EP7: Finding Product-Market Fit with B2B products, with Paul Sawaya, founder Human Interest ($1b valuation) Artwork

Paradigm Shift

Candid conversations with people building the future. We discuss breakthrough moments and hard-earned insights that shaped people, products and technologies. Hosted by Ashish Kundra and Zane Salim.

All Episodes

Paradigm Shift

EP7: Finding Product-Market Fit with B2B products, with Paul Sawaya, founder Human Interest ($1b valuation)

October 30, 2021 • Ashish Kundra and Zane Salim • Season 1 • Episode 7

Summary:

Conversation with Paul Sawaya, co-founder of Human Interest, $1b company that makes it easy for SMBs to provide 401k savings plans to their employees, and helps employees save for their retirement.

We cover so many interesting topics in this conversation across the Human Interest story, numerous lessons learnt, and technologies he's most excited about for the future.

Episode Highlights:

Early experiences like building an FB app at Bloomberg
Interviewing with Emmett Shear to interview at Justin.TV
How big company vs tiny startup experience made him want to start a company
Spending 2 years exploring problems and ideas
Challenges and lessons from wandering the wilderness
Key lesson: focussing on problems first before jumping into futuristic projects
Got excited about solving 401k access problems which led to Human Interest
Only 20% small companies offered 401ks due to setup, admin and cost challenges
Product innovation: automating operational work to lower cost structures and expand access
Building on top of back-office solutions to learn/move fast, then moving in-house
How they moved fast to get early customers for feedback and validating demand
How to spend time during exploration/sabbatical: establish a routine and talk to lots of people
Approaching GTM through an effective partnership and integration strategy
Cleaning up some shady 401k practices by offering better, sustainable solutions
Transitioning out of day to day Human Interest work to spend time with early stage startups
Exploring new ideas for his new project and things he's excited about
Exciting breakthroughs other people might be under-appreciating: Web Assembly to run high intensity server side compute locally in the browser, and AI to reimagine user interfaces
Prototyping to explore and learn new things
Ability to enter ambiguous, messy situations and find a path forward
People who gave him an early break that he is grateful for
Power of developing personal values as a guiding light

Links:

Read more about Human Interest: https://humaninterest.com/
Follow Paul on Twitter: https://twitter.com/automin

Hit subscribe to keep up with new episodes!

Follow Ashish and Zane on Twitter for summaries

Click here to give feedback — it only takes a minute.

0:04
Hello, hello, you're listening to paradigm shift, a podcast about people building the future in pivotal moments in their journey. I'm Ashish. And I'm joined by my co host, Zane. And today we're super excited to speak with Paul suya about his startup journey and CO founding human interest. human interest offers retirement investment services to startups, and recently raised $200 million at a billion dollar valuation. Paul, thanks for being here.

0:29
Thank you for having me really excited. Course,

0:32
I'm really fascinated by your background, particularly because you've worked at like very distinctive companies that each have their own kind of ethos earlier on in your career. So maybe we can start there. Do you mind telling us a bit about how you got started in startups and software in general?

0:47
Yeah. It's interesting, right. So my prior to human interest experience, to put a generously wasn't a lot, right, I hadn't done a lot before. But my love of technology and computers goes way back way back to age nine, where I begged my parents for a computer and started teaching myself HTML, JavaScript, and actually Visual Basic six. So I knew from a very early age that something involving computers or technology would be what I would spend my life doing. But I didn't really have a clear idea of what that would be. The first time I ever got paid to write code was when I interned at Bloomberg, all the way back in the summer of 2008, in New York City. And I managed to get that gig as a summer intern, despite my grades not being so great, because I built a Facebook application the summer before that went viral. So they saw that, like, this person has talent. And here, I was working in this serious financial company. And I was really excited to be paid to write code, honestly. And I really enjoyed it. And we're very hard that summer. But remember leaving thinking, you know, at the end of the day, all I'm doing is making numbers appear in a grid on a terminal. And I don't really find that interesting. I remember thinking, I'm probably never going to work in finance or financial technology ever again, which is kind of famous last words. And, you know, as we get into the human interest story, I'd love to connect that back because it was kind of eye opening to me to realize how that work could be fulfilling. We're very proud of my work at Newburgh, we shipped a feature that is still used today. It's if you ever in front of Bloomberg terminals SDS H go, it's the foreign exchange dashboard. I recently met a trader who is works in forex. I mentioned the feature the function on the Bloomberg terminal. He's like, Oh, my God, I use that all the time, which is kind of what you want to hear as someone who builds things, right. But people are using your software. I love it. The real pivotal moment for me, though, was the summer afterwards, I had an interview at Microsoft. And I was so excited for it right past the phone interview, and they flew me out to Seattle, and I did the interview on site. And I completely bombed it totally screwed it up. I was so devastated because I thought you know, I had Bloomberg, which is my stepping stone. Microsoft has the big leagues. That was my shot, and I blew it. So I came back to my college and East Coast pretty disappointed about that, and decided kind of a desperate move, just research startups like small companies that might want to hire me because kind of a, the intern hiring season was over. And I thought I was out of options. I randomly applied to the startup called Justin TV that I found on this random website called Hacker News. And I'd found Hacker News because I've written a blog post and someone submitted me there. So I saw it. And so it was a referrer. And I started reading it. So there's a job section I applied and didn't hear back from them for a week. So I seemed like they weren't interested. I just sent my resume and a code sample into a black hole. I remember very vividly, I was headed to the college dining hall to get some food and my phone buzzed. And it picked it up. And it was it was Emmett from Justin tv, then the the CTO now of course, the CEO of Twitch. And he asked me about a very specific line in my Python program, and it sent it as a sample and explained to him why I did what I did. And then he just hung up on me. And then a few hours later, I got an email from Justin Khan saying we'd love to have you for the summer like sweet, okay, the culture of working at an early stage startup like Justin tv, back in 2009. couldn't have been more different from working at Bloomberg, right? It was maybe 20 ish people, the founders themselves were like 26 years old. And it wasn't really a well oiled, well run company. And I think Michael and Justin would be the first two minutes, especially about those days, but there was a strong feeling like everyone there was very smart. And we're kind of all in this together. And everyone has to be a little bit creative. You know, the engineers aren't just being handed designs and being told to build them. They're not just talking about code. They're talking to everyone in the company, right? They're like brainstorming new features with the CEO. They're trying to find new ways to kind of make the product more useful. They're coming with features on their own and shipping them. This sounds like fun. And I looked at founders who were very smart, but also very inexperienced. And I thought we could probably do that too. And and that planted that seed in my mind. In fact, a lot of the people I met that summer went on to be startup founders. One of them actually asked me to try and drop out of college and work with him on a project. I declined. You know, his project was going anywhere. He wasn't even technical, but he later went on to to prove me wrong and start a company called Zero. cater, which is like a very successful YC company that does office catering. So there was a lot of entrepreneurial energy there a lot of people taking initiative. And after that summer at Justin TV, I knew, I knew I wanted to do a startup someday. I later Later interned at Apple, and then in turn two and worked at Mozilla. And those were interesting experiences, but really only reinforced by learned to Justin TV. One, one thing I'll never forget from Apple was, my project was basically to take this open source project that Google built and kind of embedded into Safari, and WebKit. And we had to meet with lawyers in order to do it. So like my boss, and then a bunch of Apple lawyers and me were in this room, and my boss was trying to convince them that there was nothing, there was nothing wrong with us using open source. And it just blew my mind, we had to meet with lawyers to do open source. And finally, the phrase that when the lawyers over was, Look, we're not doing anything innovative here. Okay, that's fine.

5:55
You know, we're happy to kind of like approve the licensing of getting this open source project in the safari because there's nothing, there's no IP risk, there's nothing new or innovative. And I was thinking, Man, I love using Apple products, but it can never work in environments. So that early days, you know, pre human interest that my career,

6:13
it's fascinating that you had this amazing mix of very large established companies like Bloomberg and apple, and then you had Justin TV, which was you wouldn't know at the time. But that would be this source of so much creativity and development in Silicon Valley of all the things that like, all spin off companies that spawned off it, were the things when like, after that experience, when you're heading into nitrous, they were like, I'm definitely gonna take these things into my next startup. Yeah, human interest was a very

6:41
interesting learning experience for me, right? I mean, I went in with four internships and eight months of full time work. And I had the audacity to call myself a CTO, right? A lot of startups these days are founded this way, but high energy, low experience people. But the amount you learned from being thrown in the deep end like that, it's completely changed the way I think about work. And about business. Of course, though, every startup kind of has its own challenges. And I think that the challenges that we saw at human interest, especially in kind of the early to mid days are mostly around getting people coordinated, and sharing information. So both myself and my co founder, Roger Ver very hands on people. In fact, when we wrote our company values, one of the values was keep your hands dirty, right. So don't be afraid to dive into the details no matter who you are. And we really live that value. Like, even when it was way too big of a company. For us to be doing this reasonably, we were still like, I was still like reading code reviews. You know, I was still writing code, I was on the phone doing first call engineering candidates, Roger was still deep diving into operations with individual operations associates, talking to sales reps and initial calls. And the biggest takeaway I learned from that is, you can't succeed in the early days without doing that, like you really can't outsource finding product market fit, you can't outsource the basic engineering and early days, you have to get your hands dirty. But then, as your company grows, you can't succeed unless you stop doing that. And seeing that, that that shift, where what was doing or what worked to get us to a certain point wouldn't work going forward, really left an impression on me. And when I think about my next company, or my next startup, I'm kind of obsessed with this question, right, which is, how do I know when to pull back? Because we definitely succeeded in the early days, because we didn't pull back. And then in the later days, we had scaling issues.

8:27
Let's dive into that a little more. So zooming out a little bit, what got you to work on human interests or human interests, for folks who don't know, makes it easy for small businesses to get affordable 401k plans, maybe you can tell us a little more about what it is and how you sort of got connected to that problem. And what inspired you to work on it, or human

8:45
interest. It's a great business. So I meet Roger Lee, who be my co founder and human interest. And just to set the stage here, I kind of spent two years kind of in the wilderness, right? Trying to start a startup bouncing between ideas, and co founders running out of money, doing contracting engineering gigs, to get my bank accounts back up, and then going off and trying to start companies, other people. And after two years of that, I was pretty, like I was pretty down. And I think the thing that really connected for me was the fact that, you know, I've been writing code and programming since I was nine years old. And I was so into the idea of building things. I loved it right, but hadn't really built anything that people used. And I remember thinking at that point, maybe instead of working on projects that I thought were interesting to me kind of things that are sexy from an engineering perspective, like one project I worked on previous was about a friend's startup that was kind of a futuristic kind of spreadsheet on the iPad, which is a beautiful demo, but really didn't solve anyone's problem. So instead of doing things like that, how can actually solve problems for people and make their lives better? That really coincided with when I met with Roger because he kind of showed me the current landscape of 401 ks and options for small businesses. So if you're a small business today, you probably don't offer four. Okay, I think the stat is companies that are 100 employees or fewer of those 20% Some of them offer 401k, which is a huge tragedy because the 401k is the best means we have to save for retirement as Americans, right? Social Security will not be enough for maybe we'll even be there at all. For us, IRAs are great, but you can't contribute nearly as much money as you can to a 401k. Most people don't know how to set them up or just don't set them up. So 401k is kind of a social safety net for retirement for Americans and has to be procured through your employer, employers that want to set it up, because there's a lot of manual data entry, like you add someone to your payroll system, to add them to your 401k first, last social security number, hire date, salary, all that some leaser company to mark and terminating your 401k return, you run payroll, you have to drop a payroll file into your 401k. And then And then also, just at the end of the year, there's a lot of compliance risk in having a small business 401k plan, there's a lot of risk that you may own, a fine or an extra contribution on behalf of your employees. And a lot of small business owners really didn't know how to think about that risk, because the person running HR or benefits at a five person company is the founder or is the small business owner, and they don't really have time to run a benefit like this. So human interests, innovation, main contribution was basically to just address all those needs at software. So we integrate with all the major payroll providers, right, we automate all the compliance work and do it paperless for all of our customers. So now, a 235 person company can set up a 401k with a click. And we have customers now that have over 1000 employees, we're happy to grow with our customers. And that I think, I think that was the insight, right? Extending the market a 401k is rather than just trying to be get another 401k provider, it actually holds up to, if you look at our customers now 75% of them have never offered a 401k before using a given interest, which I think validates our thesis, right, we can make this really easy to set up for companies that thought they were too small for them. And it works. So that's kind of human it's so

11:54
it's such a tragedy to think about how so many businesses in the country, small businesses that sort of employ like a vast majority of like the folks in the country in aggregate just didn't have 401k. So it's such a huge and important problem you get solving. But why is that? Why did the market not have a solution for it? Like how did we end up in this state where big companies offered 401 k's and and the smaller guys just didn't have a solution or like an economic model that works to make this available to

12:21
their employees? Because the old guard or foreign key providers? Great question, by the way, like coming in from first principles. There's all this opportunity. There's all these small businesses, why aren't Why are the needs being addressed? And the answer is, these traditional foreign key providers aren't software companies, right? Because they aren't when they sell you a foreign key, what they're really selling you is manual labor, they're have people on their end processing contributions process doing the record keeping work. And then they actually set people on your end to do a fair bit of work to do because they assume you have an HR department. So the products are priced accordingly. They're pricing in human labor, which means there's a too expensive for small businesses or they're affordable. But there's too much work posted on the small business, all of that work can be automated away with software. And we did it you know, we automate the keeping piece and the parents piece. And suddenly, it became a viable option, again, because we thought like a software company, I consider human interest to be part of this broader trend of companies that take benefits that are really only available at big companies traditionally, right, like Google or some top law firm, and make them available for small business owners. And I think I think that's terrific. And it makes me really happy to see this trend,

13:28
Roger, I think, sort of found this problem some extent and pitched you on it. How did he find it? Was he working at a small business? Was he in a company that faces like, what was added, like part of that discovery process?

13:38
It's a good question. And I get asked this a lot, right? Did I just wake up one morning and decided that I want to spend the next five years of my life fixing retirement for small businesses? And I'd love to say Yes, right. But it was incremental. So Roger has always been a big investing geek. And he kind of talked to me about the issues people have managing their 401 K's. So we went with that kind of a different version of the company. At first, our initial idea was kind of a b2c 401k advisor. So this company called Future advisor, which kind of did this in brokerage accounts, we're going to do the same for foreign key accounts. We worked on that for a while. And then we found out that basically, charging money for what we're doing would be very difficult or impossible. And that's when Roger realized that actually, his experience of 401 K's wasn't just on the investment side, but also the administrative side. He was COO at his last startup, he ran the 401k plan. And he said, Wait a minute, maybe instead of marketing to cost consumers, you can market to businesses, and it's a win win, because we're expanding access to the 401k by getting businesses that otherwise wouldn't have. It's all for the benefit. Plus, from our point of view, we're acquiring customers a company at a time, right? So instead of going out, reaching out to Angel customers wanting sensing market sensitive marketing campaigns, we're just getting them 510 or 20 at a time. And if you look at actually a lot of FinTech companies, particularly the ones that aim to do good by improving people's spending habits and urging them to invest from the investor point of view in those companies, they're very difficult Because of all this spend on CAC and b2c acquisition, so by basically going straight to businesses, we guaranteed ourselves a scalable, more scalable business, one that could actually fulfill our mission of helping everyone have a stable financial future. And that was the insight that really brought us to where human interest is today.

15:18
So correct me if I'm wrong, but one of the reasons like this market was underserved was because traditional companies that offer 401k plans had these very like human heavy sales, heavy processes, and they kind of had, you know, high cost structures, it was hard for them to serve these smaller customers were very price sensitive, you guys figured out how to sort of go to market through like a web first step, first, lower cost product solves some of these problems you could bring on these smaller businesses. But what were the economics like I imagined, it was still kind of hard, because you probably had a bunch of really small companies. And you don't really start making money till like the savings grow over time, right. So it's kind of like this, it compounds really nicely over like a long time. But those early years, I'm sure the economics are a little tricky. So help us understand what you found, and how you guys tackle some of those challenges.

16:04
It's funny, because in the early days, we were thinking about how we could price our business so that we don't run into that problem, right? I remember, we were in YC, office hours with the YC partner, Aaron Harris. And we're our first thought was, we're going to be free for employers not gonna charge them anything, or it's gonna make all their money on assets. And yes, it takes a lot of compound. But once it compounds, it's great. And I remember what he did is he went up to the whiteboard, and he drew two lines on a line chart. And he showed one is revenue growth, purely on assets, of course, was a slow crawl up from a low base. And then he said, revenue growth from SAS, and he drew like a straight diagonal line. And of course, the diagonal line didn't touch the the Aum growth line until what would be like four years out. Fortunately, that actually was part of the insight that we had fell into businesses. So the way human interest makes money today is we have two revenue streams, we charge the employer immediately a monthly fee, they also pay it annually. And then we charge essentially, the employees that people hold the assets is a very small asset base fee. And by having both of those revenue streams, it actually helps us achieve our mission, the SAS fee that we charge small businesses that addresses marketing costs very quickly, right, we can pay our salespeople Commission, we can pay for our Google ads, we can pay our partners, and actually a lot of its partner driven acquisition now and that's great. Meanwhile, having that jolt of revenue allows us to charge in assets under management fee, that's very low, very low compared to most warranties, especially small business 401, K's some of them are in excess of like one and a half percent, like almost getting to like V SL two and 20 less, which is I would argue almost criminal given sort of people that are saving retirement accounts and the investments that they're they're making. So we charge a much lower fee than that. And that allows us to see account expansion, that kind of compounding you talk about over time, which in turn allows us to offer better customer support to our plan participants, but without without having to charge the high fees. So those that kind of a two, this two revenue streams are really what makes the business scalable and actually a good experience for our customers. And obviously, we've had competitors that have tried one without the other, they're definitely foreign keys, they try to remove the asset base, it turns out, they can't scale, we offer great customer support. There's a few companies that tried doing it without the just purely charging participants and nothing else. And of course, being a charging participants a lot and they have trouble getting off the ground because they can't pay for marketing. So the two, the two revenue stream approach is really what made us a success. So it was

18:28
cost structure innovation through software and automation. There was this go to market innovation to like a self serve portal and partnerships. And there was this business model innovation around the SAS fee and the Investment Commission as well. And, you know, small businesses are always pushing back on fee. So I can kind of see why a lot of your competitors might not have had two revenue streams from day one. But sounds like you guys do that was it was a big part of the unlock one more question on that. How did you guys tackle learning about the space because it's kind of kind of like a high regulation, a lot of like, process and it's sort of like administrative stuff. There's a couple of like 20 Somethings, you're probably in your early 20s going up against this, you know, really complicated industry like, which is almost like seven, eight years ago now. How did you guys sort of get up to speed? And how did you like, work through some of those regulatory and like, you know, compliance type of questions.

19:20
It's funny that you mentioned mentioned the fact that we're just some 20 Somethings because literally I remember early customers would go to our about page, and they'd see two fresh faced young people. And they'd say, Why should I buy a 401k from you know, industry experience and you look like my son's there's something like that. A lot of it on the sales front was learning to brand ourselves in a way that was trustworthy and showed competence, right? And the way we did that, and the way we dealt with all that regulatory compliance stuff. In the early days, we worked with a third party administrator. So for first couple years, we were built on an external firm that kind of managed the back office of days for us, and they've been in industry for a long time and we pitch our customers saying you're not really trusting us so much as your Interesting that they hold the money, we're just building the nice web based software. And we did that for a couple years kind of just to teach ourselves, the industry, you know, and we were an early stage startup we couldn't hire, you know, ahead of compliance in the industry can hire a record keeper from the industry and operations person from the industry. And after, after a couple years of that we really learned the nuts and bolts, it's Warren Kay's, and he's grown enough that we can raise enough money to hire those people. And at that point, it became time to build on our own platform. And, and that was a big project that I led it human interest was building what we call wave four. It's an internal codename for basically our internal record keeping software and the organization that supports it. So we started that with two engineers. And then I hired someone who'd been a foreign key record key for all her life and was just very entrepreneurial, and knew the industry as being very stodgy. But also, she had an engineering degree and was kind of hungry to shake things up a little bit. And we built our own record keeping software internally, which many people in the industry said don't do it, you know, we chatted with people that were going to license a software, they said it would be a bargain, it only be like two or $300,000 a year for the first handful of plans. And then yeah, you have to find a local license on top of that, and yet run, you know, Visual Basic and runs on Windows, but it's like it's rock solid. And I considered it but I looked askance at it. And I said, now that we spent two years learning we have to do to be a 401k provider. But our responsibilities are, we need to like continue with one of the base insights of the company, which is we're a software company, we build our own software. So today, virtually all of human interest customers is a handful of holdouts for complex reasons. But like 99% of human just customers running our own infrastructure. And the moment we flipped that switch, we noticed a lot of really great things, our customers had a better experience, their contributions were processed much faster, we noticed that actually, we're spending less money on support, which is great, like turns out, you would think that if you lose a third party that has support people, you bring it in house, you have to pay a little more in support. When turns out, we're actually paying people to do coordination between the two organizations. So simply moving in house actually reduced our support costs. And and now we have a scalable technology platform to serve 1000s of 401k customers like 1000s of businesses, and you know, 10s of 1000s, soon to be hundreds of 1000s of individuals. You know, before that we were actually the largest platform on that third party record keepers, because most of the time who were selling for and reselling foreign keys to financial advisors to individuals, and we just showed up and put 1000s of plans on them. And they were straining under our load. And now we can set our own our own destiny. I think you see a lot of FinTech companies do this actually build on the third party and then move it in house. And we were able to execute it with the pressures of growth and start boys hasn't all those other, all those other kind of startup headaches. And it was it was really inspiring to see the team pull together to do that, because it basically set the company up for its next stage.

22:46
That's a really clever approach. And it reminds me of maybe its industry thinking what lithic did are we talked about started with it. And they offer a debit card issuing API credit card churning Pattu. And they started with the legacy system, paid through the nose dealt with latency issues, and then in the background, built their own system eventually moved off. And everything just went 10x faster after that, once they validated. Could you tell us about your first customer?

23:13
Yeah, this is a funny story. Actually, our first customer was a small business that consisted of my co founder, Roger and his mom, which he already happened to have, like he was he was using this to kind of, you know, work with his mom on projects, like she was helping him out part time or whatever. And we signed them up, you know, as if they were an external customer instead of a 401k. Just to make sure it worked. Right. After we did that we could go to our angel investors, many of them were founders and say, Hey, can we sell at your business, we already have a customer. And by the time we kind of sold our investors, we can go to our YC batchmates. And say we already have 10 customers on live on our platform. So that's kind of how we hacked that problem. Because it's very true that in something like FinTech or employee benefits, no one in the world wants to be your first customer, right? Do you want to be the first person to try the next Robinhood? You'll be the first person to invest your money in a brokerage app? Do you want to be the first person to buy health insurance from a startup like, especially if you've been inside startups, and you know, you know how much duct tape and just like how operationally crazy they are, it seems kind of insane to be the first one to try out a new platform. But someone has to do it. So there's always little hacks like that you can do. I've since started advising startups. And one thing I sometimes tell them that are b2b and kind of high trust is never sell your first customer. Like you should never go to a customer and say you will be our first find someone find a friend, find a way to be a customer of yourself. Just don't make some poor early adopter, be your absolute first guinea pig, because you're building trust by showing that you've moved things through your system before in our case, move money through your system, and then too early customers will see bugs but they won't see the really embarrassing one. Right?

24:49
That makes total sense and how to sort of ladder up and get traction. How long did it take you to get your initial product to market so kind of like you know you're trying to validate demand you have a couple of customers how quickly we're able to get a we want out How did you think about getting to default to lie around in profitability? What were some of your methods,

25:05
we were very early out the gate and Roger and I were very oriented towards kind of ship fast, sell fast. And you meet people in kind of our space who were not that way. For example, gusto, the payroll company, which originally launched ZenPayroll. Now, they do more than just payroll, but they launched the payroll company, they were in stealth, right, I think a year, just getting their software right before they even signed a single customer. And that's not a bad approach, right? You know, people want payroll, it's not like something you have to test or validate the payroll, there's a market for payroll. But building a payroll system as a startup is so fraught, especially with kind of all the nuances of tax and issuing W twos, that you can imagine wanting to be in sell for a year, that just wasn't our style. I mean, Roger, our approach was basically, let's go out and get 20 people to believe in us. So during YC, we signed 20 customers, and we said, we're gonna launch, you know, YC ends in August, are going to launch in September and October. And that gave us a month to kind of test out the platform with Rogers test company, and just kind of make sure it all really worked. And we actually had a pretty successful launch, I'm not going to say that some of our early customers didn't see bugs, but it was it was a lot less scary than we thought it would be. And in hindsight, I'm very glad we took that approach. Because there's obviously the reliability part of your platform that you want to kind of de risk right, by by getting to market early. But then there's also the feature side that you really want to get early validation on. And the moment you start putting screens in front of users, the moment they can say, this isn't clear, I'd like this change, right? Or can you do this, and that's how you start improving your product. So I think a lot of stealth startups do themselves a disservice by basically missing out on say, a year's worth of product learnings from customers. And in hindsight, I wouldn't do anything any differently.

26:43
We know what comes next, which is, you know, once you're in market, there's massive adoption, you guys have, like, I'm sure many ups and downs. But overall, a very successful company, which is now a billion dollar company just raised a massive round. Before we started, like, continue down the road, I kinda want to having talked about how you found success, I want to just look back a little bit to those two years you had where you were like wandering the desert, trying to sorta like figure out what to do. I think a lot of entrepreneurs go through that. And it's an extremely difficult time for them personally, and it's not really talked about. So when I kind of hear about like, what was your psychology at the time, you were clearly like a talented software engineer who, you know, who could have worked at Apple or Bloomberg, you done all of these traditionally, very respectable and probably like, attractive things. But you you stuck to your guns and stayed on this really difficult path where you were wandering the desert? So tell us what was what that was like, What kept you going? And what advice might you have for someone who's going through that today?

27:43
It's something that people I wish we'd write and talk more about, which is basically how to manage a time when you have no no job, right? No professional obligations, because it's a great opportunity, right? You can learn new things, you can collaborate on projects, and finally don't have a nine to five weighing you down. But it's also intimidating to have that much that much time and that much autonomy. I think the biggest mistake I made in those two years off before sort of human interest was I was still thinking like an engineer and not like a founder. So all the projects I wanted to work on. Were like I said, very, very technically cool. They were like they were sexy engineering projects, right. And I started with technologies, programming languages, techniques, I wanted to try and but how can they bend this into a product? Which is not how great products are built? Right? Great products start with a question or a problem, and then you devise a solution to them. So I think I think that's why I really had trouble getting things off the ground. Because I was looking at things through the eyes, I always had just as an engineer, and thinking about what people need is a step further. And that's something I learned a lot about at human interest. As far as advice for other people who are in this situation, I would say on kind of just the scheduling, it's a personal level, making agenda, make a routine and stick to it, right. And there's so many books on how to do this. You know, atomic habits is a very popular one for good reason, because it's great. And you know, you don't have to over optimize it don't don't like look at like the five best to do list apps, the five US calendar apps, just pick one and use it. Because starting a routine, especially early on in your sabbatical, we'll just set the tone for the rest of it. And then my other piece of advice is become a wide acceptor of introductions and conversations, like talk to everybody. Talk to say someone you met at a conference just before the pandemic, right? Talk to someone you haven't worked with in two or three years, but you missed them right? Reach out to people and and really just embrace the fact that you have only possibilities and opportunities in front of you and enjoy it because soon you're going to start narrowing down instead of you know, you're going to do anything you can imagine next you're going to say I'm interested in maybe this space, so that space and interested in products in that space and then before you know it, you're working on your next thing, but there's something very special about having that time off because it's one of the few times in your career where you can fully embrace all the possibilities in front of you.

29:59
I really like the part about enjoying it because you're right. Like, once you narrow down an idea, it's like full on execution, it's a very unique phase, to be looking at various problems and being open to new experiences. human interest is, in so many ways is like the poster child of a company, right? You found the problem early on, you have this solution to get to market in a fairly efficient way. It was a very capital efficient business, assuming CrunchBase is accurate, like you raised $100 million. But most of that was like, really in the last year, why do you think you guys were so successful?

30:30
It's so flattering to hear you say that we're a poster child, and we executed well, and I guess ultimately, we did, right, because we reached this level of success. But man, we made so many mistakes, building that company. And I just want to say that on the record, so that other entrepreneurs listening to this, know that right, that you can make a ton of mistakes, and still build a billion dollar company, I think it's more about getting the important things right, then every little thing you get wrong. The important things we got right, as we discussed the the business model and the opportunity. So we chose a large market. And we chose a business model that actually allow us to address that large markets needs. Other foreign companies that chose different business models have failed. And actually in 2015, if you want to, it's almost a natural experiment. I mean, obviously, you know, we had certain advantages being a YC company, and so on. But in 2015, I remember there were like, maybe as many as 40k startups in the mix, including some that like you've only heard about, I've only heard about once or twice, but it was kind of a it's kind of the time to start a company in this space. Especially after the success of companies like ZenPayroll, turned gusto and then Zenefits people were just thinking, let's just take every HR benefit that small businesses can have and basically do it with software. And so there were a lot of foreign key businesses and us and one other really the only two interesting startups still around. Like I said, I think the pricing that business model was a big part of it. Also, the go to market was a big part of it as well, today, something like 80% of human just customers, definitely, but the lion's share, come through partners. So we have a sales team. But we also have a partnership team. And actually the partnership team is is kind of our secret weapon, they'll go out and be a partner, partner account managers all over the country, not necessarily in every state, but we have them in the Northeast and the Southwest, right? There's they're sprinkled all over. And their job is to work with payroll companies, HR companies, right and all sizes, benefits brokers, and get to know their sales team and basically help them close clients, but also some given interest. So we act as kind of a support, maybe you're a payroll provider trying to close a new client or trying to renew a client, and you want your offering to be more sticky. Well, you can actually send your client over to human interest, and we'll sell them on your payroll offering as well as our 401k. And getting that right allowed us to reach more customers rapidly and cost effectively. Direct acquisition is great, but on the internet, it can be can be expensive, it's also hard to notch up and down. We were at a point where we felt like our sales funnel was almost perfect, right? When customers came into our funnel, they would always talk to a salesperson very high rate, once they talk to a salesperson that he closed at a higher rate than they would launch their plan at a high rate and they would churn at a very low rate. So it's like this business shows a lot of signs of having product market fit. But we really constrained because at one point we were getting most of our leads through for inbound, maybe through our blog, and SEO, maybe through through SEM and paid ads, I weren't able to grow as fast as we wanted to. It wasn't until we set up a real partnership program in the company that we really were able to step on the gas, because everything downstream of new leads work just fine. We just had to find a way to generate new leads rapidly. And partners were happy to send us over like much of their existing client base because we were making them happier with their offering. So that's really how we stepped on the gas pedal. That's another thing we did right. And then I think the other thing that you know, we did we did really well is we really embrace the tech side of our business early on. There are other 401k startups that were kind of maybe outside the Silicon Valley ecosystem that would tap things like low fees, but they wouldn't really put the effort in into into doing the payroll integration part. Well, they wouldn't put effort into the into the compliance part well, and I think they kind of like use a lot of the marketing and language of these new robo advisors and said, that's us, too. And wasn't enough. You know, you can't just paper over the stuff with marketing. I think one thing I learned in human interest is that there are structural things in the 401k business that make some businesses succeed and some fail. And if you go out into market, basically promising the world but you haven't structurally made it so that you can pay for customer support and operations, you're not going to last and your customers are going to have a bad time. I'd say it's actually borderline unethical because you're promising people a retirement savings plan, right? And you're running the risk of going under as a business in a few years. You know, we're all okay maybe with writing Ubers or DoorDash is or getting DoorDash deliveries where they're subsidized by VCs, knowing the price is going to drop in a few years before retirement savings plans running out that kind of lost money and each client, I would say is borderline unethical and we saw companies try that and fail.

34:57
One thing that really jumps out at me there that's gone. For intuitive is how to have a successful integration and partner strategy, you really need to help sell the person's product, we are integrating it and you guys, like your pitches kind of hey, integrate with us and like will help you improve conversion and sell more. And I think a lot of people that pursue integration strategies today, I see a lot of startups do it, I'm not sure they quite like, have internalized how important that is for that channel to be successful. So I think that's a really great, not an obvious insight and take away, one of the things that comes jumps out at me just like looking at your website, and having talks to me your sales reps to I almost bought your product for company ultimately went with. So this company was a spin out of a bigger company. And so they had like a legacy provider through fidelity. And so we had a plan in place for fidelity and I kind of wanted to move to your product, because you had lower fees was half a percent. And it was like, much more employee friendly and human friendly. But these transitions are always hard, like switching 401 K's is a giant undertaking, and so excited to kick the can down the road to the next cycle. And I'm sure you've you've seen that in part of like your successes, attributed to the fact that you were going after like a market that didn't have a solution in place. The point I was trying to make is, I think you're like a lot of legacy players are quite predatory in their model. And it really stinks. And like your branding around sort of you being human centric, being employee friendly, I think that's been like, at least from the outside in, I perceive that as like a huge differentiator. And I think you guys have done that really well. It's true. I

36:31
mean, there's a lot of shady things that you see in the foreign pay business, there are foreign key providers and advisors that constrain the fund options in a plan to funds that have a high fee. And then they themselves receive kickbacks from those funds. And that's actually legal, it shouldn't be legal, but it is, it's possible to offer 401k without being a fiduciary of a 401k, for example. So you know, we see ourselves as kind of part of the trend of just cleaning this stuff up. And the nice part is if you actually approach this market in a progressive way, using technology first, you can make a ton of money and also do right by people. I always said that the great thing about this business model is that people like money appears in people's pockets almost by magic, right? Because we're still making money as a company, but our customers are paying less. And of course, that dividend comes from technology and cutting out middlemen like advisors and these kind of like ridiculous mutual fund fees.

37:19
So switching gears a little bit. So we know you recently transitioned out of human interest to and are taking some time to figure out what's next. So tell us about what's next. What are you working on? What are you excited about? And where do you see opportunities? And what do you think the future looks like?

37:33
I spent a little bit time after I left few managers just focusing on working with startups because actually, I actually missed it a lot. You know, when I left human interest, it was approaching 200 employees. And it was many days past when it was just me and Roger and like a handful of employees in office working like like mad and working together. And it wanted to just kind of reconnect with that energy and talk to people that are in that stage. Because that stage is fun. It is really fun. So I kind of CO led a batch with this online sort of accelerator called xx happy to plug them they're amazing URL is xx dot team is their website. And they're always you know, they're they're running multiple batches a year, if you're an early stage startup founder and you want to work with amazing mentors in like a really intimate setting, and get 50k, I highly recommend it. It's a great pre accelerator to YC love of our xx companies do go on to yc. And we've been told that it feels a lot like the early days of YC, because it's very small and intimate about 15 companies per batch. So I did that I had a blast met a lot of great founders did some angel investing. And then after that, I started thinking about what I wanted to do next. And as fun as it is to work with founders and do investing, and then kind of have that remove, I felt like I really wanted to get my hands dirty and build another company. And a lot of my time since then has been kind of plotting what that's going to be. There are a lot of really exciting things happening in the world of technology that I think are going to create a lot of amazing companies. And you know, they're the ones that we all talk about all the time, like, you know, cryptocurrencies and blockchain, and VR, AR and AI. But I was looking for the things that I thought were maybe a little bit more nuanced, maybe things that a lot of people aren't looking at. One, for example, is WebAssembly, which is a feature that's been relatively recently activated and all of our web browsers, it basically allows any programming language that supports web assembly, which is rapidly approaching all of them to be compiled in a form that can run inside your web browser at speeds that are basically the same or very close to a native app running on your computer. And they're so fast that people are actually taking server side code and also compiling it down to web assembly and running servers on it. And to me, this is kind of a, you know, a sketch of the future of computing, where basically the programming language you use doesn't really matter. The distinction between front end and back end doesn't really matter. There's just binaries that run in this web assembly sandbox. And this opens up a lot of possibilities to do things in the web browser that people think you need to do on desktop. And we've already seen that as a large trend, right? Like a lot of people told the figma founders, there's no way you can build that inside the web browser. And of course, with some really great engineering and with Web GL, they're great. too, but I think you're going to see things like triple A games, like really high quality 3d games run inside the web browser, I think you're going to see really complex programming environments and development environments like entirely in the browser. And on the web, I think you're going to see things like servers running on AWS, also being able to connect to them and run them inside the web browser. And I just think that's really exciting. I think that's going to create a whole new class of tools and opportunities, not just to build products, but to build algae. So that's something I'm thinking about a lot. It's one of those things that I think is going to be like, like a big breakthrough, sort of like AJAX and what the figma guys are doing. I'm trying to imagine what some use cases might be. So gaming is a big one, any others that you can think bring with in terms of use cases, any kind of high end media editing, right? So Photoshop is still delivered as a desktop app, music production, audio production tools are still delivered as desktop apps. I think that could that could really benefit from being in the web, you get things like collaboration for free, you get the ability to push new code for free subscription business models are better. I mean, we've seen this I mean, the web has eaten like 80 or 90% of software at this point, right. And I think that kind of that last 10% of that high, high intensity compute, compute work, you know, with web assembly can also move on to the web. And I think that that's super exciting.

41:14
Podcast editing. Yeah, we do that today. Yeah. So that's a great one. Are there any other friends that you're trying to pull on while you're exploring new ideas? Yeah, so

41:25
I'm very excited about AI, which, like I mentioned, is something a lot of people are thinking about, but I'm specifically interested in it as a way to kind of change computer interfaces, user interfaces. So there are a lot of startups that were doing they called conversational interfaces, maybe back in like 2016 2017. Remember that that trend? The the phrase was, bots are hot, everyone's trying to build every product as a Slack bot. And it turns out that a lot of those conversational interfaces were not like any better for the user. And just learning the products UI. like by the time you learned the slash commands, you learn how to phrase your request, you just like wasted more time than will take and just find the thing you want to do in a product menu, right. And that's why I think a lot of those companies failed. But we see a lot of breakthroughs. Obviously, we've all seen like the open AI demos, right? GPT three, and Codex, that makes me wonder if you could actually no build an AI interface that is actually useful. And if that actually materializes, I think they'll actually change the way that we interact with computers in a big way. So I as a programmer, right, when I'm writing code, what I'm doing is I'm writing computer code, like I'm speaking the computers language, and then hopefully, it works. And I run it. And then I see if it has the behavior, I desire not to have to go change my code. Again, I think in the future, we might see interfaces, we're actually speaking natural language to computers. And instead of checking our programming code for mistakes, we're checking the computer for mistakes, right? So if I make a mistake, while I'm writing code, it's my fault. I wrote the wrong or the wrong code. But if I phrased my request to computer natural language, and it does a slightly wrong thing, it was a computer's fault. And then I can go in and kind of give it further instructions on what I wanted it to do. So you could imagine something like a spreadsheet. I mean, I've seen really intricate and complex spreadsheets, that human interest, like our you know, business model in Excel is like, you know, 10 tabs, and it's incredibly intricate. I can imagine a world where someday I has an understanding of the spreadsheet. And I could just say, hey, what was our What was our logo churn for July, August and September, and then also give me that number, the ratio of that number to the same last year, month to month, right? The information is contained in there, I could do that calc myself by digging into the numbers. But if a computer had enough understanding, it would be a big win for me just ask that question and get it out. And I think we're now getting to the point where the technology can support that. So another thing I'm thinking of is I'm looking at software. And I'm asking if the software has a really complicated user interface, if it would take me months to learn. Is there a way that AI could simplify my learning of a tool? Could I just ask the question, and then check the computers understanding of my request, where I've been trying to make my mind understand how the computer thing?

43:59
That's really cool. I mean, you mentioned this this when we were talking off the podcast earlier how office has like a million features. And like, I worked at Microsoft for a while. And I know that the sort of stuff people would ask the team for was stuff that already existed in the product. It's like, hey, I want this feature. And it's like, this feature exists. You just didn't find it. I don't know how to find it. And so having a AI for that sort of discovery sounds really interesting. With spreadsheets, I'm always googling how to do a pivot table to do some weird thing. And it's never quite great. So AI driven interface. Yeah. Interesting. Yeah. So how are you sort of approaching that?

44:37
I'm doing lots of prototyping. And I'm trying to work with people in kind of low stakes, collaboration ways. Like, instead of saying, let's sit down and talk about starting a company, I say, why don't we work on something for a week together, and the only goal is to learn something or to build something cool, right? And that obviously has the benefit of helping you potentially meet new co founders, but also just to learn from new people. And I think that's a really great way to spend kind of time in between things. To build lots of little things, there's something really nice about being able to build something to throw away. You don't have to spend all your mind thinking about, you know, legacy code, how's it going to be supported in the future, you know, is UI really going to scale and you get to focus on the one thing you want to learn. That's another thing. I think that that allowed me to get good at programming. When I was younger, that I was pretty prolific, I was always building lots of little projects. And it's been such a treat after being, you know, CTO of a large company, and not really doing any coding, to be able to just kind of code up little things, again, write and play with individual things like the open AI API that I want to learn about. I think that's a that's a great way to learn and explore. Paul, I'm

45:38
shifting gears a bit, we have a few questions that we ask every guest, and we can jump into those. The first is around superpowers. Everyone has a certain set of skills that they're really good at that it feels like play to them, but it's worked for others, people over the course of their career kind of learned to develop and identify their superpowers and lean on those. So I'm curious to hear from you. Like as you start this next chapter, what are some of the superpowers that you've noticed in yourself that you're trying to leverage?

46:05
Well, that's it's a very flattering question. Because I think actually, a lot of what I did after human interest was reflect on things that I'm not so great at. So now I get to reflect on the things that I feel like I was uncommonly good at with my superpowers, which, which is nice. I think one thing I've shown myself, at least to be very good at, is to kind of enter a situation, there's a lot of ambiguity and find a path forward, right? Like we were in a place in human interest, where we were realizing that we needed to move off of that, that third party record keeper, we had to build our own internal software. And it was it was overwhelming, like, we had to go out and investigate existing software packages, see if they were any good, or to design our own. We had to find custodians to partner with which custodian should do. And I'm really proud of the fact that I was able to spend a few weekends thinking and writing docs and coming up with a go forward plan. I think founders are wired a little bit different from most most people in terms of their skill set. And there's totally a startup founders skill set. Many of the employees I hired were much smarter than me much stronger engineers able to go much deeper than me, right, I could give them a piece of code, and they could find five things wrong with it and rewrite it to be just that much better. And that's lovely. But I think the skill that they lacked, which is why they were better engineers, and founders is the ability to just take a step back, see kind of everything in its early stage in its messiness, right, a company in its early stages, totally messy, it's not clear what you should focus on where you should go, a project, like our internal were competing was totally messy. And identifying a path forward and learning to adapt is a totally different skill that I happen to possess. And I think that's, that's something where I've learned I want to put myself in situations where I can use that skill, which is why I would make a very bad kind of software engineer at like, even a medium sized team. Like, personally, I would not hire myself to be engineer number 20 at a startup. But you know, I might make a good you know, founder of that same startup. I think learning what you are and not aren't aren't good at is also a great way to build rapport with your employees is kind of a side point. But when I got to a point where I realized I wasn't the best coder in my company, but I was really good at getting things started and keeping them on track, I had a much better relationship with everyone in the company. And actually, there was a really strong mutual respect between me my VP of engineering who came from this really, you know, impressive background, he met 70 person teams. And you got to a point where we knew what we were both good and bad at. And that allowed us to work together very effectively. So I like thinking about superpowers. As a general thing. I think it's helped me out a lot in my career.

48:31
I love that. Yeah, self awareness is so fundamental to figuring out what you should be spending your time add in what your weaknesses strengths are. Second question is around people that have given you a break. Like, were there situations or times in your career where you felt like things could have gone? One of many directions, but you were fortunate enough to have someone helping you out that changed things for the better? Oh, absolutely.

48:55
I know, I consider myself to be incredibly lucky to admit the people that I've met, like so many people come to mind when you asked me that question. I mean, going back in reverse chronological order, you know, Roger, my co founder, and I really started two companies. And you know, he took a bet on me and his co founder, and it worked out really well. But I'm sure he could have had access to someone a little bit more experienced. But I think he saw He saw the talent and the drive. And also the fact that we worked really well together. And he took a chance on me. You know, I took the chance to write starting companies, anyone is a bit of a leap of faith. But that worked out really well. Other people in my life. I mean, there was this engineering manager Benedetta at Mozilla who hired me, at Mozilla, when I was an intern on a different team. He actually plucked me out of that team and gave me a full time offer on his and that was really great because it got me to San Francisco, got my first taste of working in a software company professionally. And it made me realize I didn't like being an employee. Right. But you know, that was that was incredibly helpful and I'm very appreciative for him for him doing that. Back in college. Many professors took a bet on me. I went to this this kind of unorthodox school called Hampshire College where The classes you take at the school have no grades, no majors, but it's not much of a computer science program. So you can cross register at neighboring schools, I took classes at Amherst called just one of the schools, you could cross register. And Amherst College is obviously very, like academically strong school, very strong reputation, right. And some Amherst professors were a little suspicious of Hampshire College students that would register for their classes because they didn't have the state admissions criteria, prerequisites didn't apply to them leave a reputation of not showing up to class or showing up late. But there was there was one professor in particular and hammers to was really encouraging of me and kind of saw that, you know, I had a talent for computer science. And challenge me, I remember he taught a compilers class, and I just learned so much so much from that class. And, you know, sometimes see him after class, and he's kind of give me new site assignments and new ways to think about it. And I'm very appreciative of that. So I think it's really great to reflect back on that, for me, I'm just trying to reflect on how can I give people a break now, right, I'm lucky enough to be in a position where I can help people out. So when founders pitch me for an angel investment, I think, like, you know, well, my angel investment be the thing that can make this company success. Maybe I can introduce them to people, that would be lovely, right? Sometimes I talk to employees that have left human interest, right. They're trying to figure out what their next the next job is going to be. And I'm happy to give them advice and refer them to people that helped build that network. Many people are hired at human interests were unorthodox in terms of their resume. I remember I hired someone she worked in like game development for a while she went to Full Sail University studied game development, realize she didn't like game development. So her resume was like she worked. You know, this, I think she worked. She had this this degree in game development. And then and did some like PHP web development. And I brought her in, and she completely knocked our socks off. And she's still there's just a really strong engineer, and unlike, so happy that I was able to find her. You know, obviously, when you close some, you know, hotshot Stanford MIT graduate, that looks good. And at that feels good. But for me, the the most fulfilling part of hiring is when I hire someone whose resume looks like utterly, kind of like, utterly like normal, but you you kind of see like one or two tidbits, it's like a sign of talent here. This person is working on things in their own time. There's like a spark here, and you meet them. And they end up being like one of your best engineers, and knowing that I was able to pluck someone like that out of the out of the pile of resumes and kind of help them out. And the company is benefiting from kind of undiscovered talent that makes me so happy.

52:23
That's fantastic. Last question is around books. Are there books you've read recently? Or maybe even not? Recently, I've been lastingly. impactful.

52:32
I've recently listened to a book called The courage to be disliked. And if you're familiar with this one, love, love that. Remember that twice there. Yeah, it was recommended to me like five times and like, Okay, I got I got this book, like, there's clearly something here. And it's just a really good tool for looking at yourself, your psychology, or it's a book about psychology, looking at yourself and trying to think about what actually drives you, right? Like, what are you actually working towards? Why do you act the way you act? It's very focused on the here and now in the present, but not focusing on the future or the past, but the present. And I found it to be really intellectually interesting and helpful, I found myself referring back to it a lot. You know, ever since I started listening to it, I'm about halfway through right now. So that that book is currently on my mind. Sure. Another thing that's that's helped me out a lot, it's not really a book. In fact, it's not a book at all. But it's something that I think is a useful, a useful thing to do, is, about three years ago, I worked with with a coach and I wrote out my own personal values. And since then I put it on my bedroom wall, and I've heard of it regularly. And it's really been kind of a guide for me. So I have I have 15 values of sounds like a lot, but I could like name all of them off the top of my head. And whenever I find that I have trouble making a decision, I look at my values. And I think, Hmm, how should that guide me, because the values are kind of an articulation of who I am at my best. For example, one of them is being comfortable with discomfort and other is acknowledge my privileges and headstart in life. And that's been that's been a real guiding light for me. One of my goals for for the coming year. I'm a little a little shy, a little vulnerable about it, but I think it's the right move is I want to just publish it on my website. I think if you go to pulse Avaya comm just gonna see a list of my values, and they'll put a little link in my email signature or something like that so people can take a look. I think that kind of accountability and consistency of knowing kind of who I want to be glam at my best over over the past few years has been very helpful. And by the way, it was inspired by this thing, written by John Perry about Barlow was a very interesting guy, one of the founders of the EFF as well as a musician for the Grateful Dead. He wrote the principles of adult behavior on his 30th birthday. So if you haven't read those, I recommend reading those as well. It's kind of his own list of values, and actually inspired me to write my list around my 30th birthday stuff.

54:50
I love the idea of writing down principles. It gives you clarity on why you're doing the things you are doing and maybe why you shouldn't be doing some of the things that are doing so. Well. This has been a fun That's it conversation I really appreciate how thoughtful and reflective you are and I know I've learned a lot from this conversation and I'm sure others will too so thanks so much for coming on today

55:10
of course a lot of fun My pleasure

Transcribed by https://otter.ai