In episode 15 of the Startup Field Guide podcast, Sandhya Hegde chats with LaunchDarkly Co-Founder and CEO Edith Harbaugh. LaunchDarkly is a pioneer in the feature management space. Founded in 2014, while the agile and cloud movements were still relatively nascent, LaunchDarkly was ahead of its time. They have now scaled to a $3 billion company, serving over 4,000 customers.
‍

Be sure to check out more Startup Field Guide Podcast episodes on Spotify, Apple, and Youtube. Hosted by Unusual Ventures General Partner Sandhya Hegde (former EVP of Growth & Marketing at Amplitude), the SFG podcast uncovers the path to product-market fit for early-stage founders and product builders.
‍

‍If you are interested in learning more about some of the themes and ideas in this episode, please check out the Unusual Ventures Field Guides on segmenting early markets, customer validation, and working with innovators to find product-market fit.

‍

TL;DR

• Edith Harbaugh's a-ha! moment came from realizing that big software companies had much more elaborate platforms to release features. She and her co-founder, John Kodumal, wanted to create a feature flagging solution that would allow developers at smaller companies to roll out their code slowly, do dark launches, and control the pace at which new features are released.
• It took a few years for the market to catch up with LaunchDarkly’s value hypothesis. Launch Darkly was turned down by Y Combinator and their seed pitch was rejected by 30 VCs. When they launched on Product Hunt, they expected hundreds of sign-ups but only got one qualified lead. 
• Edith and John found more traction when they started selling to developers outside Silicon Valley who wanted an external product to manage their rollout process.
• The best advice Edith received was to “create a movement and then convince people to use the platform.” She did a lot of “brute-force” marketing by talking about feature flagging at conferences, podcasts, and other publications.

‍

‍

Episode transcript 

‍

Sandhya Hegde (00:53)

So going back all the way to 2014, when you were leading the product team at TripIt, how was the idea for LaunchDarkly idea born?
‍

Edith Harbaugh 

Yeah, so my own background is interesting because I actually started in engineering. I was an engineering manager, and I got frustrated building stuff that nobody wanted, so I became a product manager. And I actually started TripIt for Business, TripIt’s Small Business Division. And when I was at TripIt, I also started doing marketing and also had a small sales team that reported to me.
‍

So I'd really done every role within the company except for finance. And I just kept seeing the same pattern over and over where it was difficult to ship features, it was difficult to know which ones really worked. At TripIt, we had a homegrown platform that helped us do this, for example, I could roll stuff out to 10 beta users at TripIt and then do a percent rollout. I could block people, if they were in the press, we didn't want them to see an early feature, or I could enable it. And I knew that the big players like Facebook or LinkedIn had much more elaborate platforms than TripIt could afford, and it was just like, "Why is this just not a product? Why doesn't this exist as a separate product? Why do we always have to suffer and rebuild the same thing from scratch?"
‍

Sandhya Hegde (02:22)

And how did you go from that one insight to, "Okay, is this worth starting a company around, who would I start it with, finding co-founders, picking a name for the company?"
‍

Edith Harbaugh 

Well, that's interesting because my college friend, John Kodumal, and I had been talking about doing a company for years, and we actually were working on another separate idea at the time. And when I was going out and trying to do market research for the other idea is when people kept saying, Hey, this other idea you have which was a company called Continous Delete, is interesting but my real problem is I don't have the platform I had at LinkedIn, I don't have the platform I had at AWS to release features.
‍

So this was just an "a-ha moment" for me. I literally drew, not on a napkin because I carry around a notebook, but I drew on a notebook like, what if you had a platform where you could do slow rollouts, do dark launches, do betas, and then do experimentation, and then also do long-term entitlement controls? Because TripIt had a premium product just like we'd done there.
‍

And I went back and showed it to John and he's like, "Wow, I would definitely use something like this." John was at the time in engineering at Atlassian, he's like, "We have a platform like that at Atlassian, but there's nothing on the market." So it was just this total "a-ha moment" for both of us that this was a pain that we had both felt. There was nothing on the market, and this was something that we wanted to commit to.
‍

Sandhya Hegde (03:59)

And how did you go about thinking about the fact that like, what's the, "Why now?" A very popular question we as investors ask founders is, yes, this is a problem. Clearly, it has been around for a long time. Why has no one else solved it? Why would you be the team to solve it “right now?” What's the “Why now?” Was there a clear, “Why now?” for LaunchDarkly?
‍

Edith Harbaugh 

Oh, we got absolutely hammered in fundraising. YC turned us down. I think I did 30 seed pitches and got multiple nos. I have a spreadsheet somewhere of everybody who said no, like nobody got it. Like nobody.
‍

And the funny thing about being in the Valley a long time, because we started in 2014, is now I'm still friends with the same seed investors, and they're like, "Well, Edith, you were right. I just didn't see it." I'm like, "Well, thanks." And they're like, "You were right. I didn't see it, and shame on me."
‍

So the "Why now?" to answer directly, is we were super early. I think we were perhaps two or three years too early. It seemed extremely obvious to my co-founder and I that this was a good idea. What we didn't realize was that there were two things that had to get fixed before it could be successful—that had nothing to do with us. It was, one, we were both in Silicon Valley. We worked about five blocks apart in the Mission. I was at TripIt, he was at Atlassian. And to us, it was very obvious that everybody was doing agile sprints and wanting to release at least once a month if not once a week. For the rest of the universe that was not in the Bay Area, this was still an extremely new idea. People were still preaching Waterfall. Gartner actually had come out even after we started with the bimodal approach of, "Let's do yearly releases for important things."
‍

The second thing that was a huge headwind that we didn't realize was people were still moving to the Cloud. Again, this is in 2014, I'd been at TripIt, which was the number one mobile travel app. And I thought, "Of course, everything is in the Cloud." For a lot of customers, it was still this brand new idea that, hey, you don't have to have on-prem software. You could move to the Cloud and have a lot more agility and release more frequently. So if your question is, "Why us? Why now?" —  in hindsight, we were way early, and we just had dogged determination that this was a better way to make software, but it took quite a while for the market to catch up with us.
‍

Sandhya Hegde (06:34)

And I always have felt that it's better to be just a little bit too early than be extremely late, especially because in the early years, you don't need the entire market to see your vision. You only need a bunch of early adopters to see your vision and embrace it. So, as long as eventually the market comes around, you're good. 

So maybe that's a good segue to the next question, which is around this idea of the product-market fit process. And we typically divide that into two big stages. So first, just testing how strong the need for the product is, like is the value hypothesis very sound? And there I think you have a big edge because both you and John were the customers for this problem. You've already felt the pain, so you have an incredibly authentic source of conviction for it, but I would love to understand how you kind of pressure-tested that. Who did you talk to be able to say, "Yes, there are enough people who want this today in the Valley?" And then second, who were the early users? Were they small startups, bigger companies that wanted to keep moving fast? Who was the early desperate user that you ended up getting off the ground with?
‍

Edith Harbaugh

I'll say it was extremely, extremely slow, uphill sledding. It took us about a year to get our first 10 customers, and it was so much work. So I always read about the people that just put something on Hacker News and just immediately get 3,000 customers. That was not our story.
‍

Our story was we went around basically to all our old coworkers. So I'd been at TripIt, and before that I'd been at a company called Epicentric, and just asked them to use the product. And when we asked them to use the product, they would say, "This is interesting, maybe I'll try it," and then we would show up at their office and ask them again and say, "We'll come and watch you install it." Even then, it took us a while to get traction. So at the beginning, we were pitching it more as that it would help you experiment, that it would help you see which features are more successful, that it would help you measure release impact, which by the way is the pitch that we use now in 2023.
‍

What we found over and over is when we went in and talked to prospects, like I remember my friend Andrea had gone to be one of the co-founders at UrbanSitter. She's like, "Yeah, yeah, this experimentation thing is great, but we're just struggling with more basic release process things, so come back next quarter once we have the release process things sorted out." And we went back every quarter for a while and I finally was like, "John, I love the experimentation features, but I really think we need to pivot and focus more on developers."

And it was really easy for us both to make this realization because that was suddenly when we started clicking with our customers. And when I say customers, I mean any at all. And then they're like, "Yes, this is a problem that I have right now that you can help me address." So our first real, real customer that wasn't just an old coworker was somebody who came in off our landing page who used to work at LinkedIn. At the time, we just had a landing page and a “Contact Us” box, and if somebody filled it out, I, Edith, would contact them. And he said, "I want the same platform I had at LinkedIn. I want to do slow rollouts, I want to do beta testing, I want to do some validation, and I just moved to a new job that's not LinkedIn, and my internal team wants to build this and I would much rather have an external product." And it was just like, we have exactly that. It was just like this ping of what you want is what we have, and it was just this amazing feeling.

Sandhya Hegde (10:30)

How long did it take you to have something that was a stable product in production with people using it every day? Was that 12 months, 18 months, like from the time you actually started the company to saying, "Yes, we have a couple people using us in production and happy with us." How long was it?

Edith Harbaugh

So we were both, as I said, both product engineering folks. We officially started in July 2014. Before that, we'd been noodling on ideas, but we took an official “this is our first day picture.” And John from the very beginning had sprints where he would build stuff and we would validate ideas. Our first actual person using it, and I have a picture somewhere, because I took it that day, I think it was August or September, it was a friend of mine who went to a company called WidgetBox and he said, "Hey, Edith, I'll try this," basically as a favor to me. And he started using it, and that was really fun. It was a much simpler product back then. It just basically let you turn stuff on and off and it had a slider for rollouts, but we had something.

Sandhya Hegde (11:38)

I think that that velocity makes a huge difference because I think nowadays it's also maybe the bar is a little bit higher in terms of the amount of surface area you have to build if you are in a competitive market, but even then, I think a lot of startups are not able to reach that point even after 12 months, so the feedback loop is much slower when that happens. So I think the fact that in three months you had something people could try was definitely really, really valuable. What was the profile of the first 10, 20 customers? Were they all smaller startups? Were they bigger companies? Was there a pattern that you could point to? I know it was friends and family, but still, was there a pattern you could point to and say, "Okay, this is who we should focus on in the early days?"

Edith Harbaugh 

So I told you before about my buddy Grant using the product. I was very careful to say use and not pay for. So in the early days I was very careful to not charge because I didn't want to lock us into a price. When we did finally have to start charging because we'd raised our seed, I wanted to start putting some revenue on the books, it turned out that they actually had a home-grown system that they were much more happy with and they didn't want to pay anything for ours. And I'm not above making a rock-bottom offer if it's like, "Hey, just pay us $5 a month so I could put you on the books." And he's like, "No, Edith, honestly, I just did this as a favor to you. We like our home-grown much more."

That was pretty brutal to hear, but it was also true like their home-grown was better. So then we basically started trying to get our own system to parity. And what we discovered was that Silicon Valley companies were actually not very good customers to us because they're much more propense to just build something. Our best customers were external to Silicon Valley and they were folks who said, "Hey, I can't build this in-house, or I don't want to build this in-house, and I am happy to pay you money for this."

Sandhya Hegde (13:51):

It makes sense because you have much more engineering talent available in the Valley, and at the time, they weren't quite as expensive as they are right now either, so you have the ability to actually have a little spare talent on the books, you have hackathons where they can build some internal systems. That makes a lot of sense. How did you reach the external customers outside of the Valley bubble? How did you decide where to start and pick those early customers?

Edith Harbaugh 

A lot of, honestly, brute-force marketing. So now it seems really obvious that feature management should be a category. At the time, it was brand new. I made up the word feature management to describe what we were doing. So, a lot of blogging, a lot of blogging about why feature flagging was good, why feature toggling was good, why you would want to use feature flagging, and just constant repetition of the words so that we would do well in SEO.
‍

There was a popular article back then about feature flagging as the worst kind of technical debt, and I wrote a counter to it, "Feature Flagging is Technical Debt if You Don't Have a System Like LaunchDarkly," that InfoQ published because it was vendor-agnostic so they would publish it.
‍

The blogging helped a lot just by giving us this long-tail reach of if you're interested in this field, you'll come upon LaunchDarkly. I also started giving a lot of talks. So the best advice I got was from this guy named Javier who had just sold a company to Microsoft and he said, "Edith, you're not building a tool, you're creating a movement. You need to create a movement and then convince people to use your platform instead of the other way around." And I was like, "Okay."
‍

So I put together a talk about mistakes ... at the beginning it was just like why you should do feature flagging at all, like top 10 use cases for feature flagging, which I've said many times already, but like beta testing, dark launches, control, experimentation. And I would just give that talk anywhere that would have me. I remember going to a 20-person conference room at DevOps West Las Vegas and giving that talk. I once gave it to two people at an accelerator in SF, at which point I just sat down and chatted with them.
‍

And the other thing we did was I started a podcast with my friend Paul from CircleCI, and at the time, podcasting was still pretty new. This was 2015. And it was just us talking about software. We called it “To Be Continuous”. And what I did not realize at the time was there's just not much technical podcasts out back then, so it actually got a pretty strong following.
‍

And then things that didn't work for us in the early days were Google AdWords. There was no market for it, so nobody was really searching much for the words enough. Display ads didn't really work back then. There was no silver bullet. It was just a ton of brute-force. Oh, oh, oh! And then just we put a lot of effort into Product Hunt launch, and it's like all such things when you're a product manager, we had such high hopes. We put together a discount code, and we were expecting to get maybe 400 sign-ups out of it. We got one, which was pretty devastating at the time.
‍

It was a company called InVision, and they were like, "Hey, our CTO has built this system in the past and he doesn't want to rebuild it. Can we use LaunchDarkly?" And so I think they were officially customer number six or seven. And we just lavished care on them, and they turned out to be an amazing early customer. They're still a customer, by the way. And that guy, Ben, who was the CTO, still writes blog posts about how much he loves LaunchDarkly and how it changed his life.
‍

And the other thing is, I've told this story at conferences before about how we only got one customer, and then developers have come up after and said, "I saw your Product Hunt launch. I liked it. It just took me a long time to convince my internal organization to do this." Like, it's a slow burn. We realized, sure, if somebody used LaunchDarkly, like it can bounce around internally for six months, two years, or longer before people decide to get on board.

Sandhya Hegde (18:35):

So if you go back, so when was the first talk you gave and how long did you keep doing this? What did that window look like when you were just on the road trying to build the movement, as Javier said?

Edith Harbaugh

Oh, gosh. So I was a really shy engineer. I hated giving talks. I remember the first time we had our own meetup, I hid in the bathroom because I didn't want to talk to anybody. And the best advice I got was from one of my friends. She's like, "Edith, two things. They're all here to hear you talk. Nobody's going to make fun of you." I was like, "Okay, that's true. This is my meetup. I should have a little bit more self-confidence."
‍

And the second thing she said, and this is really funny, she's one of my best friends  — I used to sing karaoke a lot pre-COVID, and she's like, "Edith, you are one of the worst singers I know." And I'm like, "Yeah, that's true." "But when you get up on the stage, nobody looks away and everybody claps at the end. So just bottle that energy better." And she's like, "You're actually a really good performer. You just need to translate that into your talks and have more confidence." So it's like, "Okay, all right. I could do this." If you've sung bad karaoke and people still clap, you could certainly give a technical talk about something you care about.
‍

So, I put together a basic talk, and then I made a spreadsheet ... I love spreadsheets, by the way ... of every conference I thought might be interested in this talk like Continuous Lifecycle London, DevOps West, GlueCon, NDC, and I spent an afternoon just submitting the abstract to all the talks. That made it much easier because you have a canned abstract that you tweak for each one because they're like, "We want 200 words or 700 words."
‍

And that way it's also much easier to field rejection because I would look at my hit rate, I submitted it to 20 conferences and got admitted at two, Continuous Lifecycle.London was one I remember. Great, I'm going to go to London.
‍

Then you start to build up a track record because conferences want to see a video of your prior talks. So when you're starting out, you don't have that, but you can start to build that muscle and be like, "Great," and then you can ladder into bigger and bigger and bigger venues. So now, for example, I spoke at Web Summit on their main stage in front of I think like 40,000 people, which eight years ago I absolutely would've been ill-equipped to do.

Sandhya Hegde (21:06)

Given that your early customers were now in a, less kind of the fast-moving Silicon Valley startup profile and more outside of it, were they also bigger teams? What was the shape and size of the customer you were going for, and what was the implication for you as a young startup? Because you still have to prove product maturity and reliability. I'm sure these are concerns that most dev tools that have customer-facing infrastructure-as-a-service have to face. I'm curious what that process was like. How did you build trust and actually get in with these less fast-moving companies?

Edith Harbaugh 

So LaunchDarkly helps a team manage a release, and there's this myth or trope that you can go extremely bottoms-up and a developer could start using something. What we discovered was that we could not go that route because we're part of the release process. It would eventually hit this brick wall of an architect, a release manager, somebody would say, "What the heck is this thing sitting in my code, and can I trust it?"
‍

So we discovered that we really had to talk to at least a manager or a director who could get something deployed. So, this was a little bit of a chicken and an egg problem for us for a little bit, in that we had to build up that muscle of having enough reliability and security that somebody would use us. But it also became very much a competitive advantage because once we had a marquee customer like InVision, that people knew, we could say, "Well, InVision uses us. They trust us. You could trust LaunchDarkly."

Sandhya Hegde (23:25)

Any surprises for you in the early customer feedback? I know you already went through one phase where you said, "Okay, the experimentation thing is a little too early, let's not focus on that," but I'm curious if you saw more surprises that led to more effective positioning, messaging, that was rooted in early customer feedback.

Edith Harbaugh 

I think what made me really happy was how much customers loved it. We had a really early customer called LinenTech. We did have some Silicon Valley customers. I don't want to say that we didn't have any. So LinenTech was a Silicon Valley customer in downtown SF. And John, my co-founder, went on site and he said when he was there they introduced him to everybody. Everybody said “thank you.” So we just started getting feedback from our customers that they really loved it and it really changed our lives for the better. That really felt good, and it kind of gave us that early fuel that, though we weren't this runaway immediate success, we had this core that we could continue to build on.

Sandhya Hegde (24:41)

At what time did the market feel more ready? At what point did you start seeing inbound, start seeing your bottom-up motion work? Because at some point, I know you launched a self-serve trial experience, right? Could you help us understand like what was the timeline of your evolution in terms of go-to-market from when it was brute-force founder selling to like, "Oh, okay, we have a machine that's kind of starting to work for itself"?

Edith Harbaugh

Our original assumption, like I said, my co-founder was out of Atlassian. I had been at TripIt. Our original assumption was that this was going to self-serve. Like I had all the spreadsheets ... Again, I love spreadsheets ... about we're going to have this many people land on the page, they're going to come in at this rate, this is our free trial conversion. And it turned out that our early customers did not want to buy through that funnel.
‍

I talked before about an early customer, and it came time for them to buy and I'm like, "Great, just go put down your credit card." And they said, "No, this has to go through procurement." And at the time, we didn't have salespeople. We had me, and I guess I was a salesperson, but like I didn't have a dedicated AE or SE to hand it off to. And so there's a couple back-and-forths of like, "Well, go put your credit card in," and them saying like, "Well, procurement needs to work on this."
‍

And then I literally remember, because I was driving up from a race I'd done in SoCal, so I had like five hours of driving time, and I was like, "Oh geez, we're an enterprise company. I'm going to charge them a lot of money. I'm going to charge them $10,000." Now we track a 100K+ customers. That's why I'm laughing.
‍

So even though we had this self-serve funnel, nobody wanted to use it. So the reasoning that I figured out was they wanted to go through procurement, they wanted a contract, they wanted a security questionnaire, and they were certainly willing to pay money for that. So we still, by the way, have a self-serve funnel. If you're listening to this podcast, you can show up at LaunchDarkly, throw down a credit card, use it immediately. And we love the people that do that. But we also have a lot of folks who do want, I'd say, more of an experience with us.

Sandhya Hegde (27:10)

When did you launch the testing and production meetup? That feels like a really big part of the community identity and buzz around LaunchDarkly, at least in 2017, when I first discovered it. I would love to learn more about that.

Edith Harbaugh 

Oh, gosh. So we had a meetup from the very early days. We were in an incubator called Heavybit, which had an office in downtown SF at 9th and Folsom. And it was just this hotbed of dev tools, and I loved it. CircleCI was a cohort ahead of us. PagerDuty had just gone through. It was great because you could see what all the other companies were doing.
‍

So what companies did back then was meetups, and the joke was if you wanted free pizza, you could get it any night you wanted, but your meetups weren't supposed to be about a vendor. They were supposed to be about an idea. So ours was actually about LaunchDarkly, but we're like, okay, we can't call it a LaunchDarkly meetup, what can we call it? I think our marketing hire at the time, Andrea, who's a wonderful marketer, she was the one who decided to call it “Test in Production”. But my memory is honestly a little bit hazy now.

Sandhya Hegde (28:32)

And was that effective?

Edith Harbaugh 

To be honest, yes and no. We had a group of like 10 to 20 folks that came. So if you measure it by a ground swell, like a HashiCorp style of hundreds of people coming, absolutely not. Was it 10 to 20 people that cared a lot and would show up over and over? Yes. I still remember there were some people who made it a point to try to come to every single one, and no, not just because of the free pizza, but because they liked the speakers, they liked what we were doing.
‍

But the metrics when you're in early days with startups are sometimes so small but so meaningful. I remember our very first meetup, the one I said before I was nervous talking about, we had I think 15 people show up, but from it, two became customers that are still customers to this day.

Sandhya Hegde (29:22)

And I think you all did a really good job of surfacing the conversations from the meetup eventually on your blogs and spreading that to a wider audience because that's how I discovered Test in Production. So, it definitely had other avenues of impact that are harder to measure.

What would be your recommendations specifically to dev tool founders trying to build an audience today? What's the thing that people are doing today that you think actually works, and if you were starting LaunchDarkly right now, would you still do the meetup? Would you do something different?

Edith Harbaugh 

Yeah, and again, I want to give Andrea, who was our marketing person at the time, a lot of credit for that one because her strategy was always, you get the content, and then you slice it and dice it and distribute it as many ways as you can. So we would have a meetup that literally had, if it was rainy and there was a Warriors game, had like 15 people show up. Not 50. Once we had one during Dreamforce which had like seven, but we would get a nice video of the speakers, we would do a good blog, and then we would put it on Hacker News. We would put it on our Twitter feed, we'd put it on our LinkedIn. So we would get a lot of distribution from one 10-person or 15-person meetup. And by the way, in the video, you didn't see the size of the audience, you just saw that it was really professionally shot and it looked good.

Sandhya Hegde (31:00):

I just assumed there were like 500 people there, Edith, so great job. That's great marketing. And do you remember, what was your Series A raise like? I think this was about maybe 18 months into starting the company, you still only had a few paying customers. What was that like at the time?

Edith Harbaugh 

Yeah, it was difficult. So we started at a very different time in 2014, which I think is now perhaps similar to 2023 in that dev tools were not seen as a lucrative space for people to invest in.

Sandhya Hegde (31:40)

Yeah, I actually distinctly remember ... I've been in and out of venture a couple of times and this was my previous stint in venture before I worked at Amplitude, and I remember the general wisdom was developers don't really like paying for tools, they will use free tools, but if it's paid, they would prefer to build their own.

Edith Harbaugh

Yeah. It was definitely the contemporary wisdom. I think that the way we flipped that was people will pay for something if it saves them time. But you can never go head-to-head with a developer and tell them they can't do something. They're just going to get their back up and say of course they can build this. But if you can say, "Hey, it's going to make your life better and you can do something else," they will happily pay for a tool.
‍

So, our fundraising for our Series A was extremely difficult. We were eight people at the time. We had some early traction. I think we were around a hundred-ish customers, and I went out and tried to raise and I got no after no after no. And it was really frustrating because we were starting to see some traction and we did want to hire more people. Like I said, we were eight people, and we were eight people who really should have been at least 15.
‍

And I still remember it. The team could see me going out and fundraising. When you're eight people, they know where you are. And I couldn't get the money. It was really painful. We would get very late-stage, and then somebody at the firm would block it, which was hard. It was hard for me.
‍

And finally, because I did finance back then, because we were eight people, I looked at all our spreadsheets and I could see the money coming in. I could see our burn. And I told the team, "Look, we're fine. We're not going to get the splashy $10 million raise, but we are very near break-even." In fact, next month, we are very, very close to break even, and then we're throwing off enough cash that we can start to hire a person every two months or so, depending on the sales we bring in. So like, "Your jobs are safe, the company is safe, we're just not going to get the splashy infusion."
‍

And the team was really good. They were like, "Oh, we get that. That's great." So we started hiring our ninth person, and in the meantime, we'd put out this big bid for a contract and it came through, and then suddenly everything changed because we crossed some revenue milestones and then we could get our Series A. So both things came true at once that we were a going concern. Like I knew that we were going to make it, we just wouldn't have the TechCrunch byline... and then we got the TechCrunch byline too. By the way, I don't think TechCrunch covered it. I think it was VentureBeat, but …
‍

But around that time, I don't know if it's still so true for startups, but Series A was seen as just this really hard thing to get. There were all these articles back then about the Series A crunch, and I actually wrote a blog post called, "Revenue Is Its Own Series A." You can get money from two sources, you can get it from VCs, you can get it from customers. I actually like customers better. No offense to our VCs, love you, but customers are a renewable resource.

Sandhya Hegde 

They don't even ask you for shares in return.

Edith Harbaugh 

Well, a customer will use the product, love the product, buy more, go to another job, buy more there, and you'll transform their life.

Sandhya Hegde (35:17)

And I think the Series A crunch you talked about is very much what we are already seeing right now, starting late 2022, and we'll probably be seeing for at least all of 2023. I'm curious, what would be your advice for Seed Stage founders today? Either those who are just getting started or have raised a bunch of funding but just haven't hit the revenue milestone they are at yet, that they need to, what would be your advice for them?

Edith Harbaugh 

So, funding is never an end, funding is the start to the next phase. So every round we did, I would think about, how am I going to get our metrics together to get our next round of funding or to operate the company as a self-sufficient business? So we always,  had this mindset that we would be okay even if we don't get VC funding. Where I see companies get into trouble is if they had a really easy round, for example, if they're a hotshot out of Facebook or Google and raised a $6 million Seed round and assume your next round is going to be easy, because you just sometimes run into a wall then, whereas I said earlier that our early fundraising was so hard that I never took a round for granted. Even when funding was easy, I never thought it was just going to come. I always thought, "How am I going to run a business?"

Sandhya Hegde (36:50)

I'm curious, Edith, how do you invest in your own CEO evolution? Having seen this upfront so many times, I feel like as a CEO, you just have to become a different person and a different leader every six months, at first, and then maybe every year or two you have to evolve so much. How do you think about that as a founder?

Edith Harbaugh 

As a founder, your role changes every three-to-six months, if not more. In the very early days of the company when it was John, my co-founder, and myself, I mean I was an IC. I had the title of CEO, but I also did everything else. When we were eight folks, I think I was more of a team lead. I still did a lot of IC tasks, and then once you're around 30 is when you actually start having an executive team, even a very early one, and start acting less as an IC and more as a leader.
‍

Now we're a 500-plus company, which is global, and my role is very, very different than it was eight years ago. I think it's just a continual evolution of looking at everything that you were doing and figuring out if somebody else should be doing it besides you.

Sandhya Hegde (38:11)

How do you figure out what it is you need to change next? In terms of just what you spend time on, how do you decide, okay, it's time to stop doing this or start doing something new? Is it based on feedback from your company? Do you work with exec coaches? Like what's been your go-to?

Edith Harbaugh

So in the early days, I said we'd originally planned to be self-serve, and then we discovered that our customers didn't want to buy that way. So I was the first salesperson for a while, and then what I realized was that I was the bottleneck. People would say, "Where's my write-up? Where's my proposal?" And I would be perpetually behind, because I was also trying to do marketing and fundraise and also run payroll, by the way. So I said, "Okay, look, we need a full-time salesperson." And we hired somebody just to do sales.
‍

Another thing I wished I'd hired in the early days was I did all the ops too. Literally, I got on the phone with the healthcare company to negotiate our contracts, which in hindsight, I absolutely cringe because it's stressful. It's actually a ton of work. I wish that we had hired an ops person earlier because it was just taking a lot of my time. 
‍

I think it's tempting for people to say, "Oh, this is the one thing that led to their success." We just honestly tried and ground it out, and I hope that's inspirational to folks that you might have a wonderful idea, which I think LaunchDarkly was, but there was still so much hard work and it wasn't just me, it was my co-founder, it was the early team, and I hope I've given credit to people like Andrea, who was our first marketer. But it is a lot of work to get to success, but it is very rewarding.

Sandhya Hegde (40:07)

Awesome. Well, Edith, this has been exceptionally informative. Thank you so much for sharing the LaunchDarkly product-market fit story with us and our audience. We'll be sure to send you questions that are coming up from this. And a big thank you for your contribution to the dev tool community. I think there are so many founders who see LaunchDarkly as an inspiration and know that the VCs don't always understand what developers will pay money for, so thank you so much.

‍