Spenser Skates, the CEO of Amplitude, took a very deliberate approach to becoming a founder and left as little to chance as possible. Learn about the steps he took to prepare himself, and how he went on to build a multi-million dollar analytics business with 100 employees.
Hello everybody, this is Courtland from IndieHackers.com , where I talk to the founders of profitable internet businesses. In each episode I try to get a sense of how the founder got started with their business and entrepreneurship in general, and the goal is to learn what goes on behind the scenes so that the rest of us can learn to build our own successful companies.
Today I'm talking to Spenser Skates. Spenser is the CEO and co-founder of a San Francisco based startup called Amplitude. Thanks for coming on the show Spenser!
Absolutely Courtland, super excited to talk to you and the audience. You know it's funny, I wish Indie Hackers had been around when we first started. It was seven years ago when I first got into startups and I spent a huge amount of time just reading the stories of early stage companies, and how they got started, and all of that. Definitely super excited to pass along what I've learned since then.
Yeah I'm super excited to have you! I think Amplitude is an awesome company. The way that I would describe it is that, it's an analytics tool that you plug into your website or your app, and it will tell you exactly how your users are behaving so you can make better decisions and build a better product... is that an accurate description?
That's right on point. Just stepping back for a second, I think a lot of people hear the word "analytics" and there's like, hundreds of analytics companies that all do pretty similar sounding things, like, "Hey, we're going to help you make better decisions, and we are the most scalable and easiest to use UI, and we let you ask the most in-depth questions." but our focus is a very unique one, which is we focus on the product, we focus on helping you understand how your users are interacting with your products so that you can build a better product, so being able to understand the level of engagement people and more importantly, let me understand what's behind that... so why people are engaging or not engaging with the product, how the usage of different features impact things like conversion or long term retention, so ultimately you can take that and use that to understand how to improve your product, and to ultimately build a better one.
Yeah, I would love to deep dive into this later on, because I think there's a lot of confusion around when exactly people should be adopting these analytics tools and how they can get the most out of them, and get the right data for their companies, so it's just a subject I would really love to pick your brain on.
Absolutely, it's interesting, I think it's definitely not something that makes sense in every use case and there's definitely a time and a place for it. I'm happy to talk about it.
Can you give us a sense of how big you guys are as a company, or your revenue, or how many people are using Amplitude?
We have tens of thousands of products that send us data, and we have hundreds of customers on the enterprise side. We launched it four years ago, since then we've grown quite a bit. We have a hundred folks that work here today. We just raised our series C of $30million about five or six months back, now we're on a trajectory to keep up and continue that growth, so we've done pretty well over the last few years.
That's a tremendous amount of progress to make in just four years. One thing I really want to get into is how you became who you are today, because today you're running this business with a hundred employees, you're raising tens of millions of dollars from investors... how did you decide to become this kind of person, was there a point in your life where you just said to yourself "Hey I can be a tech founder if I want to"?
I really appreciate that question because I definitely think it's not one that enough people ask and talk about, and something I spent a really long time thinking about before jumping in. So it actually started back when I was in school at MIT, one of the things I realized coming out is that... you could decide to do almost anything in life, not that that's going to be easy or you can do everything, but hey you can play whatever character you want to play. I spent a really long time thinking and trying to understand, what do I know about myself and what do I know about what's important for me to do?
Not to get too much into the philosophical, but I realized a few things about myself. One, I've always been motivated, and ambitious, and driven, but at the same time I've always wanted to do that on my own terms. Then the other big one was that, I saw the people that were most successful in their lives and I wanted to understand common themes. What was it that really successful athletes, or entertainers, or business people, or researchers, or anyone in life, what was it that they had in common? One big thread that stood out to me was that they all were in it to do it for something beyond themselves, and they really wanted to help other people. My take away from that was I realized I wanted to set up my life in a similar way, and I saw a lot of people who were very unhappy with their careers and where they had gotten. Whether it's they got into medicine or something else, they had chosen it because they thought it was prestigious, or they thought their parents wanted them to do it, or they felt they could make a lot of money, whereas on the flip side the people who were the happiest, and most fulfilled, and the most successful were doing it because they wanted to help other people, or they wanted to make the world a better place, or they wanted to serve others, so that was just a really really strong thing that resonated with me when I was trying to understand what my role in life was. I spent some time looking around and one of my takeaways was that, the best way for me to create something that helps other people is through building a company, and that was going to be the most impactful way I could do that, because if you look at who I was, a new grad coming out of school, I knew some coding, I had seen other people take that path and that was going to be the most impactful way I could help others.
I looked at a bunch of industries at the time. I looked at consulting, I looked at finance, I actually spent a year in finance before coming out and starting a company, I looked at going to a big tech company, I looked at going into academic research... out of everything out there I had found that starting a company is by far the highest leverage way to impact other people. I ended up looking at a bunch of other folks at the time who had done that, so the founders of Dropbox were a few years ahead of me at MIT, Patrick and John at Stripe were from MIT, there was a bunch of folks that, with very little experience or training came out and made these massively impactful companies. So I said, hey, is this something I'm capable of doing? So I spent a long time reading lots of stories about early stage startups. I read Jessica Livingston's, Founders at Work, and a bunch of books like that. After about a year of research I was like, you know what, I think this is something I'm capable of doing and I could quite honestly dedicate my life to, for sure the next few years of my life, and maybe if that works out the longer term. For me it was really about, how do I make the biggest impact that I can? The cool thing about starting a company is that as we scale, if we're successful that's a really good thing for so many folks, it's a good thing for all the people I get work with, it's a good thing for my co-founder, it's a good thing for all the customers we help, it's a good thing for our investors, it's a good thing for the broader startup environment, it's just a win on so many counts. I wouldn't rather be doing anything else.
That's such a great answer and it's funny because I bring a lot of people onto the show, and I think one of the most common pieces of advice that I hear people say is, "Stop reading, stop thinking, just jump in and do it!" and you on the other hand, are a very analytical guy and your approach was to spend a lot of time reading. In fact, you had a very dedicated and thoughtful research path that you set out for yourself. What would be your advice for aspiring founders who might be completely inexperienced but trying to learn more? You mentioned reading founders at work by Jessica Livingston, but is there anything else you would read or things you would pay attention to?
Founders at Work was great. There's Startups Open Source, which is kind of the next gen version of it. I know a lot of the stuff that you have on Indie Hackers is great. Just being able to understand the really really early stage stories of as many startups as possible, I definitely think if you want to start a company go do it. There are a lot of folks who spend tons of time just reading and talking about doing it, but don't actually make the leap, but there's nothing that gets you more progress than starting. If you know starting a company is what you want to do, then go do it.
The piece of advice that I will give is that, the place that I don't see spend enough time is that just being really thoughtful around who they are and what they want to get out of life, and that's more life advice than startup advice in particular. I don't know, that's just my personality and who I am, but that's always something that I wanted to make sure of. One of the very influential books that I read before was, I don't know if you've ever read, Meditations by Marcus Aurelius, but a lot of his philosophy has resonated with me, just about how to live a good life. I'm all for spending a really really long time being thoughtful about what it is you want to do and why. I think too many people will just jump to the next thing every year, and only spend a week or a few weeks thinking about it without deeply reflecting on who they are and what their role is in life. If you think about how many years you spend over the course of your career, you're likely to work for like thirty, forty, or in some cases fifty years. That's a long time, not that your career is the only thing but it will end up being a big part of your identity as a person, so spending the time to be really thoughtful about that is... I see a lot of folks who really don't spend enough time on that and end up jumping from thing to thing, and are continually unhappy because they don't introspect at the level that I would for myself.
Exactly. If you start from what you want your life to be and what you want to get out of life as a human being, then you're much less likely to later on find yourself in a situation where you're working on something that ends up conflicting with your goals, and the kind of person that you want to be.
Exactly, one of the things that was cool that I realized, you know you always see movies or TV shows of these people who live crazy lives as doctors, or lawyers, or in the military, or business people, or stars, or whatever, and what I realized is you could literally choose to be any of them. You're not going to get there on day one, it's going to take many many years or decades in some cases, but you can choose to be any character you want, so being thoughtful about that is super important.
There's actually a really cool anecdote that if you don't mind I would love to share.
Go for it!
So one of the most inspiring stories to me when I was trying to figure out how to spend my life, was I was trying to read about people who had impacted the world in a really big way, and there was this one name that really stuck with me is this guy named Norman Borlaug. He was the Nobel Peace Prize winner in 1970, he was an agriculture expert and he started his career in the 50's and 60's, and what he wanted to do was bring best practices around farming and how to grow crops to the developing world. At the time, the population growth was exploding so we were at like three billion people, or something like that, and adding a billion every decade or two. The rate at which people would grow was really really fast, but the problem was that if you looked at how much food was available to feed those people it was actually limited. There was not enough farmland, just in the entire world, to feed everyone and to feed the growing population, so you kind of had this exponential growth curve of people but then you had this limited food supply, and it's like, what's going to happen? There is a bunch of famous books written about this exact problem, one famous one is called Limits to Growth, and it just talked about how the future would be a really really bleak one, and how that we'd run out of resources, there would be mass starvation, and disease, and famine, and war, and the few people left would be fighting over what limited resources they have.
Anyway, that was kind of the prevailing thought at the time and the only way to combat that was to enact huge population controls, so limiting how many kids people could have, just really awful things, and what ended up happening, as we know today, we have 7-8 billion around and by and large people are fed. There are definitely folks who still go hungry, but those are problems more of distributions then of just the total amount of food available. The reason that we're able to sustain the population that we have is because of largely the efforts of this one guy, Norman Borlaug, so what he did was he went down in the late 50's, to Mexico to do a bunch of research on how to bring modern farming techniques, and he created this special strain of wheat that ended up having a much higher yield. It had a bunch of characteristics, it was disease resistant, it was high yield, it had this attribute called dwarfism which is important for crops, and as a result Mexico's wheat farming quadrupled over the course of a few years and they went from a food importer, they constantly had to import food from the United States to feed their growing population, to a food exporter. That was huge, now it wasn't a worry. Food stability, and production, and being able to feed the growing population wasn't a concern. Then a few years after that he went to India and did the exact same thing there, and India was even in worse shape than Mexico at the time. India had just gotten independence and it split with Pakistan, you had this massive growing population with a super poor country, just a recipe for disaster in so many different ways, but he did the same thing and as a result India became self-sufficient in terms of food growth, and it was kind of 2.5x in terms of the overall ability to produce food. Then after that he went to Pakistan, he went to Africa, he went to a few other places to create new strains of crops and bring modern farming techniques, and as a result if you look at the net impact of what he's done, it's super clear that he saved probably hundreds of millions of lives, if not a billion lives. Now for me, that's real impact!
We talk about making a billion dollars here in Silicon Valley, you know what real impact is? Saving a billion lives. That story really stuck with me and resonated with me, and you know, he got the Nobel Peace Prize in 1970, but not many people understand the impact he has to hundreds of millions of people, but even more, our trajectory as the human race, and it's like, woah, if I can have a tiny fraction of the impact that someone like that has had on the world, that would be incredible. It was like, hey sign me up, I'll dedicate the rest of my life, no question, and I will be totally confident in that, and that's really important because I think a lot of folks go through life and they end up realizing what they want is very different from what they're doing, so being able to find something where you can say, hey let me dedicate my life to this without having any doubts about that, or any regrets about what you do is super important and a good way to find long term success. When I look at what we're doing here at Amplitude, we're nowhere near on the scale of an impact of someone like Norman Borlaug, but I think about my talents and what I'm good at, and it's like what do I know how to do? Well I know how to build and sell software, so let me do the most good as we can through that.
That's such an inspiring example, and also such a difficult person to compare yourself to when you're planning out your life. It's pretty rare, especially for people who haven't had any experience, to have a successful business. It's pretty rare to have a success right out of the gate, but the failures that you have early on can often teach you lessons at a deeper level then you would learn just by reading about what other people are doing, so in that spirit I would love to get a sense of what were some of the first companies that you started before Amplitude, and the projects that you worked on that may have ultimately failed to take off?
Yeah absolutely! I was just a kid starting out of college so I didn't know anything about anything, about starting companies. You know, how do we even pick something that's worthwhile to work on? I ended up convincing my co-founder Curtis, who is a really close friend of mine from MIT to leave his job at Google, and we said what should we work on? We worked on a bunch of different projects but one of the first real ones to get traction was this one called Sonalight. Sonalight Text by Voice, and it was a voice recognition application that helped you send and receive text messages by talking to your phone. The idea was that this was a safe way that you could interact with your phone while driving, because you didn't need to press any buttons or look at your screen. You could just have a conversation with your phone, and send and receive text messages. We launched right around the same time as Siri, which was a really good thing for us because a lot of people were looking for a similar type of experience on Android. That was the very very first idea that we worked on. We didn't know anything about voice recognition, we didn't know anything about mobile apps, we didn't know if this would be a good bat, a bad bat, you know, anything in between. We were just like hey, this doesn't exist, we think we can build it, it will help some people, so let's do that and take it from there.
We started that in the summer of 2011, six and a half years ago from now. We launched the product and it was awesome, we got thousands of downloads. We were really excited, we ended up getting into Y Combinator as a result of that and we were in the winter 2012 batch. Coming out of Y Combinator we had this really cool demo day presentation, where we'd show off the technology and I'd take my phone, I would be like a magician doing a trick to the audience, so I'd take my phone and I'd show it to the audience, then I'd put it in my jacket pocket, and then I'd be like, "Look, I'm talking to my phone and having a conversation with it without touching it at all, woah, this is the coolest thing ever!" This is before Amazon Alexa and Google Home, and all that stuff. It was a pretty cool demo and the entire room was like "Woah this is so cool!" and Curtis and I felt on top of the world, and we were really really jazzed about doing this really cool demo, and we have this really cool piece of technology, and on top of that we had the most press mentions coming out of demo day of any startup in that batch, out of sixty startups, and there were some really good companies in that batch. We were like, woah this is awesome, we're the best! We fulfilled our dream of starting a company, like nothing could go wrong.
We did it.
Exactly, we did it! I'm not still doing Sonalight today so it looks like things ended up going down a very different direction. Coming out of demo day, we had a lot of press which led to even more downloads, and more users, but what happened really quickly is users would not stick around in the product. They would use it, it'd be a cool tech demo, it'd be a fun piece of technology, but they would not integrate it into their day to day lives. Very, very few people would actually continue to engage with the app for more than a month. As a result, after all that press and all the users, our user base actually started shrinking. We looked at this and we were in panic, we were like "Oh my God, what's happening!?" like I thought we were on this exponential curve and it's no longer happening, what's going on here. Because we were engineers we spent a lot of the time digging into the data and we really wanted to understand what was going on, and we ended up spending half our engineering effort because after coming off this big high from demo day, we were facing the death of the company in some sense, because users just wouldn't stick around.
I was like, oh man, am I going to be a failure? I had spent a year before this building startups up as this big thing in my head, and I thought it was going to be this really successful thing. My parents are worried I gave up a good job in finance and all my friends are doing really cool jobs at their companies, and getting paid a lot of money, and I'm not getting paid anything. I don't even have anything to show for it after a year of working on Sonalight, and I was just in this really hard place. I was kind of questioning what my purpose is and if this thing is for me. Anyway that was kind of going on and we ended up doing trying to understand why was it that users were not retaining. Unfortunately all the tools on the market at the time wouldn't do this out of the box for you, so we had to build our own systems to do that and we ended up finding out the best predictor of long term retention in the app was whether the voice recognition actually worked. As it turns out the voice recognition did not work that well, most users actually would have a number of failed recognition attempts within the application, or they'd accumulate a good number of failures over time, so after you got enough failures you wouldn't actually continue coming back to the app, because it's like, well I could try to have a conversation with my phone and send a text message but it gets it wrong so often it's not even worth it.
That was the core of our retention problem, then we looked into voice recognition and we said, it's going to be really hard for us, just two engineers, be able to improve the quality of the voice recognition, so we were like, this doesn't make sense as an idea to work on, let's go work on something else. That was a hard moment because all those emotions that I had talked about earlier kind of built up in that moment, it's like okay, we have one thing that didn't work out and a bunch of side projects even before that, that were pretty hit or miss, so would we even be able to find something that we would be successful about.
I read some of the things that you posted online about this period of learning, where you spent a year just trying to prepare yourself to become a founder and to decide if you really could be a founder, and one of your biggest takeaways you said, was that, almost all successful startups have this point about a year to a year and a half into them where things look hopeless, and the founders have probably just given up... was that this point for you guys?
Yeah absolutely, that was definitely it. We were lucky in that, I had spent a lot of time reading the stories of startups, so we recognized that we worked on this thing for a year, it didn't work out, that's okay, and we can still find success, and in fact the path to success will almost certainly have an event like this along the way, and that's okay. We ended up taking what we had learned from building analytics and we had showed it to a few people, and they had expressed interest. They weren't as interested in the voice recognition technology but they were actually pretty interested in the analytics, because they had the same problem in their product where they were really trying to understand what creates a successful product, and how can I understand how my users are interacting with it, that was a moment for us where we were like hey, I had saw a lot of other companies that had gone through this same sort of failure so the fact that we didn't have something that worked out, even after a year, was okay, so if we were willing to stick it out for longer then it'd be pretty likely that we could eventually find some sort of success, so that was kind of a big turning point for us saying, it's okay Sonalight didn't work out, let's go work on the next thing.
Yeah, I think it's so helpful to have that realization before you get to that point as well. Because I think if you end up running into this road bump, then you talk to people and they're like, "Oh don't worry, this happens to everybody." it's kind of small comfort. Versus, when you're preparing for that to be something that you run into and it happens, and you think okay, I was ready for this, it's just a matter of pushing through the admittedly very dark place that you're in, and finding the next thing to work on.
Courtland I totally agree with you, I think even outside of startups and a lot of businesses I see, even if you're starting a restaurant, it'll take a year of no success to actually get to the place where you have the potential shot at success later on. That is just kind of the barrier to entry in my mind.
I think it's interesting to think about the sort of dichotomy between, on one hand starting the exact product that you want to build and having a very specific mission, and on the other hand, kind of stumbling into a really successful and promising business idea, then sort of having to grow to actually care about it over time. Do you think you fit more into the latter category, since Amplitude was something that you guys really didn't start off building and you kind of stumbled upon it?
Yeah absolutely, definitely it was not super intentional. It's not like we had this grand vision for what product analytics would become from day one. We definitely have come to realize what the real opportunity is, and the way people build products today is pretty messed up in that there can be a much better way for them to understand how to build a great product, so that has come to us but it's taken us five years to get there and to figure that out. Really early on, we actually really focused on mobile analytics because we had saw mobile is growing a lot. The needs of people on mobile are pretty different from the needs of the analytics tools that came beforehand, so let's create something custom for that. It definitely took a while for that to morph and shape into what Amplitude is now, and now we have a vision of where we're going, and like, why we have the potential to have this massive impact, but especially at the time we were just like hey, this is an immediate problem that needs solving, let's take a stab at it and see where it goes from there.
So one of the interesting challenges with that is at the same time that you're trying to figure out what you want your company to be and what kind of impact that you want to have, you're also trying to solve the very practical problems of growing, and hiring, so what were the early days of Amplitude like when it was just you and your co-founder, and what kind of concerns were at the top of your mind back then?
So the biggest one that I think that we took away from Sonalight is we didn't spend enough time talking with customers. We did some user research studies where we had people use the application and see what worked and what didn't, but we didn't really spend time getting to know our customers. We spent probably 95% of our time building and engineering things, and we spent very little time interacting with customers. When we started Amplitude it was like, let's change that and let's flip that, so let's see if we can actually spend too much time talking to customers, because everyone says you should talk to customers more. Let's take that to the extreme, let's see if it's possible for us to talk to customers too much, so before we built anything we said, okay, for the first month let's set ourselves a goal of talking to thirty people who could potentially be customers of the software. Not build anything, not have anything to demo, not have any product, let's just get out there and talk to people who have this problem, and that was huge! That did a bunch of different things, one, it gave us much more conviction that if we built something it'd be useful, two, it really helped us understand what was it that frustrated people about existing solutions.
I remember back then the big name in mobile analytics was Flurry, and there were all sorts of issues with Flurry about the data being inaccurate and not being in real time, the sorts of reports that you able to get out of it being very limited. It convinced us and gave us a path to say hey, we know if we solve these problems that there will be people that will find this useful and be willing to pay us money. That was a big change, so we started off just talking to people before building anything. Once we did that, it was an iterative loop for about a year after that where we alternated between talking to prospective customers, and then we built what they asked, then we gave it to them for free, then we asked them what would make this really useful to you, so we went back and we built that, and we just kept doing that over and over again for about a year. We started to get a real understanding of analytics as a problem and how do you scale a system to massive data volumes, and how do you really build something that's useful for people.
The one mistake I would say we made is we didn't ask for money early enough, so we'd end up spending a lot of time building something for someone, where they didn't really value it. I think it was definitely great that we got out there and talked to a lot of prospective customers but the one really big mistake is we didn't ask for some amount of money. You don't have to ask for much, but just some amount of commitment from their end that they value it, even like $50 or $100/month would've been a really big deal, and that would have allowed us to focus in. The one big mistake is that we spent a lot of time building for people who didn't really have as much pain as we thought they did, they'd say they had the pain but when it came time to see how much you're willing to pay to fix this pain, well nothing at all, okay well we probably shouldn't build for you, let's go spend time talking to other people. I think we were so afraid that nobody would want our product that we weren't willing to say no to people we should have said no to, so that was the one mistake. You know, we probably could have gotten through that phase faster if we had got in front of that but it was fine, we ended up doing this really good loop of building, giving that to more people, taking their feedback, and building more, and so on, until we had something valuable as an analytics product.
All of the insights that you just said are so valuable and so important to have, and it goes back to what I was saying earlier which is, that you can spend a lot of time reading stuff and I totally advocate, I'm analytical as well. I think you should read things and you should be prepared for what's to come, but there's something about actually doing it... like it's so easy to hear people say, "Talk to your customers!" and you think, yeah yeah yeah, that's great advice. Then you don't do it, or it's so easy to have retention issues and hear about retention issues, and not really worry about them, but once you've actually experienced those issues and once you actually experience the benefits of talking to customers, then it's so ingrained into your soul at that point, that you just don't make that mistake again, you become a lot more resistant to it. So I think learning from experience, this is the part of your story that really shows how valuable that is.
Yeah, absolutely. There's definitely a lot that we got wrong, firstly choosing Sonalight as an idea and going through the ups and downs of that. Then, we could have definitely could have been a lot faster in starting up Amplitude but there was no real way for us to understand that, and internalize that, and appreciate that without having to go through that as a failure. If you just told me at the time, "Hey, you should spend 50% of your time selling the product and asking money from customers." you know, I wouldn't have even known how to do that, and it was going through the pain of not knowing how to do that, that eventually led me to really step back and see what it would take to fix this as an issue.
I think what's also cool is that, like, you guys were so conscious about understanding where your weaknesses were and sort of almost over correcting for them. When I started Indie Hackers for example, I had a very similar situation where I looked at everything I had done in the past and thought, "Jeez Courtland, you spend way too much time coding and you can get so sucked into the coding that you don't do anything else." so I'm only allowing myself to work on an idea that involves very little code, so that I won't even be able to fall prey to that problem.
That's absolutely right.
Yeah, I think just having the ability to look at what you've done wrong in the past, and of course that requires having past failures and experiences, then having a discipline to correct for them is extremely underrated. One thing you mentioned was that you were going through this loop for about a year where you would talk to customers and sort of build these things for free, and see how they reacted... were you consciously going through this loop, or was it just something you sort of ended up doing?
It was definitely a conscious decision... so the thing about great engineers is that, if you're an engineer your natural instinct is to solve a problem, so if someone says, "Hey, I have a problem doing this." you're always like, hey I can fix that for you, let me build something that can help solve that. That part is easy and all you have to do as an engineer to add onto that is like, go talk to customers. The only asterisks I would have added in this case is, ask them for money as well. You don't have to ask them for a lot but just ask them for something, and just alternate between those two things. If you keep doing those two core things really well, if you understand what people's problems are and then build something to solve them, and you do both of those things really well, then, yeah that is the core motion of success for any software business. Anything else that people think is starting a company doesn't really matter. You don't really need to market yourself, you don't need to have a great product strategy, or the perfect messaging, or hey I need to sort out all these things on my finances, it's like no no no, all you have to do is build the product, sell it... build it, sell it, build it, sell it, that's it. When you're very very early on that's the core motion that will lead to building a successful company.
That's such a simple model... an engine that will run your company and I think it will appeal to a lot of developers out there who don't necessarily want to spend a lot of time doing other things that might end up being unnecessary for them to learn.
Yeah, if you're an engineer the only thing is, get in front of customers. If you can ask them for money, I definitely wish I had, but at least getting in front of customers is a million times better than not.
So how were you guys funding yourselves during this entire period of exploration and building, were you still living off of the money you got from YC the year before or did you guys raise additional funding?
We raised a tiny amount of additional, but it was not much, it was like $70k of additional funding. Curtis and I were lucky in that we had saved up a bunch of money from when he was at Google for a year, and I was in finance for a year, and we both managed to save up some money from those experiences so we were able to live on that, plus the money that YC had given us. In fact, we didn't even pay ourselves salaries for about the first two or three years, and it wasn't even that much when we got to that point. We had just saved up the money from the jobs that we had, and both Curtis and I were relatively frugal people. We were just very cheap on the apartment, and on our lifestyles, and on everything, and we were both very lucky not to have any debt from student loans or anything else, so as long as we didn't spend a lot of money we could sustain ourselves just based on what we had saved up.
Yeah that's such an important thing to do, because you just give yourself a bunch of extra chances to succeed that you wouldn't otherwise have if you were eating through money so fast that you had to quit in six months, or three months.
That would be my one piece of advice, if anyone feels really passionate about starting a company and wants to do it, I would say do it. The one exception I would say you should probably strongly consider not doing it, is if you have a lot of debt for whatever reason or you're not able to live. You know, you have debt, or bills, or whatever else you have to pay off, so that's the one exception to the rule of start a company if you want to start a company.
That's one of the cool things about being a programmer too, and why a lot of programmers end up becoming tech founders, in addition to to obvious reasons, because if things don't work out you've got a solid skill set that you can always fall back to.
Absolutely, especially as an engineer. You have no reason to worry about not being able to find a job or find opportunity.
So one of the cool things about your story and the success that you've had with Amplitude, is that analytics is an extremely crowded space. You mentioned this earlier that it seems like there's new analytics tools popping up all the time, and even on this podcast I've talked to Josh Pigford of Baremetrics, I've talked to the guys behind Segment, I've talked to David Darmanin at Hotjar, and you at Amplitude, and you guys are all doing extremely well despite the fierce competition. Why do you think that is, and how worried were you about your competitors early on when you were first starting?
So we're not, we're always paying attention, but for us as long as we're solving problems for our customers you don't really have anything to worry about for competition. Stepping back, I think one of the mistakes people make is they use pre existing terms to bucket everything into a market. People think of it as a market for analytics, when that's actually a really deceiving way to looking at what's going on because analytics is just a tool to help you accomplish something else. No one gets out of bed super excited to work on analytics for the sake of analytics. Once you start thinking about how and why is the analytics used, there's actually very few people that we compete directly with, so within the realm of product analytics, when you think about helping product teams build a better product, we think of it, while there are hundreds of companies in analytics, we think of ourselves as one of the only ones out there that does product analytics. Yeah, with analytics, there's hundreds of companies... solving the problem of the product team there's very, very few. If you can understand, like hey, I'm solving this person's problem in this very specific way that other folks are not and they're coming to me because I'm the best thing for them, yeah you just don't worry about the competition.
I think of Slack, and Slack is a great example of this, if you think about messaging tools there are hundreds, or thousands, or probably even more than that of messaging tools and applications out there. Whether it's email, or text message, or things like Whatsapp, or chat applications, or HipChat, there are hundreds or thousands of messaging tools out there, but that's not the problem that Slack solves. Slack is really focused on helping teams collaborate better and they think that problem hasn't been solved, and as a result they found a niche within that for themselves that they've been the best at and most successful at. In the same way we found product analytics for ourselves and we feel we've been the most successful at that. Looking at it in terms of what's the market, that doesn't make any sense at all, think about what problem am I solving for who, is there a group of people out there who I can solve a problem for and as long as there is, you don't have to worry about competition or anything else, just whether you can solve that problem for that group of users.
I think that's such a good way to look at it. There's so many early stage founders who are motivated and excited to start something new but they're not sure what to work on, and I think the default thing is for people to say, "Okay, let me look at something that doesn't exist at all. I need to do something that's completely unique and unheard of." whereas when you look companies that are more successful, a lot of them do fall into these buckets where it seems like they have a ton of competitors and it seems like they're not all that unique, but they're actually providing a lot of value to specific people.
Yeah, the mistake is that you're looking at it, and we had this with Sonalight, I was like hey, I have to do something totally new and novel, and the way to look at it is instead, hey, I have to solve a problem for someone that is not solved for them right now. If you look at it from that perspective, then yeah, people's wants Courtland, are unlimited. Humans will always want more stuff, so as long as you can solve a problem that they have that hasn't been solved for them, then you're in great shape.
Yeah, and it goes back to the story you told about the guy who kind of revolutionized farming and helped so many people. Farming wasn't exactly new, it wasn't like he was inventing something that no had ever even conceived of before. He was taking things that worked in one place, and bringing them to other people who want something similar.
Exactly, he solved the problem of hey, we don't have enough food in food production within the developing world. That's the problem he solved. A lot of the farming techniques he used in terms of creating crops that were high yield, and robust, and disease resistant, had already existed within the United States and Western Europe, and a few other places, but he was taking those solutions to places that hadn't solved that problem yet, so as long as you're solving a problem for somebody somewhere, that's all that matters.
So let's talk about analytics for a little bit, and I know you're biased because obviously you run Amplitude, but what's your philosophy on how people should use analytics tools and get the most out of them, especially early stage companies who might not be in a place where they have a product that's even worth making incremental improvements to?
So I'd say the first thing is before we even talk about analytics or anything else, the first thing is make sure you're talking to your customers. That's even something we still do today here at Amplitude. If you look at how our own product team spends it's time, they actually spend way more time talking and interfacing with our customers and prospective customers, than they do in analytics. We definitely use analytics quite a bit internally but we spend even more time talking with customers, so if you're going to choose only one thing to do when you're early on... talk with and interface with your customers.
I just want to interrupt here for a second to let people know, because I know we keep saying talk to customers but for some people it's not obvious why you need to do that. My take has always been as a founder, or a creator, or a product person, you don't know exactly what people find valuable and it's very difficult for you on your own to just sort of magically come up with all of the answers, so when you talk to people you're actually learning what they value, so you prevent yourself from spending months or years headed in the wrong direction because you're focused on building your hypothesis for what it is your customers want, as opposed to the reality of what they want.
Courtland, I'm glad you said that because this is actually really important for those listening and who are wondering why people keeping saying, talk to your customers. The really important thing to understand, the difference between when you're at another job, or when you're at school, working on things is that in the world of startups and business, the problems are not well defined, so defining the right problem is a job and super important in itself. We have this saying here at Amplitude, which is that product's job is not actually to come up with a solution. Product's job is to define the right problem. Recognizing that that's a hard thing in itself, defining the right problem, I think the false expectation that a lot of people have is that they'll be handed the ideal problem down from the heavens, and the perfect definition, and know exactly what it is to build perfectly for all things.
The reality is you don't, you don't know at all. You don't know what problem you should solve. You're never going to know perfectly what problem to solve, so viewing that as a thing you always have to invest in, just like hey, you always have to continue to improve your product and ship more code, and make improvements, it's like you always have to update your understanding of the problem and really invest behind that as a thing in itself. Just knowing the right problem to solve is massively valuable, and basically avoids all the situations you were talking about. If you don't know what the right problem to solve is, it's like you solved an incredibly hard math problem but for a test that didn't matter, or there wasn't even a test around it, or you don't get credit for it, so you only get credit for solving problems that people actually have and those are unknown by default, you just don't know what they are, so viewing things as hey, I always need to spend time updating my understanding and definition of the problem I'm solving as a task in itself is super important.
I would say spend half your time doing that, that's what talking to customers is. That's what sales is, sales is uncovering the problem of the person you're talking to, you don't know that by default. Split your time between the two. It sounds crazy to spend half your time defining the problem, but that's actually how you should think about it. Spend 50% of your time talking to customers to update your understanding of the problem and define the right problem, and spend 50% of your time solving it, because the worst is you define a crappy problem nobody needs solved and you spend 95% of your time on the wrong problem. That would be the challenge I have to anyone who's starting a company from an engineering background is to spend half their time defining and understanding the problem they're solving.
I think that's such a great way to look at it, and I liked that you mentioned the misleading experience that you get by being employed at a bigger company or being a student where...
...you're handed well defined problems.
Exactly, you're handed these well defined problems and there's this entire layer that exists between you, the programmer who's implementing the solution, and the people who are actually talking to customers, and understanding what problems customers have, and it's easy to fool yourself into thinking you understand that.
That's exactly right.
So first of all, companies need to understand their customers and talk to their customers, but what else can they get out of analytics tools because I think the most common thing that I hear from founders, especially the founders of early stage companies, is that they set up their Google analytics, they set up Segment, they set up Amplitude, they set up whatever tool they're using, Mixpanel, and then they don't know what questions to ask, or they just kind of look at the metrics and they don't really change how they're building their companies. Is there such a thing as being too early stage for analytics or are they just using the tools wrong?
I'd say, if you're already talking to customers and you have a few thousand users or more, like on a daily, weekly, or monthly basis, then that's the right time for you to start using analytics. If you're earlier than that, if you only have a few hundred users, or if you're not talking to your customers yet, don't use analytics. This is coming from an analytics CEO, so you know that even with my biases that's what I'd say. That's the first thing, now once you get beyond a few thousand users it becomes very impractical to interface, and interact, and talk with them all directly, so that's where analytics comes in and is very very powerful. The way I'd look at it is if you don't know what to track, or how to track it, or how to think about it, the first thing you want to do is focus on your retention. A lot of people focus instead on their daily active users, but the problem is that's the output. The input to that is your product and how good your product is, and the best measure of that is the retention of your customer base. The way to view your users is, someone has taken a lot of time and gone out their way, and out of all the things in the universe they landed on your thing, and they found your thing to solve their problem, or what they think solves their problem, and they started trying it out, and your ability to convince those people you can solve their problems and they keep coming back is the best indicator of whether you're actually building a good product and solving people's problems.
The first thing I would focus on once you get to the stage where you can actually start thinking about analytics is understanding whether your users are coming back and retaining. If very few of your users are then that's a problem. Benchmarks vary by industry and are all over the place on this one, but if you want a very simple baseline, the standard baseline that I use with folks is that after the first day, after using it for a day, 40% of your new users should come back the next day, then 10% of those should still be around by day seven. If you're hitting those benchmarks, then you have a good product, but if you're below those then you should be much more introspective and ask yourself, why is it people are not coming back enough and spend time understanding that. There's a bunch of ways to understand that from an analytics perspective, with Amplitude, what you'd want to look at is what is it that people who retain have in common and what is that people who don't retain have in common. Typically what you want to look at what features people use and what features people who retain use versus those who don't retain, that will tell you the difference. That will tell you, hey, these features are really valuable and I should emphasize them in my product, and these features are really not.
So a quick example of that is there was this mobile app named Calm, one of our customers, who was an early stage startup, they've since grown quite a bit, and they wanted to understand this exact same thing. They wanted to understand, how can I improve my retention. They looked at a bunch of different features that people used and they found a really interesting feature. They found this one feature that if you used it, you were three times as likely to retain long term as if you didn't use it, for the baseline. The interesting thing is that only 1% of their user base actually used the feature. What the feature was, it was a reminder, so Calm is a meditation app that helps you achieve mindfulness and it walks you through a bunch of meditation sessions, and the feature that these highly retained users used was a reminder to meditate at the right time. The really interesting thing is that they had buried this behind three or four different screens within the application, and when they saw that data they realized that this reminding people to meditate at the right time is actually as important, if not more important than the entire rest of the application that actually guides you through meditations, so they took that feature, they brought it to the front of their application, and as a result they massively increased their retention. That was huge, because they created a better product from understanding what features were really valuable to people, so that's a really small example of the process of how to go about using analytics, and product analytics to create a better product.
I'm really curious about those numbers that you got, 40% retention by day two and 10% by day seven, did those numbers come out of your personal history building products or is it something that you've noticed and analyzed in working with Amplitude customers, and trying to help them?
It's actually been around before Amplitude has existed, this came from the gaming industry which has been a pioneer with analytics. If you look at some of the stuff, let's say Zynga did in 2011 or 2012, they were very much thinking about all of these different things. It's been since reinforced by a lot of what we've seen with the products that we work with, and if that you have less retention then that you really want to focus on increasing that and nothing else, but if you have more than that, that's when other things can become higher priorities. That's just kind of one of these analytics industry standard benchmarks that we've seen and we've found to be true.
So we're running toward the end, I've got a couple more questions for you if you've got time. One of the first is that, I saw an interesting talk last year by Gale Goodman, who runs this company called Constant Contact, and her talk was the long, slow SaaS ramp of death. She was basically talking about how at her company, they constantly had to find new growth channels because whenever one started working too well, they would eventually exhaust it and hit the point of diminishing returns. Has that been true of you guys at Amplitude? Because you guys have grown so much, the number of products you have sending you data and employees you have at your company, you have to imagine that you've been through several stages, or has it been just kind of one smooth, uninterrupted path of growth for you guys?
It definitely has been, where we've had to find new growth channels. One of the things is that we're less sophisticated than you might think at finding channels for us to grow, that's still something we're trying to understand. We've had really good success with two things in particular over the last year or two. One of them is just reaching out directly to folks, so reaching out to folks who have product analytics as a problem and just sending them an email, just saying, hey are you having any of these issues, if so we'd love to talk more. Then the other one that we found to be helpful is what we call product analytics summits, where we spend a bunch of time hosting an event that teaches people the basics about product analytics, and what product analytics is, and some examples of how to do it, so those have been the two that have been most successful for us in recent history. I definitely think it's something that I'm not an expert at, I came from the build and sell, those are the two things that I know how to do really well so that's something that we're still developing our good ability to do, but it's absolutely true that the way you get your first ten customers can be pretty different from the way you get the next hundred, from the next thousand, an so on.
Especially as you get bigger, there's no playbook for exactly how you're going to grow your particular company and acquire the users that you want, and you always have to be experimenting and trying new things.
One of the comments I do want to say on this is, one of the funny things is, I don't know if you know Dropbox has this referral program where if you refer someone you get more space, I don't know if you're familiar with that...
Yeah it's super famous, everyone tried to emulate it for years.
Exactly, so everyone's tried to emulate that and I don't think I've heard of a single other instance where that has worked. My big takeaway, I was at a panel of Y Cominator founders a few months back, and people asked the question that you did about what channels are successful and everyone said they tried out the Dropbox referral program, and it didn't work for their product. My takeaway from that is the way that you reach your audience is just so different by industry, it's good to know what other people do but it's impossible to take what has worked for another company, and customer, and user base, and apply it to your company, so you just need to figure out what are effective ways to get in front of people who have the problem that I solve, and the channels that you use for that are very very different. Don't try to be Dropbox, don't try to be Amplitude, just try to be you.
Sage advice, be you. I usually end these episodes by asking you to give advice to early stage founders but I think you've given a ton of advice so let's do something different here. I'll ask you to tell me a story, but not a real story, what I want you to do is pretend that you are Spenser Skates with the ideal version of yourself ten years from now, in 2027, what is your life like?
Ooh, that's a really hard one! That's an interesting question to think about what things were like ten years ago and that's something I care a lot about getting right. It's so hard to look out, I can barely tell you what's going to happen next year, never mind a few years out, never mind ten years out...
...what I mean is, who do you want to be in ten years?
In terms of who I want to be, I'd love Amplitude to go on, to continue to be successful, we continue to grow, we continue to solve this problem of helping people build better products. I think there are a lot really interesting things way beyond analytics that can help people build better products, that we have the potential to do, and we have the opportunity to do if we're successful, so I would love for us to get the opportunity to tackle some of those. I think we have to get the product analytics problem solved first. No matter what happens to Amplitude it would be awesome if in ten years from now, Amplitude and me as a part of it, could still be around and we're still successfully helping other companies... if I do nothing else in my life but help companies build better products I think that's awesome. I would say I definitely don't know a lot, there are a few things I know about myself for sure in ten years, and what I'm going to be like. I know I'm going to really care about making the world a better place and helping other people, I know that that will probably still be true. I know I'll still be motivated, other than that it's really hard to say, it's really hard to understand what I'm going to be like in ten years and what the Spenser of 2027 is going to want, but at least I'm hopeful that if I can continue making Amplitude a success, and making everyone here at Amplitude a success, and making all our customers successful, that would be awesome.
Well I wish you the best of luck and I hope you continue to make your customers successful. Can you tell listeners where they can go to learn more about you personally and the things that you're working on at Amplitude?
Yeah, so me personally, that's a good question. I've done some AMA's in the past so if you just Google me, I think I did one with a product community a few months back. If you just want to learn more about product analytics we have a ton on our website at Amplitude.com . I'm always super happy to help out folks in the early stage, so feel free to reach out to me direct. I'm happy to provide more prospective and advice, and that's one of the things I definitely owe to the community because there were a lot of folks that helped me when I was early stage, so I'm always more than happy to help out folks. Feel free to send me a note directly at [email protected] as well.
Alright, thanks so much for coming on the show Spenser!
Absolutely Courtland, I really appreciate talking to you. I love what you're doing with Indie Hackers, I'm glad there are most resources today for folks at the really early stage and I feel super happy to be a part of that.
If you enjoyed listening to this conversation and you want a really easy way to support the podcast, why don't you head over to iTunes and leave us a quick rating or even a review?
If you're looking for an easy way to get there, just go to IndieHackers.com/review and that should open up iTunes on your computer. I read pretty much all the reviews that you guys leave over there and they really help other people to discover the show, so thanks a ton for your support! In addition, if you are running an internet business or if it's something you'd like to do in the future, you should join me and whole bunch of other founders on the IndieHackers.com/forum . It's a great place to get help with pretty much any problem that you might encounter while growing your business, like how to come up with an idea or feedback on a product that you're working on. I try to spend a couple hours a day answering questions and giving people feedback, and getting to know everyone, so I really hope to see you there. That's IndieHackers.com/forum .
Thanks for listening and I'll see you next time!
Did you know Indie Hackers has a newsletter?
Sign up to get insights, takeaways, and exclusive content from each new episode, directly from the host, Courtland Allen.