by Y Combinator8/4/2016
Suhail Doshi, cofounder of Mixpanel, on creating demand by educating journalists and VCs.
Kat : Where did the idea for Mixpanel come from?
Suhail : At Slide, when I left, I had thought that it was a bit silly that both Slide and all its competitors were building analytics tools in-house. That seemed very strange. Why were they building internal tools when there were clearly other analytics companies in the world? There seemed to be a gap in the market. It wasn’t clear whether the product they were building internally was widely applicable, but it just seemed like such a rational logical thing, like wouldn’t you want to know how people are using your products? Wouldn’t you want to use the best metrics to figure out whether your products are objectively good or not? That’s what I took away from Slide. It was the fact that they were doing it, all their competitors were doing it, and I thought, “Why are they all building it? What if we just built it and then they just paid us?”
Kat : Right. All of these larger companies were building these internal teams to build the software in-house, to build analytics in-house, and you decided, “Why don’t we just do it and sell to all of them?”
Suhail : Exactly. It just seemed silly because Slide had about eight to ten people doing data analysis for them full-time, and then they had a bunch of infrastructure, lots of servers, MySQL servers, things like that. So it just seemed odd to me that you calculate all that in and you’re going, “Wow, Slide’s spending like a million a year just to do this thing. If I could do this for them at half the price, maybe they would pay for that.”
Kat : Can you paint the picture of what the analytics market was like back when you started?
Suhail : Yes. Well, it’s interesting because when we started Mixpanel, I wouldn’t say that we put together this very complex business plan and did all this market research to determine what the competition looked like and what we could do. But there was something that was important to us, which was that when we started Mixpanel early on, we made this decision to not track pageviews. The reason why was because we felt that there are all these companies that already did this, and Google Analytics was pretty good. They’re still pretty good at it. We thought to ourselves, “What extra value could we really add to the world if we track pageviews.” This notion was somewhat risky at the time because it was very odd to build an analytics company that did not track pageviews.
We thought to ourselves, “Well, if we can’t add any extra value because we track pageviews more so than something like Google Analytics, then we ought to do something else. If we can’t figure out what that something else is, well, it didn’t have legs anyway, and we deserve to die as a company.” It was a harsh view of the world in terms of what we thought we should do. What we decided was really like, “Well, with the advent of AJAX and Web 2.0, websites are becoming less page to page based.” If you can imagine this, it’s not really like GeoCities where you just click on these bright blue links and you’re going to the next page anymore. It’s actually more like you’re going to Pandora and you’re playing a song, but there was no page load. Or, you’re on Facebook and clicking through photos, but you’re not loading a new page. It’s just taking you to the next photo.
We thought that that was the future. We said that the way people are going to build things in the browser, it’s going to be more rich. This is very obvious in 2016. You wouldn’t expect anyone building a product any other way, but this is when Web 2.0 basically happened. We thought to ourselves, “Well, if you’re not going page to page, then how are we going to measure all these things that people are doing?” Then we thought to ourselves, “Well, the right unit of measurement then is probably an action, an event, something that’s happening on that page.”
We thought, “Well, let’s just track engagement instead. Let’s track the actions that people take, and let’s make it really easy to track those actions because those actions are probably more interesting than someone just going to a page.” In fact, those metrics are better. By the way, we saw companies like Slide benefit from being able to track engagements. It was somewhat twofold, we said to ourselves, “If we can’t figure it out and we can’t get people to do this, then we deserve to die as a company, and we’ll figure out something else. But if this does work, and we think it could work, we should definitely do it.”
Kat : You had to basically educate the market, educate people that pageview tracking doesn’t make sense and that tracking engagement is much more powerful. You still have to do that. What kind of things did you do early on to help people understand that tracking engagement and actions was the way to go?
Suhail : Right. There were a bunch of people that were popularizing this idea that you shouldn’t track vanity metrics – downloads and installs and pageviews – things like that. Number of users that have ever signed up for your service. LinkedIn does that, for example. They just say, “We have 300 million registered users.”
Kat : Great. How many active?
Suhail : Yeah, how many active. That’s like an obvious thing that people ask now. One of our first steps was riding that wave, and a lot of people helped us inadvertently, so we didn’t have to do a whole lot there. But then we wanted to start being more proactive about it, one tactic that we started to use was that we started to educate journalists. That was key. We educated journalists and we educated VCs. We knew that if we educated both of those parties they would ask founders and entrepreneurs, “What are your actual metrics? We don’t believe in these other ones.”
For example, one person that published something was Liz Gannes when she was working at AllThingsD. In conjunction with me and Mark, we just called out. We said, “Companies should stop tracking bullshit metrics.” We came out together and said that. So we started educating all these journalists over time and saying, “Hey, you wrote something about how LinkedIn has registered users. That’s silly. Because doesn’t it matter how many actives they have?” This is such a logical thing to people that it’s not particularly difficult to…
Kat : To sell them on that idea.
Suhail : Yeah, right. So we did that and then we educated investors. We tried to make sure that. Investors like this because they were savvy enough to realize this but it was getting hard for them to push entrepreneurs to all care about it.
Kat : Did you figure it would filter down into their portfolio companies?
Suhail : Exactly. We felt that investors liked it because they got better numbers for them to evaluate the businesses, and we liked it because it forced the founders to track better things and then buy Mixpanel. By convincing these two parties we were able to get a huge portion of Silicon Valley to get to the point where no one here is talking about how many pages they do, or how many registered users they have anymore. How many downloads are they doing? Nobody cares.
Y Combinator created a new model for funding early stage startups. Twice a year we invest a small amount of money ($150k) in a large number of startups (recently 200). The startups move to Silicon