Founded in 2011, Segment has raised more than $100M and boasts an impressive client base with thousands of companies across 71 countries. Peter Reinhardt, Segment’s founder and CEO, discusses his approach to product-market fit, fundraising, sales, content marketing, and more.
The early days of Segment
Q: How did you come up with the idea for Segment?
A: To be fair, we started with a few awful ideas. I grew up in Seattle with an interest in space and math, so when I got into MIT, I thought I would enter math/space engineering. It wasn’t my calling however, so my two roommates and I decided to start a company.
At first, we had this idea to build a classroom tool—where students could press a button to show how confused they are over the course material during a classroom period and then there was a graph that could show how confused they were over time. It was a horrible idea. We ended up sitting in the back of a few classrooms after we launched the pilot program and found that students were just on Facebook the entire time. It was even worse because we had raised $600k, so we called back all the investors saying “This was a bad idea, do you want your money back?” Interestingly, they said that they had invested in the team, not the idea, so they wanted us to find something else to do with the money.
Q: How did you get from that idea to Segment?
A: Well, after that we decided to build a metrics analytics tool to compete with products like Google Tag Manager and Implement Mixpanel. We spent over a year trying to figure out how we could differentiate, but it was just a bad market—too crowded. After about a year and a half, we ran out of the money in December of 2011 and needed help. We approached Y-Combinator, and they said, “so just to be clear you spent half a million, and you have nothing to show for it?” That gave us pause, so we took a step back.
We remembered the first week, when we had started working on this project and couldn’t figure out which tool to use: Tag Manager or Mixpanel? Rather than figuring out which to use, we started writing code that would implement them all and figure out the most optimal one all on its own. Slowly our product began to evolve. It’s funny because we were getting objections from potential customers who were saying “We’ve already installed Mixpanel, so we don’t want to use another competitor,” which is the exact problem we were there to solve.
Before we launched in December 2012, we open-sourced our code, built this beautiful landing page, and posted it up on Hacker News with a pointer to a hosted beta to see what happens.
Q: That was what took off?
A: Exactly. Segment flipped the industry over on its head. It was a complicated and crowded space with so many analytic tools and advertising tools. Since the area was over-crowded, we turned that into our biggest weapon. Engineers couldn’t even figure out how to incorporate everything, so because Segment integrated all these components, engineers needed it to understand the market. Since then, it’s been a pretty smooth curve, but I’d say the longest struggle was finding that product-market fit.
Product market fit
Q: Looking back, can you see any ways to avoid the year and a half you spent wandering around, trying to find product-market fit?
A: I don’t think so. Most companies wander around for a while: Airbnb and Slack, to name a few. Slack was building a game before it became the successful chat platform we know today. As long as you’re humble in the face of the truth of what the world is, and as long as you stay persistent, you’ll find your way.
Q: As you’re wandering around, how do you know when you’ve found product-market fit?
A: If you’re not sure whether or not you have product-market fit, you don’t have it. There are a lot of metrics you can use: whether or not people will pay, if people are willing to share, etc., but you can always find an exception, so looking at specific metrics isn’t an accurate representation. With product-market fit, it’s really about the feeling. It’s also dangerous, however, because you can feel like you’ve almost found it and get stuck in a loop. With our product-market fit, there was a long time where we were down before it suddenly just exploded.
Q: Since you went through so many experiments and ideas, how do you think about iteration and revision?
A: Founders take way too long to kill ideas. If you’re looking for how to do it well, CodeAcademy is a notable exception. They tested about a dozen ideas for two months, and three days before Demo Day they decided to scrap their previous idea and launch what became their platform today.
Q: When you’re testing out all these different options, how much of the product do you need to build out?
A: I think the only answer to that is, “build as little as you can get away with.” For Segment, it was 500 lines of code. There’s plenty of companies that have gotten away with less, but it’s certainly not 100,000 lines of code. We did that for our education tool, and it was totally unnecessary.
Q: You’ve raised over $100M total, up through the Series C you completed last year. What advice do you have for the fundraising process?
A: Try to compress fundraising in the shortest amount of time as possible. Twelve months before the raise, I meet with investors to explain our strategy and trajectory. This way, the fundraising process can be really short because you’ve already built rapport when you meet with them to raise. Although some will have other opinions, I believe that contacting an investor when you’re not raising is actually super productive—I spend about 2 hours a week (1 hour each) meeting with VCs. It’s important to start building these relationships and consistently explain your strategy. I also use these VC meetings to gauge whether or not they’re going to add value to us in the form of strategy and creative tactics.
Q: What do you send over before these investor meetings? A deck?
A: No deck, just intros. We never reached out cold. And for those intros, we focus on momentum rather what the product is itself. We don’t reveal finances, no cash, no gross market, no financial metrics. We talk about change and momentum and don’t let anyone pin down specifically what’s happening in this company. And if they ask, we say we’re not fundraising right now, but we’d be happy to chat about it later.
Q: Did you raise between your initial $600K and first round?
A: No, and I think that was essential to our success: It’s important to keep your costs low and be incredibly conservative before you’re super certain of product-market fit.
Q: After you successfully launched, how long did you wait before raising again?
A: We waited as long as possible, which was ultimately disastrous. We thought that we would be able to raise instantaneously as we had done during seed. Ultimately we were raising with about only $25k left in our bank account. At the end of the day, it worked out because our product-market fit was so strong—we were getting about 100 new clients every week. Our pitch deck was just a graph with an upwards and a few logos lined up at the bottom. In retrospect, it was a horrible idea raising so late, but we were lucky.
Q: There are hundreds of micro-seed rounds now, backed by founders who are confident in the mindset that if they have more money, they’ll figure it out. How do you avoid this mindset?
A: If you’re in that micro-seed stage, you have to be incredibly conservative with cash. Most companies take a year and a half to three years to find something, so you have to stay alive that long. At Segment, there were 4 of us, and we were paying ourselves $30k per year. There’s a good essay on this topic by Paul Graham (Why to Start a Startup in a Bad Economy), where the gist is: If you survive long enough, you’ll eventually find something. You just have to be a cockroach of the corporate world.
Q: Looking back, what allowed you to raise at the impressive levels you did?
A: I think one key was working with the investment firms’ portfolio companies. We had already been put on the investor’s radar, so they were immediately curious as to who we are, which created immense traction.
Q: Before you found product-market fit, you didn’t have a great revenue number. To get there and get attention, how much did you spend on marketing and sales?
A: The frustrating thing about Segment is that our best channels are difficult to attribute. We can’t see through any sort of attribution of how customers discover Segment, and even the feedback isn’t super helpful because they just tell us, “I’ve heard about you from the internet.”
For us, the best approach has been long-form content, where you publish an article, and it gets shared. We’ve found it incredibly valuable to invest in things that produce long-term results: pure op-ed and advertising can saturate the channel, which can juice the growth curve for a little bit but won’t sustain it.
If we were to have the chance to do this all over again, I would absolutely invest a lot more in long-form content. It works, and I regret that we didn’t always invest in it.
Q: How did you land on long-form content as a strategy, and what were some of the hacks you used to create content, put it out on the right channels, and see how customers and readers were engaging?
A: We started reading Hacker News in 2008/2009 and really liked the community, so we wanted to be a part of it. It definitely didn’t go well at first, but we started to notice what content did well and what didn’t. We were able to get a better sense of the quality bar and what structure works. Looking back, it probably took us about 3-4 years to start doing it well. Now, we’re much faster. When we experiment with LinkedIn, for example by publishing a series of videos each week, it takes us a couple of months to get to the point of proficiency.
We also did make sure to test whether the content was going to resonate with the community before we published it. In this case, it was my co-founder’s brother—we’d send it to him and hear his feedback. It was a very honest impression with good feedback because he was a business owner who was busy, didn’t have a lot of free time, and wasn’t based in the Bay Area.
Q: How did you get the content written?
A: I don’t believe that content written outside your company is very effective because it’s not authentic. We experimented a bit with this, but we consistently found that outsourcing never produced content nearly as well.
We’ve found that developers responded really positively to things that were written in-house. The only problem was that it often took us months just to get one piece out because we would mull over the fine details over and over. Our co-founder Calvin spends 10-12 hours just editing things for a single blog post. That said, when we published it, readers could tell that we were truly invested in the content we put out.
Q: For content marketing, how do you prioritize different channels?
A: 80% of our leads today come inbound: Half come from customers telling others, and the other half is brand awareness. We’ve created this stream where we focus on the company’s engineers, which will lead to the marketing team, which then goes to the CTO who makes the decision. Our marketing strategy may look a bit weird from the outside, but we’ve narrowed down the process, so we better understand what we need to prioritize.
Q: When would you recommend as a cadence to content creation?
A: We’ve always been very focused on the immediate-problem in our business, so we didn’t create any content during product-market fit, and then we were concerned about traffic and awareness, so marketing became the biggest thing, and then our problems shifted again. We would keep switching around—sometimes we wouldn’t write content for nine months, write for two months, and then not write for another six months because we switched our entire focus to sales. This is not ideal though—consistency is key to getting your audience engaged and keeping them engaged.
Q: What’s unique about your sales team and their success?
A: Our top performing sales rep doesn’t have a background in tech at all—it drives home the point that great sales is about following-up and understanding the customer; it doesn’t matter what product you have to sell.
As for our customer-engagement employees, they’re not commissioned, which is important, because this way you can tell customers that they’re purely there to make sure that the client is happy.
Q: What was your strategy during the first few deals?
A: It took me almost a year to close our first two deals. By the end of that year, we had hired a salesperson and a team-support member, and they were far more effective. The first sales representative was amazingly successful. He took the first couple of weeks to test things out: a new angle or a new message. He then figured things out with metaphors and would do this pitch where he said the market place had moved from the Serengeti, which has a homogenous population, to the rainforest, which has a lot of diversity, so you need Segment to connect all these various types. It was brilliant.
Q: When putting the product out into the market, what’s the best way to go about finding the right level to price it at?
A: After about six months we hired a sales advisor, and we went into our first sales meeting together. As we were walking into the meeting, he said, “Okay you’re going to ask for $120k.” We had been charging $120 at the time, so I thought he was crazy and said no. He said that he’d quit if I didn’t, so I went in, and when they asked about the price, I said $120k. They said that I was crazy, it was too high, so they offered $10k, and we ended up settling at $18k.
We just continued to use that strategy. The big takeaway: Unless you ask, you won’t ever find the price ceiling.
Q: Did you ever have problems with competitors?
A: Not really. We have three competitors: one in Cincinnati, which targeted our customers and would undercut us. It’s never a good idea to compete on price, though, so when they shut down, we found that we hadn’t lost too much. There’s another one in China that just popped up on our radar because a client actually told us. There’s also one in New York which competes at price and just undercuts at the end of the sales cycle.
Q: You have open source code on your website—how do you position this when you’re trying to make sales?
A: I think there’s a bit of a misconception in terms of how our sales are done. The open source attracts the developers to the ecosystem, but it doesn’t solve the business problem. Imagine you’re an engineer at a company and marketing asks you to show them analytic tools. You can create something, but they’ll keep coming back asking you to change things. The open source doesn’t completely solve the problem, but the hosted version does because it hands the keys to the business team. That’s the secret: No one actually takes the open source code; they just point their business team in our direction telling them that we have something that would be helpful.
- If you’re not sure whether you have product-market fit, you don’t.
- VCs invest in early-stage companies for the team, not the product.
- You can turn an overcrowded market into your biggest weapon.
- Deep, high quality, evergreen content can have a long-term impact where quick-fix marketing simply booms and fizzles.
- Customers are often willing to pay much more than you expect—even to the tune of 5-10X.
- Meeting with VCs throughout the year, even when you’re not fundraising, often pays dividends later.