Sean Byrnes sold Flurry to Yahoo and launched Outlier AI

likes
+ Add to

What do you do when you’ve sold the company that you’ve poured your blood, sweat and tears into? Launch another company, of course.

Sean Byrnes is the CEO & Co-Founder of Outlier.ai which delivers automated business analysis for some of the largest consumer businesses in the world.

Sean built Flurry into a successful analytics platform for mobile apps and sold it to Yahoo in 2014. He then launched Outlier.ai.

Sean Byrnes

Sean Byrnes

Outlier

Sean Byrnes is the CEO & Co-Founder of Outlier which delivers automated business analysis for some of the largest consumer businesses in the world.

roll-angle

Full Interview Transcript

Andrew: Hey there, freedom fighters. My name is Andrew Warner. I’m the founder of Mixergy where I interview entrepreneurs about how they built their businesses. I remember when my wife was working at Yahoo, she was doing their social impact. She was the head of Yahoo for Good. The big talk there was about how Yahoo was changing partially through these acquisitions of . . . Yeah, Flurry was a big one. And at the time, it felt to me like Flurry was this company that just hit it, that was so impressive that they sold for a big number but also they were going to change Yahoo through that acquisition.

And it wasn’t until I got prepped for this interview that I realized how many setbacks this analytics company had. I still feels unfathomable to me. Sean Byrnes is the founder. He’s here to talk about a little bit about that and a little bit also about this new company that he created after . . . I wonder if he burned out or what happened when he took some time away. The new company is called Outlier, again in analytics, again in data. This time he’s saying, “Look, big businesses have lots of data. They need someone, some company, some person to analyze it. We’re going to be the one to do it. outlier.ai is the website. They automate business analysis.

We’re going to find out about how these two great companies were created thanks to two phenomenal sponsors. The first one is one admittedly I’ve not done a great ad for. I love the company, I’ve got to get better at it. It’s Ahrefs. If you want to know what’s happening with links to your website and how to improve through SEO and otherwise, check out ahrefs.com. And then second is the company you use if you want to hire developers for AI and other things, it’s called Toptal. I’ll talk about those later. Sean, good to have you here.

Sean: Thanks for having me, man. I’m excited to talk about these things. What a coincidence? I turned out to be the founder of Flurry we’re talking about it today. That’s great coincidence.

Andrew: Wait. You are the founder of Flurry, aren’t you?

Sean: Absolutely.

Andrew: Yeah, I thought so. Hey. Do you remember the day that you sold?

Sean: I do, but selling as you know is not a day, it’s a process, it’s an adventure, it takes time. But absolutely. I definitely remember the entire thing. Honestly, I remember the entire process as it was yesterday and also the same time as if it was 50 years ago. When you start companies, each one is like a lifetime, and so I’ve been doing this new company for about four years now, and so in some ways, it feels like I’ve been working on it my whole life. In some ways, it feels like we sold Flurry just yesterday.

Andrew: I know that it must have taken a long time because there was even . . . Back when people were using Secret. It was even a Secret post saying, “Our company is being sold.” But I can’t say for how much or to who.

Sean: I remember that, yeah. That is absolutely the case.

Andrew: Was there a day when you can just like exhale and go, “All right. We did it.” Was there that day? Was it when you signed a contract? Was there anything like that?

Sean: It was about three years after the acquisition happened, I think, roughly. It’s just so hard to let go. And so you think . . . When I started Flurry it was 2005.

Andrew: Okay.

Sean: It was the early days of mobile. Your state of the art cell phone was a flip phone or a feature phone that you got for free usually with your cell phone plan. And mobile as a business in the U.S. really hadn’t arrived yet. And it took us nine years and a lot of ups and downs that we’ll get into to build three into what was one of the largest top five analytics and ad companies in the world. And then we sold it to Yahoo.

But I don’t know that you can sell a company like that nine years of your life and just let go the next day. That’s not how it works. All of my friends were at the company, everything we had invested in was there. It was really years later that I didn’t feel the weight of that entire effort anymore. And of course, by then I had this new company with this new weight on my shoulder. So, it’s the kind of thing where you’re either welcome to burden and enjoy that weight or you probably do something else.

Andrew: And so you welcomed that burden. You like it.

Sean: I do. I find that I’m the kind of person who needs pressure to do my best. I get bored easily. I like to do lots of different things, but I am at my best when there is a lot of pressure and there’s deadlines, and specifically a competitive environment where I can either win or lose. And in the days of Flurry, there were generations of competitors we had to face down. And the same thing is true with Outlier today. And that’s really when I go home and I want to be proud of the work that I do and enjoy what I’m doing I have to be in that environment. So that’s why I keep coming back. I mean, I could have walked away from entrepreneurship after selling Flurry, but I decided to come back because there was just no other pursuit I found that has this kind of focusing effort where you can do so many things and still be so focused on one kind of outcome, one kind of success.

Andrew: I want to hear about what launched Flurry because I think the way that you think about finding ideas is useful. But let me start with the end here. I see a LinkedIn post saying that you sold for about $300 million. I see a TechCrunch post saying that you sold for $200 million. What’s the number?

Sean: Oh, they don’t let me tell it. It was part of the deal. It’s a higher than all of those, though, if that helps you.

Andrew: Wow. Okay.

Sean: It’s surprising how underreported these acquisitions are, but it’s not surprising because the acquisition price is up to interpretation, right? So do you include the employee retention packages as part of the acquisition price? Do you include the holdbacks? Do you include the escrows? Do you include the equity value at the time of the acquisition or the equity value at the time of the vesting? These things are all very hazy, so it’s not surprising there’s lots of numbers reported, but it was actually higher than all those. It was a great outcome. Everybody did very well. I was very proud of them.

Andrew: Okay. It does give me a sense of what you built. You told our producer, “I believe in starting with the problem.” What’s the problem that you found that launched Flurry?

Sean: And a quick side note on why I start with problems. I think it’s easy to fall in love with the solution, to fall in love with the product you want to build or the thing you want to do. And the problem . . . That ends up being an issue because it’s inflexible. What if that product doesn’t work? What if customers change? But the thing about problems is problems as long as you choose them well, can be essentially a north star for you forever in that company. It’s always something you want to solve.

So with Flurry, interesting enough, we started out wanting to solve the simple problem of getting our email. So we . . . I graduated from grad school in 2001, 2005 I was still of the digital native generation, loved email, loved instant messaging. And we had these feature phones. One of the things that when you move from like an academic campus to the real world is that you go from a world where there were computers everywhere, internet connection everywhere, to having this feature phone in our pocket which did very little. It plays phone calls and maybe it sent text messages. And unless you bought a BlackBerry back then, you couldn’t get your email, you couldn’t get the news, nothing on your phone. You had to wait until you were in front of a computer and find an internet connection.

And so Flurry was started to solve this problem of connectivity, to bring email and news, these sorts of things to those devices we all had in our pockets, and specifically to do it in a way that hadn’t been done before. So, there were a lot of companies that would produce apps and services for those same feature phones, but you had to distribute them through Verizon or through AT&T, through their app stores. And consumers couldn’t go to those app stores, nobody did, but there you could go start them.

Flurry was started to be kind of a direct to consumer approach with the idea that in the internet, I don’t go to Verizon to get to Google, I go to Google. I don’t go to Verizon to get to Facebook, I just go to Facebook. And that would happen to mobile, and so Flurry was started to be the first of that new generation of mobile company where we would bring messaging, news to your feature device direct. You wouldn’t have to go through anybody else. You just go to flurry.com, we install our app, and you’re off to the races.

And it was, in some ways, an exciting problem because it was trying to usher in the world that we wanted it to be. But it was also challenging because you’re going in the face of these enormous behemoths in the form of the carriers and these systemic challenges that we had to overcome. But that was the problem we started with. And as like all great problems, it turned out we were solving something totally different than we thought we were.

Andrew: The problem you thought you were solving was, “How do I get email on my phone?” The problem you actually were solving is, “How do I get any kind of information on my small phone?”

Sean: Exactly. And so we ended up launching this kind of email service on your feature phone in the U.S. in 2005 and it was an abysmal failure. It was a huge failure. Nobody, frankly, in the U.S. wanted to check email on their phone. They didn’t think about their phone as a computer. They were thinking about it as a phone. It turned out, though, that you know, who found a lot of value in a direct to consumer email solution designed for feature phones was the rest of the world. So Flurry caught fire in Malaysia, Indonesia, India, the developing world because they didn’t have PCs at home. They would go to an internet cafe and they create an email account, then they go home, and if they wanted to check their email, they’d have to go back to the internet cafe.

With Flurry, because we were direct to consumer, you could be in Indonesia, and get on your phone, go to Flurry, download our service and be up and running with your email on your phone. So, Flurry became the first widely successful direct consumer mobile company because the developing world was so thirsty for solutions, but all of the incumbent large developers were going through Verizon, they were going through AT&T. Nobody was sending their team to integrate with the national carrier in Burma.

But since we were direct to consumer, it didn’t matter, so we were everywhere, and it became the start of what eventually became mobile 2.0 or direct consumer mobile apps, which still wasn’t the problem we end up solving. This was still not what we ended up doing because it turns out that one of the core problems there is once you’re successful mobile company in that world, there are no tools, there’s no support. Nobody’s ever done it before, right? Nobody’s ever figured out how do you track how many users you have? How do you make money with these mobile apps in these countries? And so we had to build a whole infrastructure around analytics and advertising to support what was becoming this large mobile consumer business.

Andrew: Were you doing this for yourself first?

Sean: Absolutely. We were building these tools for ourselves.

Andrew: By the way, as you’re talking, I’m kind of going back into the Internet Archive to get a sense of how it started, and I could see here it was Flurry Mail. There was a plus version of the site. Am I right?

Sean: That’s right.

Andrew: And then there’s something called Mobs, which I think is what you called your group section. Do I have that right? And then there’s . . .

Sean: Absolutely.

Andrew: And so you were offering all this on phones, and you wanted to know for yourself, who’s using what or what were you try to understand with this?

Sean: Exactly. The state of the art of metric back then if you were a company like Flurry with a very successful mobile business was how many downloads you had. Downloads was the only metric you had. So, imagine that you’re running this business and the only number you have to manage your business is how many people have downloaded your app?

Andrew: Okay.

Sean: Which is insane. You don’t know if they ever launched it, you don’t know if they use it, you don’t know if they deleted it. All you know is that they downloaded it.

Andrew: And this was an app. It wasn’t like WAP or something.

Sean: These were. Yeah. These were all apps downloaded on your phone.

Andrew: So if I had a Motorola flip phone, I could install an app back then?

Sean: Oh, yeah, absolutely.

Andrew: Got it. All right. And so I would install an app, and you would know that someone installed an app, but you wouldn’t know, did they use it? Did they prefer Flurry Mail or did they prefer Mobs? Is the feeds the hot part of our site or is it actually the feeds that nobody cares about and we should get rid of? Got it. And you said, “We need this for ourselves. Let’s create it for ourselves.”

Sean: Absolutely. That’s right because nothing out there. There was literally no infrastructure.

Andrew: By the way, at this point was this . . . Was the business successful at this point?

Sean: We were successful in the way that consumer venture-back businesses are successful and that we were growing fast but not making any money. So, the metric of growth and engagement was high, but we weren’t making any money.

Andrew: Okay.

Sean: And eventually, I think that what we realized in having built this infrastructure is I would go to mobile conferences, I would talk about Flurry and all the things we knew about how often people were checking their email on their phone. And all the other CEOs would come up to me afterwards and say, “How do you know that? How do you have those metrics? All I have is downloads.” And I would talk about these platforms we’ve developed. And I’m joking . . . I’m not even joking. They would say, “Listen, I need that. I will pay you whatever it costs to get access to it.”

And I’m a pretty thick-headed guy, but if somebody offers to pay me whatever cost for something, eventually, it gets through my thick skull and I’m like, “Huh, I wonder if there’s a bigger opportunity to become a platform ourselves.” And that was the decision we ended up making in 2008 to essentially get out of the direct-to-consumer mobile business ourselves and make our platform available for other developers. And it turned out to be good timing.

Andrew: How many sales did you wait to get before you switched to that?

Sean: I don’t know that the mobile app business ever generated a lot of revenue at all, frankly, because most of our . . . We were . . .

Andrew: I mean, how sure did you get before you said, “You know what? We should give up this thing that we’d been based on to jump into the new one”? What did you do to say that, “This is going to work,” before you did it?

Sean: I would love to tell you it was a very scientific process and very well-thought through. I will tell you what I had personally was in very strong conviction that there was a bigger opportunity than just building a successful app that there was a big sea change happening. We were watching people change fundamentally how they used their phones and we just believed very strongly that this was a behavioral change that will go wider, and the idea that everybody would be doing it.

If you are on the inside of a successful consumer internet company, and obviously, you’ve been here, one of the things you always worry about is consumers are fickle. What happens when the other shoe drops? What happens when you’re no longer considered the hot app, it’s something else that’s come out? And for me, the question was, “Do you keep investing in this your own apps or can you make a bigger business in becoming a platform?” And the belief was, we can make a bigger business out of it.

Now, this decision was less scientific because if you remember the iPhone had come out in 2007, and Steve Jobs is on stage at this iPhone announcement saying it will never run apps. It’s all going to run the web. And here is Flurry which is a mobile app company moving to being a mobile app platform for native apps in the face of one of the great visionaries of our time saying that mobile apps are dead, long live the mobile web. My investors, one of my investors told me, she was like, “Listen, you’ll never raise another dime of venture capital money again if you make this move. It’s not a good idea.”

And the reality was I think that we just had faith that there was a bigger business there. It was hard to see it from the outside. A lot of it was us watching and seeing how fundamentally we’re changing consumer behavior. And something that made it look like a genius move was right after we decided to make that shift, Apple launched the app store, and so they gave in. They decided to have native apps. And so there was Flurry at ground zero of the app revolution with a platform built for mobile apps refined over three years of using it ourselves. And Apple decided to like pivot right into where we were and so did Google thereafter, and all of a sudden, things exploded and there really weren’t a lot of other solutions. And starting at that point in the fall of 2008, Flurry as a business doubled every six months for the next six years.

Andrew: Wow. Wow-wee. Yeah. I see here the App Store launched July 10th, 2008. I’m looking at your 2009 website, not only do I see the iPhone on there, but I also see Google’s answer to the iPhone which was their first Android phone with the keyboard that was on T-Mobile back before T-Mobile was a real competitor. And the thing that I see come across beyond the interesting old phones is completely free, completely free throughout you said that. What was . . . as a guy who was being offered lots of money and switched because finally, someone was offering to pay, why did you keep featuring free, free, free and where was the money?

Sean: So, when we made this pivot, one of the things we did was think through the future. Right? Okay. So we were moving from a consumer business to a platform. How do we think it’s going to develop in the coming years? This is a big bet. Let’s try to understand it. And what we did was we looked at the internet, the web, and we realized that web analytics was gravitating to lower and lower prices. Google Analytics was out there at that point, it was driving down the cost of analytics on websites. And we realized that if there was going to be a price where a race to the bottom, the best thing to do is start out free. And you can’t commoditize free. If we started out with a free analytics platform, there was no pressure on the business. We could roll it out and grow.

The other thing that we noticed was while there were some big players out there willing to pay a lot of money, most people hadn’t really figured out how much money they were going to make from mobile yet. They didn’t know how much . . . what their lifetime value of a customer was, they didn’t know what their business model was. And so it was hard to budget for tools like Flurry because they weren’t quite sure how much it was worth to them because they didn’t know how much the business was worth. And so starting out with analytics being free meant that there was no barrier to adoption, that there was no way to commoditize out from under us. And in the early days of the market that’s what you needed to be successful.

At the same time, though, then how do you make money? And so when we started, even though we didn’t launch our ads business till years later when we made that pivot, that was the plan was to use the data from all these analytics to power ad network on mobile that would be more intelligent, more accurate because of the data. And years later, we followed up with that.

Andrew: I see here also in that early website, “Flurry’s platform supports millions of application downloads across 1,200 devices in over 200 countries around the world.” We’re talking then beyond the iPhone and beyond Google. Am I right?

Sean: Yeah. As I said, we started out as a feature phone platform, the Nokias and Motorolas of the world had enormously large number of devices in enormously large number of places. And one of the early attractive aspects of the Flurry platform was that wherever your apps were, we could help support them because, again, we had built this for ourselves over three years where we had to do the same. So, it ended up being a powerful . . . The powerful story then was, this is for mobile developers, by mobile developers. We’ve lived your life, we’ve been through those wars, we know what you need and we built this for you.

Andrew: All right. I’ll talk about my first sponsor. Also, a data company called . . . I thought they were called ahrefs.com. I see other people just call them hrefs. It’s ahrefs.com. I actually search on your site there to get a sense, like, where are you getting your traffic? Where are you getting your domain credibility? And what I discovered was if I was like a competitor of yours and I wanted to write great content and I wanted to reach out to other people to link to me, I realized, you guys are really big on getting traffic from articles about artificial intelligence, so Forbes’ prediction for artificial intelligence for 2019, “13 Industries Soon To Be Revolutionized By Artificial Intelligence,” and so on.

So if I was also in artificial intelligence space and wanted somebody to write about me, I would see all the people who are writing about you who have a really high domain rank, I just reach out to them and say, “We’re also in this space.” Or maybe if I was . . . Actually, that’s the easy one, easy win. Then I get to see if I can keep clicking over here. And this is so insightful to see what’s going on with other people to get a sense of which broken links you have on your site. So if I wanted traffic from you, I might say, “Well, look at this. This guy is actually sending a link to a couple of broken sites. I should just reach out to him and say, “My site can do what’s on your site, except it’s broken on your site. I mean, the links that you’re sending out to my competitors are broken. Link over to me.” I can also see what keywords are sending traffic to you.

This thing is so beautiful the way that this site works. I’m underselling it because I am not as knowledgeable about how to use this data to create content based on data, and I’m learning more and more as I use Ahrefs. I imagine if somebody was using it a lot would know even more about how to do this. So here’s what I suggest, go to ahrefs.com, sign up for their seven-day trial period for like seven bucks. Just look through your site and look through a few other competitors’ sites.

I think the basic easy wins are going to stand out, like, who’s linking to your competitors and also has broken links on their site that could be . . . you tell them, “You should be linking to me.” Where are your competitors getting their traffic? What keywords? What writers who should also be writing about you? And then do what I did. Go to their training, which I got some free training from them and learn how to keep improving your content in a way that will get you more traffic. It’s ahrefs.com, ahrefs.com. They’re not even giving us a /mixergy at the end. They’re keeping it super simple. Do know them, by the way?

Sean: I have not used it before. No, but it sounds like a valuable service.

Andrew: They’re getting so . . . Even their competitors are raving about how good they’re doing. I’m trying to fly . . . I’m flying to Singapore, I want to interview the founder. I don’t think he’s going to see me. I think he’s trying to be super secretive. I want to find him. When you started doing the ad business, how did that go for you?

Sean: Well, the advantage of having a new platform is that it levels the playing field, all the incumbents, their advantages are negated when a new platform comes out. That could be smartphones, it could be the internet. It’s the best time to be an entrepreneur is when you have this kind of weird inflection point in technology with a new platform . . .

Andrew: Before Google Analytics comes in, there is no Google Analytics yet, you are not [doing 00:20:47] that.

Sean: Then there’s nothing. Frankly, it’s not just that. All of the existing large websites had no apps, and so two people in a garage building apps had just the same chance of getting adopted as the big companies. And so what ended up happening was that there was such a hunger and a thirst for advertising on mobile to grow to try to reach audiences and there really weren’t a lot of solutions. And at that point, Flurry analytics was so widespread because it was free, because it was the industry-leading product that launching the ads platform wasn’t that hard. And in fact, we never needed a lot of the customers of the analytics service, the free customers to convert to our ads platform. In fact, I think we only ever had 10% of that audience, but the audience was so large that 10% of those apps was enormous. And so it took off very quickly. And in fact, the problem never was demand, it was always supply. No matter how much supply you had in terms of how many ads you can show and what you can promote . . .

Andrew: Really?

Sean: . . . there was always more buyers who wanted to spend more. Oh, that was the great problem of that time was that mobile was growing so fast. You imagine how fast [inaudible 00:21:54].

Andrew: I’m looking at articles about this. Yeah, yeah. I remember that it was growing fast. I remember . . . I’m looking at TechCrunch old articles about you. I remember Michael Arrington saying even when the iPhone just started launching apps, if you’re not a mobile company, if you’re missing that, then . . . I forget how he made fun of people who were missing it. I see that. So then when was the challenge? When was the period then where you told our producer that you guys got closed . . . What was it? You were insolvent three times? You were basically earning minimum wage for years?

Sean: Oh, yeah, absolutely. Let’s go back in time a little bit. So when we were mobile app company, it was difficult back then, you’re talking about 2005, 2006, 2007. You’re in the rubble of a dot-com bubble having burst, right? Venture capital investment was hard. And we were a consumer mobile company and we weren’t making money. And while we were successful in that regard, if you don’t generate cash flow through your business, you have to have investment. And so it was during that period of trying to build that business, prove that mobile was a thing that we struggled a lot.

But even after in 2008 when we made that pivot, the App Store, as you mentioned, came out in July of 2008. The other thing that happened in 2008 was a financial crisis, right? Higher market just exploded overnight. I think two-thirds of all the startup companies I knew went out of business in a six to nine-month period because people stopped buying, people stop spending, everything contracted. Investors stopped investing. We were lucky in that Flurry had always been very capital efficient.

So while we hadn’t made money, we’d had an ethos of being very conservative in what we spend. So we were the most capital efficient company in the portfolio of our investors. So when we needed more money to keep growing to keep up with how fast the business was going, our investors were willing to invest in us, our inside round from internal investors because nobody was investing in new companies during that period, and so we survived that period by bringing in outside investment. But in the middle, you have these periods between when you need the money and then arrives that don’t materialize. And so, yeah. There was a lot of periods where we were insolvent three times, minimum wage. I mean, I didn’t paid for the first year and a half, two years of the company being in business.

Andrew: Why did you continue to do that? You were a guy who worked as a software architect for Verizon. You’re a person with an MBA. Why continue to work for peanuts?

Sean: I mean, honestly, part of it was privilege. I mean, I came from a family that supported me. I didn’t graduate from school with a ton of student debt. I mean, I had a lot of student debt but not a ton of it. I had a wife that was a lawyer. There were a lot of things in life that set it up to make it easier for me to take those kinds of risks, to be honest with you. I never felt like I was gambling everything. I always had health insurance because my wife was working.

Andrew: [From 00:24:40] your wife. Yeah.

Sean: Obviously, lawyers get paid quite a lot of money. It was never like, it was a huge sacrifice. Was it a sacrifice on the opportunity cost for me of things I could have been doing? Sure, but, again, I love doing this, so I couldn’t imagine doing anything else. There were moments when I would worry that I had wasted the three years or the four years that I had done it that it was all for naught, that I had given up getting paid and I should have just gone to work for Google or something. There were definitely moments like that, but I think in the end even if the company failed, I would have not had any regrets in terms of what I learned and the adventure that I had.

But I was very privileged to be able to do it, to be honest with you. It was very lucky and part of why I spend so much time giving back because not everybody has that privilege, not everybody is that lucky to have that opportunity to take those risks without a lot of downside.

Andrew: Before we get to outlier.ai, your new company. Why did you decide to sell it to Yahoo? Why’d you decide to sell it all? I read some articles theorizing, but I’d love to hear from you.

Sean: Oh, no. I’ll tell you the truth. So there’s a few things happening. For me personally, it was nine years. It was 2005 to 2014, I was burnt out. I was actually just . . . There’s a point where you just don’t have any more to give, especially with all the ups and downs that I mentioned and I needed a break. So personally for me, I was just so burnt out, I couldn’t do it anymore, I needed to break.

I think from the company perspective, two things were happening. One is that mobile was becoming quickly no longer its own category. Mobile was blurring with the web. It was clear that to be successful, you couldn’t just be a mobile analytics platform anymore, you had to be a mobile and web platform, which means you’re competing with a whole different set of competitors. And so it’s the question of, okay, you jump into this world where you’re the big fish in this pond over here, you jump into the ocean where you may be a big fish, but there’s lots of other bigger fish out there.

And the final part is, we have a lot of investors, frankly, venture funds, typically, as I’m sure you now have a life cycle around 10 years or so. When you get up against that 10-year mark, the investors are interested in exiting. They want to close out that fund and focus on the other funds, and that was true of a lot of our early investors too.

So there’s a lot of things that came together to make it the right timing. And, of course, the good news was Yahoo at the time was being hit by Wall Street for not having a mobile strategy, and Flurry represented for them a mobile strategy in a box and an executive team that was well proven in this new world. And it was not a coincidence that after that acquisition happened that the Flurry executive team became the executive team for a lot of Yahoo and that Flurry became such a prominent part of their entire mobile strategy. So, I think overall, it was just a matching of the needs and the interest and the timing.

Andrew: Weren’t you . . . Were you working out of the old Flickr offices here in San Francisco?

Sean: I did not go to work for Yahoo. So as the acquisition I didn’t work for . . .

Andrew: Then why did it take you three years to . . . I think you said three years to finally come to a point where you could exhale?

Sean: I think that if you’re . . . At least me, I put a lot of myself into these companies. There’s a lot of Sean that goes into these businesses that I’m building, partly because one of my core motivating factors is I want to create environments in places of work that people want to work. I feel like a lot of companies don’t treat employees like people, they treat them like resources and widgets. And it was important for me to create a place people wanted to work and place I wanted to work, and I invest so much in that that it’s more than a company, it’s more than a legal entity, it’s more than the money, it’s the environment, it’s looking forward to going to work every day and being surrounded by people that you like, that you learn from, that you collaborate with.

And that doesn’t just go away. I mean, some of my best friends were working there and still working there years after I left. In fact, we hired some of them to work at Outlier years after I left. And so I had ties back to that whole adventure the entire entity for many years. So it was really letting go of the community we’ve built both of the customers and the people and the employees, then the legal entity and some sort of transaction that happened.

Andrew: So, you’re a firm believer, you told our producer that you start with a problem, not an idea. What’s the problem that led you to outlier.ai?

Sean: So it actually started at Flurry. So we had at the peak of Flurry . . . When we got acquired, we had about 500,000 customers around the world, and I was lucky enough to travel and meet as many of them as I could in places ranging from the U.S. to Europe, across the world. And wherever I would go, however big or small the company was, whatever vertical it was in, they would ask me the same question. They would say, “Sean, I love this data that you’re giving me. Free analytics is great. But what do I look for? What am I supposed to look for in all this data?”

And I kept getting that question everywhere. And it sounds like a really simple question at the beginning, and when you started thinking about it, what you realize is these companies had so much data that all the coping mechanisms that they had in spreadsheets and dashboards were breaking down and they really didn’t know what questions to ask in the first place. And so Outlier was a . . . rose out of my belief that business intelligence needed to change, that the state of the art of tools today was very good at answering questions, but wasn’t good at asking them, and we needed a set of tools that would bring those questions to you. So, if you run a large business, you wouldn’t start your day trying to figure out what you should be asking, it would bring it to you and the business would speak to you in a way to tell you what’s going on. And . . .

Andrew: Give me a concrete . . . Give me an example of a client who could have used Outlier if it existed back when you were at Flurry. What’s a problem that they had? Even if you don’t mention the name, just anything that’s specific. I’ll understand what you are looking at.

Sean: Well, I’ll tell you an example in terms of a case study we have here at Outlier is a good example. So we work with one of the largest fast food restaurant chains in the world. They have tens of thousands of locations, hundreds of food items they sell per store. You’re talking about massive amounts of data. And so you’re talking about millions of just dimensions in terms of product by store, by time, by location. It’s crazy how much data there is. And so nobody was looking at any of those individual parts.

One of the things that we found when Outlier first got deployed is these companies were learning life-changing things about their business. So this particular fast food chain, the first time they integrated Outlier, they’re running it on their system. One of the questions that brought to them was, “Why was this one store, one specific store out of tens of thousands selling fountain drinks twice as high as they had been a month earlier?” And when you drill into that question what Outlier had found is this one the store was selling fountain drinks pretty consistently and fountain drink sales went to zero for three weeks. And after it was zero for three weeks and it came back from zero, it was almost twice what had been before.

And this is a big deal because fountain drinks are the highest margin item you sell in a fast food restaurant. This is a lot of money. And what happened was this company called up that store manager to say, “What happened? What’s going on with your fountain sales?” It turned out their fountain drink sales had gone to zero because this store had water damage. They had to do renovations and fix the store. And after the renovations were done, they reopened the store, but they didn’t set it back to the layout that corporate had dictated. Most fast food restaurants have a layout that corporate dictates.

They had inadvertently experimented with an entirely new layout because they just hadn’t set it back after the renovations. That new layout turns out to be conducive to fountain drinks sales and sells twice as many fountain drinks as before, a change they’re now rolling out to their entire set of restaurants.

The key thing there is that insight was probably hiding in their data forever. It might never have been found by a person. The data had it, it was there, it was in clear as day, but there’s no way to find it. There was no way to go in there and find it and the system like Outlier using artificial intelligence pulled it out immediately for them because artificial intelligence . . .

Andrew: Because nobody is saying, “Show me all the stores that have a rapid increase in sales.” They’re not thinking to look for that.

Sean: Well, to imagine you’d have to look at every food item, every store because, again, this was one store of tens of thousands. [inaudible 00:32:26] one food item out of hundreds. Even if we as people wanted to look through them all, there’s not enough hours in a year for us to look through every single one of them continuously. And that’s the beauty of artificial intelligence is not limited by the hours in the day. It can look through all your data constantly and find these questions for you.

The one thing AI can’t do is it can’t answer the question for you. It requires that general manager to call the store manager to figure out what happened to take the decision. What it does is it gives you an advantage by finding those opportunities and those problems in your data. So it’s exciting to watch it happen because I will be honest, when I started Outlier and I had that problem in my mind that I mentioned, I really didn’t know if you could build a product to do it. I had no idea if it was possible to build a solution. I really I had no idea. I felt like somebody had to solve it.

The first nine months of Outlier was myself and my cofounder renting ourselves out as consultants to companies saying, “Give us all your data. We’ll just go look through it and we will tell you what we find, and you tell me if what we find is useful, unexpected and has value,” because I was looking, “Does the unknown have value? Do people . . . Can people take action based on the unknown? Are there hidden gems in their data?”

Andrew: And you specifically wanted to be a consultant to understand whether it was useful? Were you doing it yourself by hand or were you using software to figure it out?

Sean: Absolutely. No, we were doing it by hand.

Andrew: Just looking by hand to see, “Can we do it?” And do you remember what you discovered that helped you see, “Yes, this is something that should be turned into software that can solve it and a business that can solve it?

Sean: Absolutely. Every single company we worked with, we found something that changed the entire course of their business every single time. It was amazing. It was by far more impactful than I imagined it. And the reason we did it as consultants is before we invested the time in writing software, before we spent all the time in artificial intelligence and mathematics, we wanted to know, was it worth it? Is it even going to matter in the end? And it turned out it did.

At the same time we were also looking for what kinds of insights, what kind of unknown things had the most value? What should the product do? What should it find that’s most impactful? And so those nine months probably saved us a decade of trial and error just in software development because we knew exactly what we wanted it to do, how it needed to do it, and how customers would use it before we even got started.

Andrew: I highlighted that so boldly in my notes here from your conversation with our producer, Brian Benson. You told him, “We learned a few things from the consultant work.” And I highlighted, I go, “Why is this guy doing consultant work? There must be something in there.” Here’s one. Zero effort to integration, you said, is one of the things. What do you mean by that?

Sean: So one of the things we learned in that process was customers have a lot of . . . business have a lot of data, but if it requires a week or three weeks or a month to get set up and try something new, they just don’t have time for that. Nobody has time for that. And if you put on top of that, we’re selling artificial intelligence, something that people still think is science fiction, they’re not going to spend a month of time on something that they’re not even convinced might work. And so it needed to be so easy to integrate, so simple to click on a button and get up and running, but there’s no excuse not to try it so that they can get that immediate payoff.

In fact, our original customers, we would go into a sales meeting with Outlier and we would tell them what we did and they’d say, “Okay, cool, I want to try it.” And we would say, “Great. Let’s do it right now.” And they would literally create their account and click the button right there in that first sales meeting and see the magic happening on their own data. We don’t do that anymore, but that was definitely what we wanted to get to is that zero effort because it removes all that friction, all that skepticism, all those questions that can come up because it’s just try it, there’s no reason not to. It’s so simple. Why would you not give it a go?

Andrew: You also recognized that companies already have a lot of different data sources. And on your website, I was surprised by how many different . . . You guys plug into a lot of different sources. I was surprised by how many different analytics packages are just commonplace. So that’s another thing that you discovered.

And the final one was, it had to be really easy to use. Users getting the most value of the unknown or executives who didn’t have a lot of time, so you had to come up with a realization like the one that you just mentioned about fountain drinks.

By the way, as you were doing that I went to see, “What could the freaking change be that could have such a big impact?” And it made sense. From one side of the cash register to the other. I imagine now almost every fast food restaurant that I know seems to have a situation where the drinks are after you pay, like, the Chipotle-style, and maybe if we move up before then I might want it before I order and remember and be teased to buy it.

Sean: Exactly. And like all great insights, it sounds obvious in retrospect. And if only you’d ask the right question, we would have come to that conclusion and that’s why questions are so powerful. Questions really are the competitive advantage of tomorrow.

But to your point about users, one of the things we found too, this is funny about outlier is that the average user doesn’t have a degree in statistics. Your average business user is somebody who just runs their business. They know their business well. And what we found was that Outlier had to present these questions in natural language, something that you could read. It couldn’t just be charts, it couldn’t be technical jargon. And it had to be something that explain it to you because it felt accessible.

But even more than that what we found was people wanted to absorb these things in a personal way. And so, for example, Outlier send you a daily email with the questions you should be asking about your business. That email when we first started sending it many years ago came from like the Outlier notification bot. And what we found was people would start opening them on the first day and then slowly over time open them less and less.

But when that email, we change it, so . . . It was still the Outlier system sending that email, but it was sent from your customer success manager, your Carla or Chris. Engagement went over time. People were able to open it and use it and engage much more and I was very curious why such a simple change had such a big impact. It turned out that a lot of those users, they felt like a robot lecturing about their business was pedantic, but Carla or Chris I can trust them, they’re on my team. I believe in them and what they’re doing.

And in fact, some of our customers believe that Carla and Chris were doing all the work. And that personification of the AI was an important part of the adoption people needed. It’s the same reason people don’t feel comfortable in a self-driving car if there’s not a human sitting in the driver’s seat. There’s a level of comfort that you need to think about and . . .

Andrew: You’re saying Carl would be your client who signing up, the emails would still come from outlier.ai’s domain . . . Oh sorry, from your service based on data that Carl gave you access to, an email now comes from Carl too to whoever the decision maker is.

Sean: No, no. So, Carla would be your account manager at Outlier.

Andrew: Oh, got it. Got it. Got it.

Sean: So you bond and you’re getting an email from Carla every day, and so it feels like Carla is emailing you and saying, “Hey, Bob. Listen, check these things out.”

Andrew: What a small change, then.

Sean: It’s such a subtle thing. And what you learn is that a lot of products, at least this is what I’ve learned in my career. A lot of products are as much psychology as there a technology, in that, you have to think about how people receive them, how they process them, how they think about them is at least as important as all of the mathematics that we do in the data science and the artificial intelligence. And that’s such a key aspect to it. Subtle changes have an enormously different approach and absorption by customers.

Andrew: Let me talk about my second sponsor and then come back and talk about why the first customers you picked, actually, even though you earn money from them, apparently they were a mistake, and what we can learn from that. I’ll do real quick ad for Toptal and tell anyone who’s listening, if they hear anything in these interviews that they want to add to their business like let’s say artificial intelligence, someone’s listening and saying, “I think we need this. I think we can use it. We don’t have a developer who can do it.”

You go to Toptal and you’ll be able to hire a developer who not only can do it, but has done it to accompany just like yours or for a company just like yours. And so that’ll work with anything, like maybe you decide that you want to get into voice assistance, you don’t have a developer who can do it, go to Toptal and tell them what you’re looking for, they will find that developer.

And one of the reasons why people who I interview use Toptal is that you can hire quickly and hire by first talking to a matcher who understands what you’re looking for and then goes and gets those people from within their network. If anyone out there wants to go use them or even you, Sean, wants to use Toptal, you should use not just toptal.com, and it’s top as in top of your head, tal as in talent, but toptal.com/mixergy will get you 80 hours of Toptal developer credit when you pay for your first 80 hours in addition to a no-risk trial period of up to two weeks. That’s toptal.com/mixergy, toptal.com/mixergy. Who were the first users, the first customers and why were they not the right fit?

Sean: So we had a hypothesis the best fit for Outlier was going to be small businesses who can’t hire an analyst, and so we would be their virtual analysts. We would automate that function for them. And what we found out very early on was that they did have this problem, they didn’t have analysts, but they also didn’t have any resources or any time at all. If you’re a small company, you’re struggling just to keep the lights on, you’re struggling to make it to tomorrow. If we bring you this amazing unknown insight, you probably don’t have the resources to act on it.

And so it was an interesting experiment. And a lesson we probably had to learn the hard way. Frankly, I don’t know there’s any way to know that ahead of time, because these customers would swear up and down before they started using it that Outlier was going to be powerful for them, and it was only after they were set up and using it that they realized that they couldn’t act on it. It turned out we were a much better fit for larger enterprises who have plenty of resources but their business is so complex, they don’t know where to look and no number of humans looking through all their data will help them. So that was part of that customer development process. But I’ve always found customer development is a series of mistakes that lead you to the answer where you try a lot of things and they don’t work and all they do is help you refine and focus.

So, one of the first documents we had at Outlier before the product was our target customer profile. Who are we selling to? And every lesson we learned would get put into that document and help refine where we were going, and then documenting got to be pretty long. You learn subtle things about, what kind of company? What are the characteristics to look for? First, what do you need to look for these? And over time, it became this kind of large growing knowledge base about all these lessons that we learned the hard way because the only thing that’s wrong with learning lessons and mistakes is if you don’t remember them. And so we codified them all and it became the forcing function that helped us kind of hone in on our target market.

Andrew: Was it hard to go back to entrepreneurship? Was it hard to go back to not being on a roll, not having momentum?

Sean: Oh, yeah. Yeah, yeah. Let me tell you, a success is an addictive thing. I think once you’ve been successful, it’s the kind of thing where you want to feel more of that feeling. And granted, I was very lucky. And I’ve been very lucky in my career. And when we got started with Outlier, I think a lot of my friends assume that, “Well, you’ve been a successful entrepreneur. I’m sure investors are lining up to give you money. It’s going to be super easy the second time around.” And I’ll tell you that is not how it works.

Andrew: Really? Why not?

Sean: Well, I think that it is easier as a second-timer entrepreneur to get that first meeting with an investor or maybe a first meeting with a customer, but that’s really where it ends. From there on, you’re like everybody else out in the universe. There was more people that have told me that Outlier could never work. It was science fiction or this was not a problem that customers have.

I had a lot of people tell me, “Just go back and do another mobile company. You’re the mobile guy, just go do what you know.” And I think it wasn’t until about two or three years in that people started to realize that, “Well, actually, Outlier might be the future of what business intelligence is. This actually might be as important as Flurry was for mobile and the business intelligence industry.” But that was a years into the venture.

People just . . . I think there’s a lot of risk in the unknown and I think the more aggressive you are, the harder it is to convince them your vision is correct. And as a second-timer entrepreneur, it was scary. Like, I was scared that, “What if I fail? People will think that Flurry was a fluke, that I was lucky the first time around, but clearly, he doesn’t have the skills to make it happen a second time.” And I just had to live with the idea that the risk was worth it.

So, did I feel as acute pain as I did the first time around not getting paid? No. Obviously, not. That wasn’t an issue. But I still felt the pressure to want to succeed, to want to win, to want to prove that I could make it. And frankly, like I said before, that I wanted to build an environment, I wanted to work in and recruit the kind of people and have the kind of team I look forward to working with every day. And I’ve done that and I would say that . . .

The way I like to talk about entrepreneurship the second time around is the first time you start a company, it’s like a roller coaster. The first time you go to a roller coaster, you’re white-knuckling at the entire time because you don’t know when to be scared, so you’re just scared the whole time. And the second time on that same roller coaster, man, you know when to be scared. So, most of the time you’re relaxed, you’re enjoying it, but when you know to be scared, you’re more scared than you were the first time because you know what’s coming. And that’s what it’s like. I’m probably much more confident in most of the journey than I was the first time, but when things get tight, I get much more scared. We have not been insolvent in the life of Outlier because I experienced that too much at Flurry and that’s probably because I am more scared of that than I was.

Andrew: Could you tell the audience about the time that you sent your investors your burn rate, your monthly burn rate?

Sean: So when Outlier first got started, I was still fiscally conservative. I learned that lesson with Flurry and so we were burning as little money as possible. When we raised our first investment round, I like to send investors updates every month. I want them to know what they invested in. I want them to know how we’re doing. And the very first month, this was the first year, we were consulting, we were just . . . My co-founder and I consulted. We weren’t spending any money. We didn’t have any servers or anything. So I think our burn rate that month was $200 and I sent this investor updates saying, “These are what we’re doing. This is how things are going and our burn rate is $200.”

And I got this very awkward email from one of our investors saying, “Listen, the burn rate seems high. Are you sure that it should be that high so soon? We have to plan for the long term.” And I was very surprised and I was like, “I think $200 is pretty low.” And I realized that they’re so used to seeing money quoted in 1,000s that he thought it was $200,000 a month. And I was like, “No, no. There’s no K, it’s just $200.” And it was hilarious and a funny anecdote that got shared at many investor meetings and stuff, but it was funny that it was so low that the investors just didn’t know how to process it. But frankly . . .

Andrew: I had the same thing. When I was looking through the notes here, I saw that and then I said, “Okay, 200,000 a month, they’re getting started.” It wasn’t until I read Brian’s notes on your response to the investor that I realized, “Oh, I should go back and reread that. It is 200, not 200K.”

Sean: Yeah, it’s easy to do. But this goes back to . . . I feel like the best way to be successful is to last as long as you can and to give yourself as many chances as you possibly can. And one way to do that is control your cost basis. You don’t control a lot on this business, but you do control how much you burn.

Andrew: You told Brian, “The difference between a successful investor . . . a successful startup and others is . . . ” Excuse me. ” . . . a successful founder versus a failed one is persistence.” What do you mean by that?

Sean: I came into this business feeling like the most important thing to be a founder was to be really creative or really visionary or really smart. And what I realized is . . .

Andrew: Or a great salesperson.

Sean: Oh, something like that. And really, the person is what matters because there’s so many dark days where everybody is telling you, “No,” and giving up is so easy. And like I said, at some point to be successful, you have to get lucky, and the best way to do that is to persist and not give up and proceed through. And so I think that at this point through all of the hundreds of founders I’ve worked with now in the last 10 years and everybody, persistence is really what matters.

I think a lot of why Flurry succeeded was my persistence, not giving up in the 2008 financial crisis, not giving up when we were not making money as a mobile app developer. And Outlier was not giving up when people thought this was science fiction or not the future. Not giving up and continuing is the most important thing you can do because if you give up, obviously, you failed. It’s a dangerous sword, though, because if you don’t give up on something that was never going to be successful, you, by definition, waste your time.

Andrew: So how do you know when you should be adjusting or when you should stop being . . . How do you know the difference between persistence and bullheadedness?

Sean: It’s a hard question. I wish I had an answer for you. There’s founders I’ve worked with that have spent six years longer on something than they should have and been left with nothing but credit card debt and depression. And there’s founders that I know that have pushed through five years and become worth a lot of money. It’s an unclear line and ends up being your contract with yourself and your life about how much you . . .

Andrew: What do you mean by that?

Sean: So, I mean, you think about entrepreneurship in lots of ways. It is an opportunity to do something to build a business. It’s a career pursuit, it’s a job that you have, or it’s years of your life. And so how do you want to spend those years of your life if . . . Honestly, after Outlier is over, and hopefully, it takes under nine years for us to become public and a huge business, will I have enough years left that I want to do it again? How many 10-year hunks of my life do I have to spend on a company? This life is so many. I’m not going to live forever. And so becomes your contract with yourself about how many of those are you willing to gamble.

Like, what if you spend those 10 years and you’re left with nothing? Is that okay? Is that an outcome you’re okay with? And so there is no clear answer. I think, in fact, the problem is that the best opportunities look like the worst ideas in the moment because they haven’t materialized yet. I mean, frankly, had Apple not launched the App Store in 2008, would Flurry have survived? Would it have become as big as it is? Probably, not. I mean, honestly, the App Store a year later might have killed us. So there’s a lot of things . . .

Andrew: You mean if they even came out with it, but it took another year, it would have killed you.

Sean: We might have been out of business. I mean, I don’t know, right? There’s no way to know. There’s a lot of unknowns in there. It’s tough. So I focus on what I can control, which is, is what I’m pursuing worth it? Do I think the potential is there for me? Are those years of my life worth it? And the answer is always been yes for me.

Andrew: The one thing that I keep coming back to with Outlier, actually, two things, one is the domain and the interesting story around there. The other one that I’m going to kick myself for not asking later is, I feel like with Outlier you’re saying somewhere in the data is some kind of value, and I wonder how easy that is to communicate to a client and convert a prospect into a client when you’re not saying, “Somewhere in the data is more sales. I will find it.” or, “Somewhere in the data is reduction of churn.” But you’re saying, “Somewhere in that data is value.” And you’ve always said that from what I can see on your website. Is that tough?

Sean: It turns out no, and the reason is . . . And this is a good question because this was one of the risk factors going in. And it turns out no, and the reason is that people have enough tools, enough data to know what they’re missing things, but they’re not finding out they missed it till months later. So, imagine you’re constantly getting to the train station, but you’re always missing the train. It doesn’t help. All you know is that I missed something big that cost me millions and tens of millions of dollars.

So when we go talk about all Outlier, I don’t have to sell them, all I have to do is connect them with that whatever happened and everyone has a story. Every business has a story. And they will tell you those stories. They’ll say, “Oh, we missed this huge fraud,” or, “We missed this big opportunity in customer behavior.” And that’s all it takes. There’s so much pent up demand to get ahead of it.

The other thing that helps us out is that there’s such a dearth of data scientists in the world and all these companies are trying to hire data scientists because they’ve already decided, “We have this problem. We need to get head of data. Let’s hire some people.” And there just aren’t enough people to hire. So, they look at Outlier, they’ve already made the decision they want to do something, they just had no idea there was a software solution for it. And once they find out, they want to try it and once they try it, they want to buy it. Our conversion rate from pilot customers to actual customers last year was like 100% because it’s a little bit like somebody is navigating their car using a paper map and you hand them a GPS unit, are they ever going to give back the GPS unit and go back to paper maps? Absolutely, not.

Andrew: Can you say what your revenues are?

Sean: What’s that? In the millions of dollars.

Andrew: That’s all you’ll say right now.

Sean: Mm-hmm.

Andrew: Outlier.ai. How did you get the domain and talk about what happened there?

Sean: So, what I learned early on in the first year of Outlier was I had to mention that we’re using artificial intelligence because before I mentioned AI people just didn’t believe software could analyze your business, but then I would say, “But it’s using artificial intelligence,” and they’d go, “Ooh, that sounds really sophisticated.” So I was like, “Okay. It has to be there. So we need to get a domain, let’s get outlier.ai.

And .ai wasn’t a super popular domain back then. It’s a small set of islands, I think in the Pacific. And there was only one registrar on the internet that would sell you .ai domains, and so I went there to buy outlier.ai. And it was one of those websites that you feel like when you put your credit card in there, it’s immediately going to be stolen. It was very sketchy. But I was like, “Listen, this is important. Let me give it a go.”

And so I went in there and I tried to buy outlier.ai and I pressed the purchase and I was like, “That was a mistake. I shouldn’t have done that.” And lo and behold, four days later, my domain is still not live, I’m like, “Yep, my identity was stolen. It’s definitely a scam.” But I went on to that website and I asked them, I was like, “Listen, what’s up with the domain?” And they responded, they said, “Listen, the person who runs that domain he only works on Sunday afternoons, so you have to wait for Sunday afternoon to come around and he will go over and type your domain into their root server and you’ll be up and running.” And that’s exactly what happened. So .ai was . . . I actually regret not buying a lot of .ai domains back then. I probably would have made at least as much money off of that as I will off of Outlier because it’s caught fire, but back then it was the Wild West.

Andrew: And you’re right. Even when I was preparing for this, as soon as I saw ai in your name, I thought, “Okay. I’ve heard a lot about the power of ai. I get it. There’s something in here that I want to find out.” And it gave it a lot of credibility. Meanwhile, you’re saying when you first started doing this as a consultant, you were doing it by hand. So, the data is there, the artificial intelligence is just software that makes what we’re doing or what we could do faster and scale better. All right, for anyone who wants to go check it out . . .

Sean: I have a secret, by the way.

Andrew: What is it?

Sean: Quickly a secret. If you want to build an AI company, this is the formula. You basically hire people or do the job yourself, and slowly use AI to replace pieces of that over time, and eventually have the service that you want. It’s how every successful AI company gets built, is you just do . . .

Andrew: What’s the difference between that versus automation? Don’t you usually look for automation? AI you’re looking for patterns. Automation you’re looking for repetition. Am I right?

Sean: That’s true. The thing is AI ends up being able to automate more than traditional software, and so you can eat more of that experience, you can be more ambitious as to what you’re doing. But nobody starts out just by building an AI system and launching into the world. Everybody starts out having humans do it and slowly eating away from the back. And the customer may never know. The customer may never know that it’s not people doing it anymore. It’s a tried and true practice for building AI companies.

Andrew: You know what? I just talked to Neil Patel who runs a digital agency that’s growing really fast. I said, “Why are you taking these smaller companies when you realize . . . ” And this, Sean, this is something you guys realize that Outlier enterprise is a bunch of better client for you. And he said, “I’m taking the smaller people because now my people are doing the work for them and I’m slowly bringing in . . . ” I think he said he’s spending $2 million in artificial intelligence, slowly automating it expecting that in the future the automation will take over every part of what they’re doing for the smaller clients.

All right. I’ve learned a lot from you. Your website is outlier.ai. Don’t also have a personal site. I got so many tabs open for you. You were writing for a while, weren’t you?

Sean: Yeah. You can find . . . So, seanonstartups.co. It is my blog.

Andrew: There it is.

Sean: It has all random rambling about starting companies and being a founder.

Andrew: Also seanbyrnes.com is a thing?

Sean: That’s right. I have lots of websites. It turns out they’re very cheap.

Andrew: There’s no sean.ai yet? It’s all you personally writing your stuff?

Sean: That’s right.

Andrew: And I want to thank the two sponsors who made this interview happen. The first if you need to hire developers to do artificial intelligence or anything else within your business, go talk to Toptal. Check them out at toptal.com/mixergy. The second is a company that will help you figure out how to get more traffic by helping you analyze your traffic sources, your competitors and others. Check out ahrefs.com. That’s ahres.com/ nothing, actually, just .com. And finally, I’m going to be doing marathons all over the world. My goal is to run one on every continent by the end of the year. By the time you listen to this, I should have four down and I’m still working on an Antarctica. I’ll update you guys on all that. Sean, thanks so much for doing . . . Do you have any hobbies? You don’t, do you? Just working.

Sean: I don’t have time . . . I have two young kids and a startup company, man. That’s plenty for me.

Andrew: Yeah, I get it. Thanks so much for taking the time here. Bye, Sean.

Sean: Thanks for having me. It was great.

Andrew: Bet.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.

x