hungry for best practices, they could never get enough. After we were thinking about that all the time, then we began to evolve more in realizing that it's not just feature flagging as an activity, but really the management of it is where the usefulness came from of having a tool.
Dana: In creating a new category, prior to that everybody was always just basically building their own. I remember you talking a lot about buy vs build, literally breaking down how to build a feature flagging system, which is a ton of work. I thought that was an interesting way of basically just exposing how much work it is to maintain that and to convince people to move over the line to just pay you for it.
Andrea: "You want to do this? Have you thought about this part? Have you thought about this part and this part?" We're here to help you on it, but when you start to realize what a handful it is you might come back and buy the hammer from us instead of making it yourself.
Graham used to talk about Stripe in saying that "Nobody actually wants
to build their own internal payment
system," and that, "It's a schlepp problem
that everybody else hopefully could
outsource, and create an entire category
and company around it." It seems like LaunchDarkly is
in that space as well.
I did have some questions about this early social proof stuff, or the partners and influencers. It's a catch 22 in that nobody wants to lend their name to you when you're obscure and a reputation risk for people, but you need those people to get more people to build more momentum. What does it look like in those first days? Like is it literally asking for retweets, or who did you approach first?
Andrea: I think one really
amazing thing that LaunchDarkly had
was Edith, who's the Kevin Bacon of
the developer tool world. Edith
knows so many great people and so many people respect her,
so I think we were a bit maybe
different than the average company in that
way, of having a lot of people that were ready to go to bat for us.
But I think it's really taking
an interest in maybe other people, and trying to find
ways that you're tying what they're doing back
to you. Just like, going
out there and trying to find the
For us, what that looked like was, say, going to conferences. Niche conferences where people really wanted to be talking about the bleeding edge. Another place for us was looking for people that had built this internally at other companies. I think a lot of early audience champions were people that had this at their company before, and they were like "Yes, I need this again. I don't want to live without a platform like this. I'm happy to help spread the word."
Dana: Sometimes you see this with other types of devtools as well, there's this creative-- It's like you point to the Netflixes and the Google-scale things of the world and you're like, "We're going to offer this scale of something to a normal company that doesn't have tens of thousands of developers that can be resourced on this." How important was it to use, I think you used Facebook as a comparison, maybe?
Andrea: Edith started with this message before I even started, and I think it was really powerful, to say "You can have this too." It was a message, I think, that always as a marketer you're like "We've talked about this too much. People get it, right?" You don't realize that when you have a message like that, you just need to keep telling it again and again and again on different mediums. That's what we did. I think Edith's blog posts were like, "The secrets to Facebook engineering." And then it's like, "The secrets to Google's engineering culture." It's like, again and again and different ways to really drive home that message.
Dana: Did you have an actual brainstorm list of topics that you just kept using over and over again? How did it get recycled in all these different formats?
Andrea: I think we have, just like in the top of our head, "These are the three messages that are really working for us right now," and really just having the whole company come together and have these feedback moments once a week. Where it's like, "What are we hearing out there? What's working now?" Because you're getting a lot of feedback, which is great, but what is the feedback that says, "This is the way to go" or "This is not the way to go?" Part of it is that intuition of "This one just really feels like it's hitting. Let's keep trying it." There's no secret magic bullet on that.
Dana: Does "Hitting" mean, like, retweets and top of funnel traffic? How well measured is that?
Andrea: I would perform
simple experiments sometimes. It's
easy to throw up a Twitter ad post
and see which one is getting more
love. Or like I said, with the show
Hacker News post, when you actually
see how people are responding to it
and realizing that they actually get the
value from it, that's when you know
that's a way that you're going.
I think that happened for us with understanding that management was a really important part of our value. When we realized it's for teams, that the teams are getting the most value out of it, then we started pushing the messaging towards that way, but there was the talk that I did also with Andy Raskin for Heavybit before that. That goes into how to find that message and how that exercise brings you to that result. Which I think is what we were thinking about all along and trying to make that part of how we were determining our messaging.
Dana: Makes sense. There's a question from the audience about early success metrics, so before you have product-market fit and before you know you're a thing, what are the KPIs and what were you measuring to show results?
Andrea: I was measuring a lot of things, but really just keeping track of them to see where there were spikes or really just understanding that. But really, the North Star for me at the beginning was the number of trials that we were getting coming in. It didn't matter if we got a huge burst of website traffic, if we were not converting those people they weren't the right people. Who cares? So really, were we getting people to commit to trying us for 30 days? And that's how they would really understand how powerful it was. It was really the only thing that really mattered to me at the beginning.
Dana: What do you think are the indications or the core levers for trial conversion?
Andrea: It's tough, because I feel like we would try to get that measurement for when we actually got a sale. We would make sure that first Edith was the salesperson and would get that feedback of, "Where did you hear about us?" Honestly, a lot of it was Hacker News podcasts. Those were two of the ones that I remember hearing the most at the very beginning. Those were two of the biggest.
Dana: They seem to match your early persona. Even if the top of funnel wasn't necessarily as high as what a Forbes would be or whatever, it does seem like it's highly targeted to the right type of audience for you. As it starts to get bigger and as you start to build things out, what do you feel within all of this is OK to outsource? Versus, what do you want to keep in-house for your product launch team?
Andrea: Good question. That's tough. I'm trying to think of what we even outsourced, because in the early days-- Certainly at some point you would want to bring on a lightweight PR agency. We did do that before we started to have someone dedicated to public relations and analyst relations, so that did become a thing, but more around just before Series C. There's no money to outsource--
Dana: Yeah. Do you do anything with copywriters, or even in the tactical execution of Trajectory conf. Is everybody internal?
Andrea: For Trajectory, we do have an events agency for that. We're also starting to bring on more copywriting help from an agency standpoint. In the very early days, I think even just having someone who is a junior marketing resource that may not be someone you have on your team as a full time employee, but just someone that's like "Can you write some things here and there or do a little bit of web development?" Or whatever you need, sorting something or looking something up, just that kind of person that can help you out. If you're only one marketer, you're going to need some more help there.
Dana: In order to onboard new marketers who are actually effective, you probably need to standardize something about the voice and something about the core messages. How much work did you have to do up front in order to bring these other outsiders in?
Andrea: I think documenting,
which we discussed in that talk with Andy
Raskin, was something we did do
really right around the time that we started bringing on other
marketers, and then really slicing
and dicing that. Like, "This is
just the description that we're using over
and over again. This is maybe a tagline we're using
from time to time. This is the
longer description," as simple as
that. Making sure there was a real
rally behind it so that if you do
have multiple people out there that you're
all saying the same thing.
When it came to the first hires we needed in marketing, one was for us with category creation. Another really big one was doing events, so getting in front of doing sponsored events and getting in front of those audiences. It was really important to us, also so we could have a real conversation with a person face to face and explain what we're doing, because that's why content marketing worked really well. Also because we really needed to tell them what this was all about and why it existed, rather than taking for granted like, "I should switch you out from this other tool that I already have."
Dana: I feel like you and Edith together made such a good team, because there was already innately a certain amount of trust there and then you could be like, "I know this event or this effort is going to be valuable. Please go and do it." And she's pretty prolific in it. Obviously Edith is one of the spokespeople in your organization, but who else was producing your blog posts? Not all content can be created by the CEO, especially not at this stage now.
Andrea: Totally. We were all doing it. John, our CTO and the other co-founder, was writing really great and deeply technical blog posts. I think one was for AWS, and then later it was for StackShare. How did we create this novel architecture that we have? That was of great interest to everyone, and those did extremely well for us. We even did what we called the Great Wall of Content, everyone in the whole company would stop and just produce content for this month. We had a backend engineer writing about database migrations, our designer wrote things here and there. Marketing should not just be marketing's job. That's a stage of your company that you're not going to be able to move forward until you get yourself in front of people.
Dana: That's interesting, because there's so many teams that make the engineers go and do support. There's bug fix days, there is hack days, or whatever. But there's rarely "Do some marketing or write the docs" kind of days, but they're just as important and part of the product as anything else. In the last question you talked about migration from an existing incumbent versus why you would need this tool, especially a new thing, to begin with. Because you didn't have this "We're cheaper than AWS at this," or whatever.
Can you talk a little bit more about that?
Andrea: Edith gives some
really good talks about this idea
of publishing the blueprint as
well. One thing we did was we
created a site called FeatureFlags.io, which
helped with our SEO with direct support in that,
it gave people all of the open source
libraries that they could access to start using
feature flagging. Because really,
once you start using feature flagging and understanding the benefit
there, that's when you realize you
need to treat it as a first class citizen. Feature
flagging on its own in a real way
can mess you up, and you hear horror
stories of a couple of companies that did that ,where they
thought they were turning something on and they were turning something
off. It can be super powerful, but
it really needs to be in control.
So that's just a conversation that we've had over and over again with users. Another way I think we talked about it was, "Here are the levels of feature flagging or feature management that you might be doing. So, you might be doing something at this level right now but this is how you get to this next level, or this is what this next level looks like." So people can understand, it's not just, "I'm doing feature flagging, I'm done." But there's a whole other world of use cases out there or sophistication that you could be bringing if you're the kind of company that's taking it seriously.
Dana: If you don't really know what you're doing with feature flagging, I feel like there is a lot of people that were just talking about A/B testing features or something like that in the early days. How important was it to offer a ton of early examples of use cases of how you could set this up? It's like, if you've got Google Analytics but you don't have any goals, who cares?
Andrea: I think that was one
of the most interesting things about working on
feature flagging, but also when I think of the hardest it's like
"Which thing are you talking about in feature
flagging here?" There's so many use
cases, and I think Edith gives some presentations
about "Here are ten different ways to use feature flagging." But
when it came down to it, once we got people familiar
with the concept and hear all the different use cases,
then we have that opportunity to start talking
about the benefits.
Really, just being that little bit more benefits driven is where we really started to get a lot of traction. Once people just knew what the hell we were talking about back then. It was like, "Turn something off when it goes wrong. How important is that?" Or the fact that you didn't need to show everyone this first. Like, these are really core ideas that have nothing to do with testing that are going to be really important for your company. Also, getting qualitative feedback is so important as well. Going ahead and releasing to certain folks, hearing what they think about it, or the quantitative feedback of whether or not your system broke.
Dana: Got it. I did have a question about when you moved from meetups and other people's third party things to Trajectory, like my assumption is you have to have a certain amount of product-market fit or a large enough community. This is not a small amount of effort to put on your own thing. When did you know you could do this and that people would show up?
Andrea: It was a little
scary, to be honest. I feel like I
have this in my personal life too, where I'm like "I know people
like me, but is anyone going to come to my
party?" I think what we really
needed, what felt good to us was
having the drumbeat and the success
of our Test in Production meetup before
really embarking on Trajectory. It
was so foundational and important for us to have done that, because
not only did I remember specific regulars that
came to Test in Production, showing up at Trajectory and being like
"This is great."
It wasn't even only just that, but it was also the fact that we had found a lot of these speakers in advance, and were like "We know they're great. We want to ask them to submit a CFP and maybe come back." It really just helped us understand who would be included in Trajectory, because really for us it's not just talking about feature management. It's talking about the future of software delivery and "What are those things that are part of or related to feature management that are going to be this feature that we're all in together?"
Dana: How has Trajectory changed? Did you start with really like Dev-y, or like release manager-y talks? My assumption is you're moving over to enterprise a lot more.
Andrea: One interesting
piece of feedback we've gotten is that people want
to hear more about LaunchDarkly. We
didn't really make it a user conference,
but we heard that feedback which we
weren't really expecting, but that's
one thing that's changed. We're
like, "Maybe we should have more customer speakers
or more folks from LaunchDarkly as part
of this." We're even considering, "What does it look like
now that we're in a pandemic and it's online?" And, "How do we
maybe break up that content and do it in a slightly different way
to fulfill that need?"
But as far as slightly Dev-y or not, I think what we found from the last trajectory that we did is that people were really excited about these more high-level cultural talks, Even at the more engineering manager level. so I asked some of my engineers and engineering managers, "I'd love for you to reach out to your network and invite people. I'd also just love to hear which talks you're most excited about." It wasn't necessarily the more tactical talks, as I thought it would be, but a lot of times it was these higher level things that people found cool and personally interesting. It hasn't been as much of a focus yet on needing to make sure it's one or the other.
Dana: What cascades out of that? Because you get all this excitement in a couple of days from a bunch of people. We already know your philosophy to run down silo to your product launches, and this is a weird other big bang event. How does that cascade back into everything else that you're doing?
Andrea: It's amazing how it
really starts to become a bigger part
of our partner strategy. Who is
involved with trajectory, and how
do we really make that a great
experience for them? It's become
really important with our analysts, where
we're saying, "We'd love for you to review these talks and tell us what
you think about it, and therefore
how we're moving forward and creating relationships
On the press side, we have had some really high profile groups that have been interested in specific talks and there's been some press on that. So it's really, now as we're starting to talk about next year, it's like "How does this all play into our product launch strategy? And therefore, again, back up to like the content marketing of how that's all going to work together?"
Dana: This is a question I feel like everybody's going to ask everybody this entire year, but how has primarily work from home affected both the product as well as the way that you work with your team?
Andrea: It's been a crazy
year for us. What we're seeing from
a product standpoint is that people
are actually accelerating their digital transformations, like this
is the thing that they're like, "I
was meaning to do that. I'm going
to get on it right now." So besides
the initial "What's going on in the world?"
I think things are really picking back
up in that way. It also works out that
LaunchDarkly helps people stay in
touch on what's happening.
The most important things within the product,
like having an audit log to see who
turned something on and off just now. If
you're not in the same place together, which maybe
even smaller teams that was more of a possibility, like
you're three people and you can keep asking
each other "Did you just do something?" Now it's like you
don't have that luxury.
team, I would speak the most from the
marketing side. It's that we're in
this crazy growth time right now, we're going
from small company to big company.
So it's been an interesting time to
not have the chance to be together, but
we're a close knit group and I think just the respect
for each other has been a really important foundation
while just everything around us is changing.
The different groups, different teams within the company, we're trying to all work together even more. It's just a bigger machine that we're trying to spin up. So I think it's a tough time to have a pandemic, but at the same time it's good for us that we have to put the rigor of making sure to document things or have best practices that we're all following, because that's what we're going to need as we grow to this bigger company size.
Dana: If you were starting with a company at five people now, would you use the same tactics that you used in the past? Given that everyone is stuck inside, because you were at every meetup. What would you do now if you were--? How would you come out of obscurity from your house?
Andrea: That's tough.
I think you just have to find a way to
be really extroverted on digital
things. Things are still happening,
like conferences, I see people still being really engaged
in the chat. Talking
to each other, I think it's actually a really good time
to be asking for coffee dates with people
because it's a simpler ask. We're
not trying to figure out which
coffee shop next to Caltrain we're
meeting at, and expecting you to take your time out of the day.
Just jump on Zoom with me for ten minutes and
share things with me, so I think there is
I know that getting digital things out there is probably the most important thing, instead of doing a meetup, it's a Twitch stream. Or if you can't give somebody a pitch any other way, maybe it's doing more videos. It's dependent on what your specialties are and what you're good at but I think you're still going to have to do it. You're still going to have to get out there and make noise and make friends.
Dana: You talk a little bit about the value of analysts for category creation with your team, can you elaborate a little bit more? What is the role that analysts play that the press, for instance, doesn't?
Andrea: When you're going after the enterprise, they're the ones with the direct line. I think they can be great advocates for you in that way, just spreading the word of "This is important" and putting it on the map. When you do get that first report about you with your name on it, you hope it's through the lens that you want the world to see your category. I think that part's really important as well, and also just they give really great feedback on what you might not be thinking of or lines that you need to be drawing to what other things that the group that you're targeting is thinking about right now.
Dana: I know there's definitely some industry analysts who almost act-- They're not quite resellers, but they're definitely making a recommendation to their major clients around what they should be using in that space. How conscious were you of resellers, and those connecting nodes to a larger deal flow?
Andrea: One thing I should say first, is just that I think a lot of people wait to talk to analysts too late. Don't forget to make that a priority, to brief people. I think that's how we were able to get the Gartner cool vendor, by getting in touch and in front of those folks earlier. But your question about that, I think you just try to find that the champion that really understands and gets you, and spend enough time with them that they're going to be able to make you top of mind. Because I think really what it comes down to is, they're getting briefed constantly. They're getting exposed to so many cool, interesting new things. Like, if you're not top of mind you're never going to get them to be the person that's out there telling more folks about you.
Dana: What happens when you're not in the right quadrant? You know what I mean? "Why am I here, and Datadog is there?"
Andrea: It hasn't happened to us yet.
Dana: You're your quadrant, I guess.
Andrea: Yeah, but back to
what we were talking about a little
bit before is, for us people are like "Is
that like A/B testing?" And we're kind of like, "Let's
start from the beginning." For us, it worked to
be creating our own narrative and
just being the loudest, strongest, best at it. Being
the leader in it and the thought
leader in the space means that you
get to put that in front of the
world more often, and so just really
spending time on that. That doesn't
mean that you're going to get your way, it's
really the wisdom of analysts that are going to be deciding
But just make that a really important part of your company to build that story and be supporting it and continuing to reach and show what's next there. Even though you're trying to sell what's here now, like what's that vision and what's coming next with it? For us, we're talking about that next thing that comes after continuous delivery, we're talking about progressive delivery as a movement, and then really just making sure that we're talking about it in this way and really excited and comfortable and giving folks what they need to really buy into what you're talking about.
Dana: I think Erica Brescia, she's the Bitnami co-founder and she's the COO of GitHub now, and I remember her saying, "I feel like every dev tool starts as a single tool, like a single wedge. Then there's a larger platform player, what the next part of that wedge is." In some ways, you got to be careful that you don't end up pigeonholed in this one single category that's smaller than what your larger vision is. When did you start with your paid relationship with analysts? Because I know, Gartner Cool Vendor, you don't have to pay for usually. But then when you're in the quadrant, there's generally a minimum investment that you need to make in order to get on these analysts' radar.
Andrea: Right. I want to say that was around Series B that we started paying.
Dana: OK, so around just a little bit after you started putting a PR firm on retainer?
Andrea: I think that was about the same time, it was after we'd moved to Oakland. I think we probably started about the same time, the PR agency was cheaper so we could have started that a little earlier, just not all of them. There's the more lightweight agency that's just pitching for you, versus one that can be two or three times as much.
Dana: When it comes to PR, when it comes to analysts, honestly, when it comes to marketing in this space and devtool or infrastructure or very highly technical product space, it's very hard to find the right mix of people. Somebody who is tactically proficient on the marketing side, but also will pick up, the category of what's actually happening in the ecosystem and how important it is. Is it to understand what channels are likely to hit with a particular persona, versus pulling someone out of Instagram, or something like that.
Andrea: Got it. And this is in context to the first marketing hire?
Dana: You see people who have really great marketing careers, but in a completely different industry or even at a different stage of company. How important is technical understanding of that particular ecosystem in that type of sale?
Andrea: I think for
your first hire, that's going to be tough if
they're coming from another industry. I
came myself from the publishing
industry into technology, and even
though a lot of the thinking was
the same, like just who were the players and where
do you even start with that? I
mean, unless you just download all of the Heavybit library into your
brain, maybe you'd already start to
be in a pretty good place.
But I think that it really depends on who else is on the team to support that and have that knowledge, so if you're bringing someone on to be a designer that really maybe doesn't have specific experience in devtools, you can coach them through that. Or maybe it is ARPR, that they're not the only ones crafting that knowledge or that narrative. But really just helping doing the pitch themselves. But it's all about whether or not you have somebody there to support them on that, versus not.
Dana: I think, thankfully, now there are more vision market people who have dev and infrastructure experience or B2B experience. But back then, there's a lot of educational work up front.
Andrea: There is. But I think the B2B matters sometimes, and sometimes if somebody gets that mentality a little bit, whether or not it was specifically in technology, they might really be able to get up to speed more quickly.
Dana: I think it's funny that devtools are so uncool that they're kind of cool now. This is a painkiller, it's not a vitamin for you. It's like, "There's an immediate problem we need solved now."
Andrea: I think that's what's fun about it as a marketer, this is not sugar water in any sense at all. People really want to learn, and how cool is it that it's your job not to just market but to teach people? So, that's what brought me into tech marketing.
Dana: Thank you so much.