Episode Transcript
[00:00:00] Speaker A: Foreign welcome to the Victory Show.
Hey victors. Welcome to this episode of the Victory Show. If this is the first time you're joining us, I'm Travis Cody, best selling author of 16 books and the creator of bestseller By Design. I've had the privilege of helping hundreds of business consultants, founders and entrepreneurs write and publish their own best selling books as well. And along that journey discovered a really fascinating pattern. A lot of businesses hit a revenue plateau, usually around a million dollars a year, and they struggle to break through it. So on this show, I sit down with some of the world's most successful founders, CEOs, leaders and business owners to uncover the strategies they use to overcome those plateaus and scale their business to new heights so you can do the same. So get ready for some deep insights, actionable takeaways that you can implement in your life and business. Starting now. Today's guest sits right at the intersection of data, labor markets, and the future of work. Ben Zweig is the CEO of Revelio Labs, a workforce intelligence company that's transforming how businesses and investors understand talent and human capital by unifying and enriching massive public employment databases. Revellio Labs is building a universal HR database with a simple but powerful mission to make labor markets more sophisticated and more transparent. And in addition to leading Revelo people, Ben teaches data science and the future of work at NYU Stern, sharing his expertise on data and AI and how that's reshaping careers and industries. Now, before founding Revelio, he managed workforce analytics projects at IBM's Chief Analytics Office and worked as a data scientist at an emerging markets hedge fund. With a PhD in economics and research in occupational transformation and social mobility, Ben brings deep insight to one of the most urgent conversations of our time. How we hire, work and grow in a rapidly evolving economy. Ben, thank you so much for being here. How timely is our conversation?
[00:01:58] Speaker B: Oh, my gosh, very. Yeah. And thanks again for having me. This is, this is quite the intro.
[00:02:02] Speaker A: Thanks. For people that are not aware of why what you built is so necessary, walk through some of the struggles that companies and corporations are having in terms of how it comes to employment data and why Revelio is solving that problem.
[00:02:16] Speaker B: Yeah, yeah. I mean, even just to take a step back, like, I mean, just if you think about the economy and how it breaks down based on its resources, you know, the economy is essentially made up of labor and capital. And if you, if you break that out, you know, about two thirds of economic output is attributable to labor, about one third to capital, and there is a science for allocating capital. The field of finance, you know, and finance is super sophisticated. We've all heard of it and participated in it. And we know that that's like a huge part of our economy is allocating capital scientifically in a sophisticated way. And there, there's all, there's a whole world that supports that. You know, we have accountants which exist essentially to like categorize things related to capital. And when it comes to labor markets, which is twice as big as capital, we don't have anything close to that. I mean, we don't, we don't allocate labor efficiently. And this is our time, this is how, how we spend our lives. I mean, what could be more important than that?
And you know, I mean, how do people choose their careers? You know, they, I don't know, they ask their uncle or something or like, you know, their friends, parents and you know, it's, it's so network driven, it's so not rigorous. There's no research, there's no anything army of, you know, accountant type people that, that help us categorize, that helps categorize labor and how we should think about how labor markets are organized.
So you know, job seekers are, you know, totally in the dark. And companies also, they struggle to find the people they need or even to find their roles or their requirements that they need to grow the business. It's very, very reactive.
You know, all of this is happening in the dark. So, so that's, that's kind of why we are trying to reveal what's happening and bring, bring more clarity to these markets.
[00:04:03] Speaker A: I love it. So when you talk about being transparent for, for the labor market themselves, not the corporations, but people in the labor market working for jobs, how does help them?
[00:04:12] Speaker B: Yeah, I mean one thing is, is they make more informed choices. So it, you know, one way to think about this is that you know, when someone like I have a 4 year old daughter and you know, she kind of knows what jobs are, she's like, oh well you know, mommy's a feelings doctor. And you know, there's like a teacher and like a cook and like a singer. You know, like, like she has a very, a very surface level understanding of like what jobs are even out there in the world.
And then as you go up to, to high school and even college, like you learn about other things, you know, you're like oh well now I know about lawyers and you know, and like, you know, I've kind of heard of marketing, but don't really know what it is. Our understanding of the occupational landscape is still pretty limited. And then as you get out into the workforce, you get an understanding of like a pocket of that and, and to get, you get more knowledge of like what is a product manager and what do they do? You know, that's not something we learn about in high college. And, and you know, you get more visibility into this, into this landscape that you've only had a very narrow view of. So I think just, just giving people more information to do research about. Like what is this job that you might have heard of or might not have? What does it do? Does it match your interests, like, or your skills? Like what does the day to day look like? Do people like it? Do they stay for a long time? What does that lead to? You know, how, how will this help you pursue your goals? And I mean that, that information is all out there like that, that's, there's data about that. It's just not organized.
[00:05:48] Speaker A: Yeah, you know, it's fascinating because there's been industries where I've been and where I've worked in eight or nine years and then you meet somebody and, and you, they tell you what they do in that industry and it's like, wait, you do what? Like, I didn't even know that. And like, yeah, I've been in that industry eight years and they're doing this thing and I'm going, that sounds way more interesting than what I've been doing for eight years.
Right. So what I love about this is the fact that as we're moving forward with, with you know, new generations come into the workforce. Like it sounds like this data really opens up the opportunities for people to find not just jobs that are a good match for their skill set, but that would also, they would find fairly fulfilling versus I'm, I'm becoming an accountant because you know, I'm, I'm like one of my friends, he's like, I'm from India and from India you got three choices. It's accounting journey or a doctor. Like all you have to do. Why? Because that's what the culture says you do. Right. And then they get in it and they're miserable. Right. So we're talking about now being able to come in to find a job that you're actually good at and, and you actually enjoy. So that's, that's remarkable.
[00:06:44] Speaker B: So let's out there, I mean I, I think you know, every individual wants, wants to find the career, you know, that, that, that maximizes their, their own objectives. And also every company is also struggling to find the people they need. So, so there's, you know, there, there's, there's a lot of need on both the supply and the demand side of the market. And that, that has to happen somehow. And now it happens from like gut feel and no data, you know, so, so facilitating that, that match just makes everybody better off.
[00:07:12] Speaker A: Yeah, I love it. So let's take a step back then, because you were, you were working in data with, with IBM and obviously you're exposed to, you know, a company scale that most people can't fathom. So what was it in that time with IBM that opened your eyes to this need for data? And was there a spark that you went, oh my gosh, like somebody needs to solve this problem?
[00:07:35] Speaker B: Yeah, so, so when I started day one, I met with the VP of the group and, and found out like my teammates and all that and what he said, you know, and I was, I was working on the workforce analytics team. We were, we were, man, we had this mandate to optimize the IBM workforce, which was 400,000 people. You know, that's like a lot of people.
You know, the way he said it, he was like, you know, 75% of IBM's expenses are people, and it is the least understood part of the business. Wow. And I just heard that and I was like, holy shit. Like, this is really important. Like, what could, you know, this is so underserved and you know, I'm like, here to do important work. It was like, really motivating. And, you know, we worked on tons of projects related to, you know, optimizing the IBM workforce, you know, reskilling and retention and promotion strategy and site selection and like, you know, like sourcing talent, all these, these kind of cool projects. There was also something that, that I, I got discouraged by and that was really, that we, we analyzed IBM data all day, but we had no idea what was happening outside the four walls of the company.
And you know, coming from economics, you know, we, we, we have like a framing of what strategy is. And, and if you could, if you could distill strategy into one word, it's differentiation. Like that, that is, you know, strategy on one foot is about understanding how you're differentiated. And we had no clue how we were differentiated. And no company did. You know, IBM, I think at the time was like at the cutting edge of people analytics and workforce data. But, you know, never in a million years did we think we could, we could get access to like the HR database at Oracle or Accenture or something. It was hard enough to get data internally. So then that then I was thinking like, this is. And meanwhile, at the same time I was kind of getting involved in this people analytics community in New York. And it's just like a really wonderful community. They have events and it's just, it's just a really, a really nice network. But in those, in those conversations I was starting to see that, you know, the projects I was working on, other companies were also working on them and like doing it more or less the same way. And I was just like, look at all these smart, motivated people doing basically the same work. Like, what a waste of talent. And I thought, you know, how, how could, you know, how could this be centralized? Like, wouldn't it be nice if there were kind of one, you know, entity just front loading all this effort, doing all the mappings and hard work to get it to make sense and kind of came across this idea that like, there's a lot of data in the public domain that could be used. So let's say you are an employee at a company, you are a record in their HR database. It says your title, your start date, you know, your reporting structure, all this stuff. A lot of that information is also mirrored on your Resume or your LinkedIn profile or something like that.
So, and that's, that's accessible, that's in the public domain. So if we could take, you know, the full universe of employment data that is in the public domain, we can sort of reconstruct or at least approximate, you know, the HR database for every single company, regardless of whether or not we have an affiliation to it. So that just, that idea just seemed so powerful and there were tons of problems. I mean, it's like a massive amount of data. It's like a total mess. Like, you know, it's free text and you know, it was just like a lot of hard problems to work on. But you know, that, that kind of got exciting. And also, you know, in 2017 when we were conceiving of this, that was actually just three years after the first like large language model was created. So, so that was called Word2Vec. That paper came out in 2014 and we, we started experimenting with like using what was the cutting edge large language models at the time to apply them to, to of resume and profile data and job posting data, which is of course, you know, very messy. But these models were really, really well suited to kind of, you know, create a structure around, you know, create taxonomies and job architecture around this messy employment data. And when that started working well, I was like, okay, this is going to be Something I don't know who we're going to sell to. I don't know how we're going to make money, but I got to quit my job. That was that.
[00:11:45] Speaker A: Here you're working with, you know, what a lot of people considered the golden ticket for a job working for IBM. Right. I mean, that's a big company. You know, perceived security there. So to leave, that was, it was a pretty big leap. But what I'm loving about this is you're discovering, you know, in the early 2000s, there was the phrase that was coined about, you know, we're in the information age.
[00:12:02] Speaker B: Yeah.
[00:12:02] Speaker A: And by 2017, we're realizing that that created a new problem, which was the information overload phase. Right. Where it's like information now. There's just so much hitting at us. It's like we, we just can't, we can't process it all. Your approach to launching your startup was, was pretty, I thought, pretty creative and unique. So let's talk about your first year. It was you. And then you had hired four, four local interns.
[00:12:27] Speaker B: Yeah, four interns. I had gone to high school with someone who worked as like the placements officer at the, the Master's of Data Science in, at Columbia University. I, I just, you know, I just asked her like, hey, you know, who's, who's good, who's underrated, you know, and like also, you know, available. She, she, you know, hooked me up with, with some, some people and they were amazing. I mean, they were great. And yeah, basically, you know, sat down with the four of them every morning and started, you know, working on problems, hands on keyboards, you know.
[00:13:01] Speaker A: So how long did you have them? Just three, four months.
[00:13:03] Speaker B: Yeah. Yeah. I mean, actually two of them were able to stay beyond that into the semester, but it was, you know, after that we had to find, you know, other people and we're always kind of like on this like, intern treadmill, which.
[00:13:19] Speaker A: Creates its own challenges.
[00:13:20] Speaker B: Yeah. Yeah. Really, for a long time. I mean, we didn't really get any hires until sort of the, the tail end of 2019, and we started kind of mid-2018. So, like a good year and a half.
[00:13:34] Speaker A: Year and a half before your first official for hire.
[00:13:36] Speaker B: Yeah. And then, and then we had, you know, and then there were three of us for like, I don't know, a half, like seven, eight months. And then there were like, you know, five of us for like the next, you know, six months. You know, so we were like, really small for a long time. Yeah. Just.
[00:13:54] Speaker A: Yeah, so. And I love that Right. I think, I think that's a pretty, the pretty typical journey for a lot of founders. Bootstrapping as much as you can. So let's talk about, you know, once you hit your first 10, 10 full time employees and you're starting to scale because I know you're, you're way beyond that now. Let's talk about for you as a founder, as a data scientist kind of working on this thing. Basically you, you've got interns but you're kind of in charge.
[00:14:18] Speaker B: I also had a co founder by the way, who wasn't ready to quit his job when I was. So like he joined basically kind of a year and a half in or.
[00:14:26] Speaker A: Then it started to shift.
[00:14:27] Speaker B: Yeah.
[00:14:28] Speaker A: Yeah. So then how did you, for you, like I'm more interested in, is like, you know, because being data scientists, being like boots on the ground working, eventually you're getting to a point where now you've got, you've got, you're managing a team and you're doing, you've got to do more owner stuff versus boots on the ground stuff. How, how did you personally deal with that transition?
[00:14:46] Speaker B: Yeah, I think, I think for us, like, you know, I, I, I had to like be involved in the technical side of it.
[00:14:53] Speaker A: Sure.
[00:14:54] Speaker B: That that's where I like get energized also. Like, I really like, you know, sitting and writing code and stuff like that. But, but also we needed to do sales and you know, look for maybe raising money or something, you know, so we were doing this, you know, the commercial side of things also needed to be served and I just kind of, you know, had, had come to the realization that like I, you know, I could hire, I feel very comfortable hiring strong data scientists. Like, I know who's good, I know how to like source them. I don't feel comfortable hiring good salespeople.
And that's still the case. But I think, I think I just, you know, it was, it was a better use of my time, you know, hiring data scientists and engineers and, and working on the commercial side of the business.
And so, so I think I just needed to kind of manage the data science and data engineering teams.
[00:15:52] Speaker A: Was it hard for you to let go of being like kind of in there writing the code and realizing like for right now I've got to hire other people to do that?
[00:15:59] Speaker B: Yeah, it was kind of hard to let go. Yeah. I mean, even still, you know me with these teams and I'm like, oh man, I wish I were like in this, wish I were like building this in a different way or Something. I mean, they're amazing. Like, they're. They're. Not that I am at this, but. Yeah, I still kind of feel like I like, you know, getting my hands dirty. And, you know, one thing we've instituted as a company, which I think is like, a really nice thing that we do, we have what we call, like a project week, where, you know, every week a different person kind of takes off the week and just works on whatever they think is cool and interesting, smart.
[00:16:34] Speaker A: That's great.
[00:16:35] Speaker B: Yeah, so. So people love it. I mean, it's like, you know, it's. It's a popular thing that is also like, just a really nice thing that sort of scratches the itch.
[00:16:42] Speaker A: I love it. So for you as a CEO, then what have been. Been the challenges again? You know, stepping away from scientists and stepping away from being professor into CEO. Sounds like you've. You've crossed that chasm and you're very successful with that. Can we talk a little bit about the challenges you went through to get to that point?
[00:16:58] Speaker B: It was, I mean, you know, there's a lot of chaos and just like, things go, go kind of, you know, things are. Things are messy even when they seem like they shouldn't be. So I think, I think I'd sort of, you know, gotten much more aware of, like, what can go wrong in an engineering pipeline. And at least for the beginning, we really, we were promising something that was pretty ambitious. We were telling clients, you can. You can get visibility into companies that you don't have an affiliation to, and you can understand what's going on with their workforce. And that's, That's a big. That's a big statement. And I think people were enamored by the possibility of that. So I think, I think getting interest wasn't. Wasn't the problem for us. The problem was not embarrassing ourselves when we showed them the data. You know, so the first, like, few years we were really.
We were really, you know, in this, like, embarrassment minimization paradigm. We really felt like we got to put something in front of them and, like, it's got to work. If we could just get it to, like, be functional and work and not have problems, then, like, we are ahead of the game. I mean, now we've sort of crossed that. And now, now we. We've. We've gotten, like, really stable. I mean, there's still problems that pop up, but we, you know, now we're. We're more in the paradigm where we're like, trying to produce more cool things and we, We've Sort of escaped this problem where, where it's just like, you know, issues that we, you know, landmines that we have to like get around. But for, for a while I was just like, damn, stuff is like really hard. Like we are really hard problems. And yeah, I was just like, I don't even know if this is possible. Like, I think when I started I kind of had two, two, two concerns. One, is it possible to do what we're doing? And another, is it defensible? Like, like could anyone do this? And I, I was very concerned about defensibility, about competition and the more time I spend in it, I'm just like, you know, like I'm not at all concerned about competition. Like, like there's so many, so many barriers to, you know, getting into this. But I'm really concerned about whether it's possible to do it well at all. You know, it's just like staying at the frontier of technology and engineering has been really important for us.
[00:19:11] Speaker A: That's fantastic. So now you're where you're at. You've got, you know, a little bit of stability, your model state. You could be harnessing all this data. What does the next three to five years look like, not just for Revelio Labs, but in terms what your vision is for how industry is going to actually use this data?
[00:19:28] Speaker B: Yeah, yeah. So it's, it's big. I mean, I think, I think just about where we are at. You know, we, we have 67 people right now and the vast majority are engineers or data scientists or economists. You know, we, we are very much an engineering culture and we don't really have a go to market function. We don't really have sales marketing. I mean we have a little bit, you know, but it's, it's, it's fragmented. You know, there's no, like, we don't have account management, we don't have, you know, partner channel strategy. We don't have SDRs and BDRs or an outbound motion. Like, like there's so much that we just don't do. And, and I think, you know, in some way like that's reflective of how, how I like to think about the world. I'd like to think, okay, you know, the best product wins and like, you know, and, and the, the selling it, you know, will all kind of wash away. And the, the best way to make a lot of money is to deserve a lot of money from having a good product. And, and I think, I think I'm sort of getting, you know, you know, sort of, you Know, shifting a little bit to thinking, oh, actually it is really important to like, have a strong presence, be in a lot of markets, like engage with communities, get to know how they think and talk to them. And, and so, so I think we're, we're, we're putting a lot more effort into like rebalancing our team to, toward, you know, just like having more of a commercial presence.
[00:20:45] Speaker A: Sure. The market marketing awareness is a big deal.
[00:20:48] Speaker B: Totally. Yeah. And we've been, you know, not doing much of that. And in terms of like, what, what we can enable, I mean, I think, I think there's, you know, there's a few ways to think about, you know, how to build a company. I think, you know, I think the, the startup orthodoxy really goes like this. You know, it says, you know, identify a customer need and just like build whatever you have to build to like solve that need and solve their problem and be the painkiller that like addresses a client's issue, make them super happy and bada bing, bada boom. You know, this is like the lean startup methodology. And I think we are not that kind of company. And I think we are a little bit more heterodox in a way where instead of, you know, identifying on the need, we are, we are building this core asset. So we are building something that we think is extendable to a variety of different needs. So now we have this like really nice, neat, curated workforce data that can like, be used for all sorts of things. It can be used for evaluating companies, it can be used for compensation benchmarking or retention analysis or whatever, you know, a whole bunch of things. But you know, we've got this, this core technology at the center of it that is, that is like really our, our main asset. And I think, I think the world we want to see with that is a world where this data just becomes ubiquitous across a variety of different workflows. So the, the analogy, the, the analogy I like to use is something like a Bloomberg terminal. Like in capital markets, you know, everyone's got their day job and their Bloomberg terminal right next to them, which is just like all the data that they possibly ever need to do what they need to do in their day job. And I think we need a Revelia Labs terminal for every HR professional and anyone analyzing people to just have right next to them, to be their companion, to just have the data they need in whatever their workflow is. So that's the, that's the vision we're trying to build toward. I don't know if we're Going to get there in three to five years. But, you know, I hope we get there, you know, in our lifetimes.
[00:22:43] Speaker A: I love it. Somebody's listening to this episode or reading this chapter in the book, they're like, man, this, this sounds amazing. Like, first and foremost, who's sort of like a great fit for Revellio Labs and the data that you harness. Like, is there a size of company that's like you. The guys just knock it out of.
[00:22:58] Speaker B: The park for not really size of company. I mean, we serve like a variety of different markets. So once, like, we sell a lot to hedge funds because hedge funds are in the business of evaluating companies that they're not affiliated with. And, and we are, you know, giving them, giving them a deep view into that. So that, that's one market for us Academic researchers is another market. But actually the, the, the biggest and fastest growing is really corporate hr. And within corporate HR, there's a whole variety of different, different end users. So there's talent intelligence, which is really interesting. There's people analytics, strategic workforce planning, compensation benchmarking, whole bunch of different groups.
It's the problem they have is really that they need to organize their workforce. They need to have a good understanding of their occupations, their skills, their work activities and how to, how to do kind of organizational design. And, you know, it touches everything. So I think, I think it's a nice fit for us because we're so focused on that foundational unification and language and, and that's really the part of HR that focuses on like the data Foundation.
[00:23:54] Speaker A: I love it.
[00:23:56] Speaker B: Yeah.
[00:23:56] Speaker A: So how, so how does someone find.
[00:23:57] Speaker B: Revellio Labs and connect our website, Reveliolabs.com, linkedIn for sure. I mean, that's where we have the biggest presence. You know, I'm active on LinkedIn and we have a popular newsletter and you.
[00:24:08] Speaker A: Were saying earlier you're working on a book. By the time this episode comes out, it'll probably be well down the pipeline. What's your book? What's the title and what are people going to get out of it?
[00:24:18] Speaker B: Yeah, it's about job architecture. So this is.
[00:24:20] Speaker A: There you go.
[00:24:21] Speaker B: Yeah. So the title is Job Architecture Building a Language for Workforce Intelligence. It's coming out with Wiley. So. So this is. Yeah, this book is really about, you know, how to think about taxonomies related to jobs, occupation skills and activities, how to think about taxonomy, why it's really important, like I mentioned before, you know, it touches everything. And you know, if you have good taxonomies, you can do good work. And if you have bad taxonomies, which most organizations do, your work, output will be bad. So, you know, I think at the end of the day, so much, so much of our output and so much of our analysis rests on how people are organized and, and into, into, into buckets, into categories. And, you know, taxonomies really are at the root of any science.
And, you know, if we want, if we want labor markets to be scientific, we need good taxonomies.
[00:25:10] Speaker A: I love it. Well, this has been a fantastic conversation. Thank you so much for your time. I really appreciate you taking time out of your day to be here.
[00:25:16] Speaker B: Yeah, thank you so much. This was so fun.
