E63: Peter Reinhardt, CEO and Co-founder of Segment. Since this interview Peter Reinhardt, CEO and Co-founder of Segment, and his team have raised another 175M USD to bring total funding to to $284 million and a valuation of over 1.5B.
He validated an idea he was trying to kill by publishing a link to a sign up page on Hacker News – pretending they’d built it. The response was massive and Segment built it’s 1st hugely successful product in just 5 days
Peter talks about why building a business where the target customer is developers can to some degree sidestep the question of what company size range you chase after initially. He describes Segment as a customer data infrastructure company – he wants to help customers manage all of their data from different sources. He lets us know whether or not he’s interested in building a marketing automation app on top of that and gives his view on the rise of 100% remote teams. This episode is about how Peter helped lay the foundation for building of a unicorn, by trying to kill the idea before it was brought into existence. If you love classic early validation stories, tune into the next instalment of 14 Minutes of SaaS. He also talks about how selling to developers can bypass the old chestnut of whether you want to go after SMBs or the Enterprise first.
TRANSCRIPT
Peter Reinhardt
The most important thing about a business in the early stages is finding product market fit. After product market fit it becomes go-to-market and distribution, but prior to product market fit nothing else matters. And I think it really requires actually a scepticism. And I think the failure mode for most companies prior to product market fit is that they drink their own kool-ade. They have a vision for the world – of how the world should be. And they build a product to try to make the world like that. But actually the world doesn’t give a shit.=
Stephen Cummins
Welcome to 14 minutes of SaaS, the show where you can listen to the stories and opinions of founders of the world’s most remarkable SaaS ScaleUps.
Since this interview Peter Reinhardt, CEO and Co-founder of Segment, and his team have raised another 175M USD to bring total funding to to $284 million and a valuation of over 1.5B. Peter talks about why building a business where the target customer is developers can to some degree sidestep the question of what company size range you chase after initially. He describes Segment as a customer data infrastructure company – he wants to help customers manage all of their data from different sources. He lets us know whether or not he’s interested in building a marketing automation app on top of that and gives his view on the rise of 100% remote teams.
Ok – we’ve got Peter Reinhardt, CEO of Segment here at the WebSummit. Nice to meet you.
Peter Reinhardt
Thanks for having me. Nice to meet you as well.
Stephen Cummins
Tell me a little bit about your life up to Segment
Peter Reinhardt
Sure. I grew up in Seattle and was really into math. Then went to MIT and decided math was a little too abstract. Then studied physics. Decided physics was a little too abstract. Then studied aerospace engineering. I wasn’t going to get my aerospace engineering degree. But anyway I decided aerospace engineering was a little too abstract anyway. And ended up dropping out with my roommates and starting a company. So made the whole transition from math to business. For the last seven eight years have been working on a company that’s now doing reasonably well.
Stephen Cummins
What made you decide to found Segment?
Peter Reinhardt
So originally we … myself and my roommates in MIT … we were just interested in starting a company together. So we were roommates and best friends and just really wanted to spend more and more time together working on cool things. So we actually started as a classroom lecture tool. And the idea was to give students this button to push to say ‘I’m confused’ and the professor would see this graph over time of how confused the students were with. We thought it was cool. A bunch of professors thought it was cool. We got into Y combinatory – startup incubator – with this idea. And raised like 600K with this on demo day. And then we put it in the classroom and it was just a total disaster. All the students just opened their laptops and went straight to Facebook. So we had this like big vision of how we were going to rebuild and reinvent the classroom. But the reality was that it just like didn’t solve the problem and just made engagement far worse. So sort of our first example of, you know, vision meeting reality. I think the grand vision meeting. What the world really needs sort of head on. So from there we pivoted to an analytics tool … tried to build that out and that turns out to be a very crowded market and not a great place to try to build and sell software.
So we got about a year and a half in by December 2012 and basically realize that we were failing. And we need a new idea. We had about 100K left in the bank. So over that previous year and a half we had built this really small open source library for ourselves and we called it Analytics Jazz and it had a few stars on Github and so forth. And all it did was is it could send data from your website to Google analytics, Kissmetrics, and Mixpanel. Really simple way of implementing three analytics tools at once. And yeah, it had a few stars and getting – maybe 25 stars. We were looking for sort of our last shot on goal. And my co-founder Ian was like ‘you know what, I think there’s a big business behind Analytics Jazz. I. Think there’s a big business behind this data pipeline. I can send it out to different places and I was like ‘that’s literally the worst idea I’ve ever heard right. It’s 500 lines of code. It’s already on Github and open source. I don’t understand like how do we build a business around that… that makes no sense? We fought about it all day long all four of us and I went home and I was like racking my brain trying to figure out how I was going to kill this idea. Finally figured it out and came in the next day and was like ‘okay, guys .. here’s what we’re gonna do … we’re going to build this beautiful landing page – it’s really going to pitch the value of Analytics Jazz and at the bottom it’ll have an e-mail forums where people can sign up as interested in a hosted version of the product.
And then we’ll put this on Hacker News and we’ll see what happens to be like a fair test of whether or not this thing actually is good for anybody. And I was thinking like okay ‘great it’s dead’ …. so we build the landing page, put it up on Hacker News. I start thinking about other things. Meanwhile it goes straight to the top of Hacker News. Gets a few 100 uploads – sits at number one for most of the day. It gets thousands of stars on Github, thousands of e-mail sign ups. People are reaching out to us on Linkedin demanding access to this hosted beta that doesn’t exist yet so … It was just sort of an explosive product market fit moment for us. And I was obviously wrong. We were obviously really solving a problem for the world and so we built the hosted version of the product in about five days. Launched it and by the end of the month had about 70 companies sending 10 events per second through the system. So. Very sort of landmine product market fit moment in the way
Stephen Cummins
You are the first founder I’ve met who started something by trying to kill it.
Peter Reinhardt
Yeah!
Stephen Cummins
It’s quite am amazing story I’m. Now looking at that space there’s not too many companies in it at the moment that’s. What’s the reason for that?
Peter Reinhardt
There’s actually a lot of companies in marketing technology broadly.
Stephen Cummins
Sure.
Peter Reinhardt
Right? …. So marketing technology broadly has really actually sort of exploded over the last few years. A few years ago it was maybe like 150 or so companies …. now it’s well over 6,000. So the number of marketing technology companies is… is staggering. But segment is pretty differentiated within that. So the vast majority of marketing technology companies are focused on a specific channel. So you might have, you know, a few 1,000 advertising technology companies, a few 100 analytics tools, a few 100 e-mail marketing, tools a, few 100 attribution tools and so forth. But we are in this more infrastructural layer where we’re sort of the pipes that are actually getting the data to this flowering ecosystem of all these different marketing technology tools so. In that sense we’re very differentiated and being that infrastructure layer and there’s very few, if any, sort of other players there. And primarily we’re competing with people who are sort of doing it the old fashioned way of going in manually – integrating the pipes to all these different tools, which is super laborious. For the most part they’re just not aware of an alternative.
Stephen Cummins
Ah okay. And. You, know … when I look at the people who are at the infrastructural level in this space .. one of the things that the crowdsourced data I’ve looked at show me that you are particularly strong on being easy to set up, easy to share, easy to maintain over the life cycle once a company starts using you. Is that a massive differentiator for you. Has that been part of the reason why there we go so successful. A lot of word of mouth stuff. Is it a massive differentiator?
Peter Reinhardt
Yeah. And you know we call this category customer data infrastructure and the reason why we are at the infrastructure level is because it’s targeted at developers which then gets at why it’s so easy to use. So I think some companies think about selling to SMB and you, know small, businesses and some companies think about selling to the enterprise. When you sell to developers, or through developers, thats sort of your primary channel … you often actually don’t have to make as much of a choice because developers are actually fairly similar across all different sizes of companies.
And what they weren’t… they weren’t very simple docs. They want a simple setup flow where they can get started. They want very well built products, and we’ll built SDKs for implementing and collecting the data. So they actually want a common set of things across all different sizes of companies. And, they want that sort of common data infrastructure if you will so. We we’ve differentiated I, think by approaching it from a developer perspective …. partly because we were developers to begin with but, that gives us access to sort of the entirety of the market in a very efficient sort of interesting way.
Stephen Cummins
So companies that sell into communities of developers and have a marketing aspect to them … Branch comes to mind for example. One of the big things that helped them grow has been building community. Have you a strategy … or have you had a strategy in community build?
Peter Reinhardt
We haven’t really actually. So I would say the majority of our sort of like presence in the engineering community has been more sharing open source. So branch built that community specifically around deep link sharing. Or, deep linking. Our community has been much more or less around the product specifically and more around the open source that we’ve contributed. And I think we have close to a 1,000 open source repos at this point. And, there’s like activity in a lot of them at the moment. Some are super popular like Nightmare JS is well over 15,000 GitHub stars at this point. So. Sort of like a pretty vibrant community around the open source that we built and then also fairly sort of like vibrant blog where we write about really deep infrastructure engineering problems that we’re solving. How we did a particularly database sharding in this instance or …. how we build sort of more of like a TCP IP kind of routing layer for data internal to our systems as opposed to a PREQIN infrastructure. Like really sort of like pretty deep down in the weeds – and we blog about that and that’s pretty helpful for the community. So we build more around infrastructure open source as opposed to around how to do analytics specifically.
Stephen Cummins
Your timing is obviously very good. You’re feeding that exploding marketplace right now. It’s very good timing and. So what would your vision be? I normally would ask for the short to medium term … What’s your long-term vision actually for Segment?
Peter Reinhardt
Yeah so? We’re very interested in continuing to be this customer data infrastructure layer so I think the most common question that we get is ‘are you going to become the app or are you going to build an e-mail marketing tool? Are you going to build an analytics tool? And the answer always has been … and I foresee being for a very long time if not forever like we don’t want to build the apps. We want to build the infrastructure and I think there’s a couple of interesting trends that you can see. One is that as the world is moving online the number of channels is becoming very… very large so you know. In the old world that was offline … you would go talk to someone in person or you might call them. Those are the two channels that were available. Maybe direct mail … 3 channels. In the digital world you have all, three of those plus you have e-mail, plus you have push notifications, plus you have the website, plus you have a mobile app, plus you have like an iPad, you have like all these different channels that you can track the people on Facebook etc etc. So you have. Like. This massive explosion of different channels to communicate with people on. But at the same time you need to be able to control and organise all the data that you’re using to interact with people from all those different channels and so … you have this force that sort of trying to pull together one layer so that a different layer can go nuts. And we want to be that layer where all of that consolidation of data infrastructure happens. And want to empower basically a sort of flowering ecosystem of communication channels and ways of interacting with customers on top of that.
Stephen Cummins
Very cool. I’m going. To bring it back to you now Peter. You mentioned recently you’re astonished at the rate of progress in longevity research. Tell me about your interest in that.
Peter Reinhardt
Peter Reinhardt
Yeah so I just happened to have had an opportunity to sit down with or learn from a few folks recently who are in the longevity research space and was just really inspired by what I heard. I haven’t really paid that much attention to it prior. And I thought it was … ah it’s a little cookie, or you know whatever … but was pretty shocked actually by how much research like legit research has been done in the last seven to 10 years. So for example it turns out that there are a lot of animals that don’t have increased mortality with age. So they’re just as likely to die at 20 percent of their life as they are at 80 percent of their life … which indicates they’re very different from humans. Humans, for example ,,, there’s a mortality curve … mortality per year … it goes up dramatically after roughly age 65. So that’s interesting like … Why did these animals sort of have the potential to live forever if it weren’t for accidents and infections and so forth. There’s other animals that live much longer. There are worms where they can extend the life now by a factor of 10. So there’s been like quite a few breakthroughs and it… it seems like aging is more of like a fundamental caused a lot of things like Alzheimers and cancers and so forth and so there may actually be a root cause it can be just knocked out from a generic perspective … that’s fascinating and I’m super impressed by the progress that’s been happening.
Stephen Cummins
I’m interested to know. If you feel that on premise humans will be disrupted as much as on premise software has?
Peter Reinhardt
So there are a bunch of companies that have been very successful at moving to a remote culture like Zapier or … eh ..
Stephen Cummins
Yeah loads of them like … Hotjar …
Peter Reinhardt
No, I don’t think that would work for me … but clearly works for some people and I’m sure that will go on. I think it’s not an abnormal thing now.
Stephen Cummins
So what’s your daily, routine?
Peter Reinhardt
So I wake up around 6.30. It depends on when my eight month old son wakes up .. and when he starts blowing raspberries then I usually go in and get him up and play with them for a little bit … and then take him to day-care and head into work. And I spend a good portion of my day now with… with customers and other folks externally sort of trying to understand the broader ecosystem and figuring out where segment can go next.
Stephen Cummins
If you were to give one a little bit of advice to another entrepreneur or somebody who is just spent started business … What’s the one or two key things from your experience?
Peter Reinhardt
The most important thing about a business in the early stages is finding product market fit. After product market fit it becomes go-to-market and distribution, but prior to product market fit nothing else matters. The question actually is what’s the most effective way to find product-market fit. And I think it really requires actually a scepticism. And I think the failure mode for most companies prior to product market fit is that they drink their own kool-ade. They have a vision for the world … We did this multiple times… They have a vision for the world, of how the world should be. And they build a product to try to make the world like that. But actually the world doesn’t give a shit. It doesn’t care what your vision is, right? Like the world wants what it wants. And either you’re gonna figure out what that is and provide it. Or not. And so I think a certain level of humility and willingness to be skeptical about your vision … so that you can actually go discover what it is that people want is the critical thing.
Stephen Cummins
Okay, great advice. Peter Reinhardt from segment. Thank you very much for being here on the program.
Peter Reinhardt
Thanks so much for having me.
Stephen Cummins
In the next episode, we’re still in the WebSummit. And we meet for the second time Nicolas Dessaigne, CEO and cofounder of enterprise search platform Algolia. The company is in the news having raised 110M US dollars after the interview we did. He was the second person we’ve ever interviewed. Checkout episode 2 of 14 minutes of SaaS if you’re interested in company culture and the first conversation that we had with him.
Stephen Cummins
You’ve been listening to 14 minutes of SaaS. Thanks to Mike Quill for his creativity and problem solving skills and to Ketsu for the music. This episode was brought to you by me, Stephen Cummins. If you enjoy the podcast, please don’t forget to share it with your network, subscribe to the series and give the show a rating.