Nerdy: Behind the Tutoring Surge with CEO Chuck Cohn

Welcome to the future of education. I'm Michael Horne. Looking forward to today's conversation with Chuck Cohn, the CEO, founder of Nerdy and popularly known Varsity Tutors, one of their businesses. Looking forward to today's conversation because tutoring has become such a hot spot during the pandemic as the recognition has grown that tutoring really is the way, the best way for any individual to learn anything out there in the world from K12 to higher education to in the workforce and trying to improve their lives. And what Nerdy has built I think is intriguing on many levels as we'll learn more. So I'll bring Chuck up to the virtual stage. It's good to see you, Chuck.

Great to see Michael. Thanks for having me on.

Yeah, you bet. Let's start at the top with this. I'd love to hear your personal story about how you got into founding a tutoring business to be begin with because you did this when you were pretty young. So I'd love to hear that backstory first.

Sure. So I'm 15 years in at this point. I started our consumer facing business varsity tutors 15 years ago based on my own personal experiences with tutoring. Some of them incredibly positive growing up, some of them actually quite frustrating where I couldn't find the help that I needed for a couple of big exams. And when I was in college, I was studying finance and entrepreneurship at Washington University in St. Louis and I was about to fail this Calc 2 exam. I didn't understand any of the material, couldn't click, tried to find on campus resources, tried to get help from the professor, looked online and just couldn't find anybody with the requisite experience to help me in time for this, this final exam. And two of my good friends who were deeply academically gifted and had great subject level knowledge and were passionate about math in this case, were kind enough to take a couple hours of their time to help me learn the subject. And I ended up, I think I got a B rather than fail the actual exam. And I had this epiphany that finding the help that you need shouldn't be so hard. And that was the genesis for the business. So I borrowed a thousand dollars from my parents, created a three page website, shook hands with, you know, my two friends, asked me if they'd be interested in tutoring and I'd be responsible for finding students in the local community. And very quickly we started receiving just tremendous feedback on the impact that it was having on students, the extent it was resonating as a solution with parents and consumers. And over time came to realize that this was a very, very labor and data intensive industry. And that is why other companies didn't exist or hadn't scaled and started investing in the underlying technology infrastructure so that we could deliver high quality live learning at scale in a way that hadn't been done before. And so the original business was an online platform for in person tutoring. In 2013, 2014, we launched what we call our live learning platform. And I became convinced there was an opportunity to fundamentally transform how people access expertise in live instruction. Started raising institutional capital from consumer venture capital firms and built what nowadays is the largest online tutoring and live instruction platform in the United States, with the ability for students to learn across 3,000 different subjects, four different learning formats in one comprehensive destination. And so the business has evolved a lot since then. We launched an institutional strategy this past year, but that's kind of the founding story.

So let me just make sure I understand that, because it actually started then as an offline tutoring business with an online platform that enabled the scheduling and the matchmaking in effect, but in localities, and then 2013, 14, you switched to saying there's a larger opportunity here to do online tutoring that is live synchronous, matching people on demand as they need it. Is that the big shift that I understand that, right?

Yeah, yeah. So in 2013, Google and Mozilla started kind of negotiating with one another on creating a protocol for sending audio and video packets peer to peer in a browser based setting. That became the WebRTC platform that we're using here today as an example on streamyard. And it was so enlightening, the idea of building audio and video into anything that you could possibly envision building, that we became convinced that you could digitally enhance expert interaction with learners in a way that just hadn't been done before. You could remove geographic barriers, you could get somebody the help they needed much, much faster, and you could pair them with not just the right person within a couple of miles, but within thousands of miles or worldwide. And so that's what the kind of epiphany was. And we launched it. The feedback was incredible. In those early days, it was super unstable. There was all sorts of defects and deficiencies, but it was clear that there was something special here. And that's what caused me to think, okay, we need to go out and recruit a world class management team and product and engineering organization and other functions that can help us scale this. And then each and every year, a higher proportion of customers, when given the choice between in person or online would choose online. So when Covid hit, we were actually up to about 80% of customers given both options choosing online. We'd allowed product quality to drive adoption and then shifted to 100% online and started leaning into additional learning formats like online classes, live streaming, adaptive self study and you know, and other ones that were all kind of together in one spot.

Very cool. So, so I want to just dig in a little bit more because when I wrote my first book, disrupting class in 2008, we looked at all these areas of what I would say, non consumption. Right. Where should there be markets that don't exist today because people can't get in for affordability, accessibility, whatever it is. And tutoring seemed like the biggest one in the sense that, you know, the stats that I kept seeing were 1 out of 5 students or individuals that need tutoring actually get access to it. That leaves 80% if you will, non consuming. At the same time, it seemed like every few years there was a new entrant into the tutoring space that, you know, some would go B2B, someone go B2C. Most of them started trying online increasingly. I think that started with like Skype or WebCT sort of connectivity. You obviously built your own platform. I'd love you to go a layer deeper on what, you know, what you found to be the distinguishing features of what you've built and why you've seen that success into a 150 million plus revenue company compared to the others that maybe jump up, get some traction and then sort of stall out and leave a very diffuse marketplace, if you will.

Sure, that's a good question. And I think to. Maybe the first part of your question oriented around markets that didn't previously exist and why they didn't exist. There was an inherent discovery problem in this market where it was very difficult to go out and find a solution, find a tutor, find an expert, find a that had the requisite expertise that you needed, great communication skills and an interest in actually helping you. And that's where the Internet and technology could just fundamentally transform and solve that discovery problem. So I think that's, that's kind of one basic element. The second is you need technology to scale and really instrument and capture all the data so that it can then inform the personalization of the experience. And that's another thing that you know, that we've been able to accomplish and is possible through an online platform that just isn't possible to the same extent in person. So those are kind of two different foundational elements that I think were present in the beginning. But there's this famous study that was done In, I think, 1981 by Benjamin Bloom, which I'm sure you're familiar with, everybody in education is familiar with, that basically documents the impact that tutoring can have on students. And it shows that students who get one on one tutoring from a subject matter expert over a extended period of time with the same individual who's then capable of changing and optimizing the learning in real time based on how the student is or isn't understanding it ultimately can perform two standard deviations better than classroom preparation. The challenge, despite that insight, is that historically it was so labor intensive and so costly that nobody could figure out how to scale it. But we're now to the point where online platforms exist. And we built an online platform that allows us to deploy tutoring and other forms of small group intervention at scale across large population groups to an extent that was never possible previously. And that's something that we're focused on right now with our varsity tutors for schools, institutional efforts where we think we can just have a tremendous impact on helping students get that intervention that drives disproportionately good outcomes, but do so in a way that is very efficient.

So I want to stay on this before we. Before we switch to the district piece, because I want to go there next, because I think it's a fascinating area right now. But before we go there, just stay on one more beat. Because other companies presumably had access to these same technology enablers, right, to start matchmaking technology to better personalize the matchmaking and things of that nature, help solve the discovery problem, but also get better matches. What made you all or what distinguishes you all in that market from some of those other efforts that are out there in your mind that makes that work better, for lack of a better word.

So people come to us looking for an expert in the truest sense of the word. And while it's easy to describe, it's incredibly hard to deliver at scale. And what I've observed is other companies that have come and gone just haven't been willing to put in the incredibly hard work that is necessary to figure out how you maintain that super, super high quality bar, but do so in a way that's very efficient and not labor intensive. So once upon a time, I used to interview each and every prospective expert for the platform at a coffee shop. And it would take three hours, and there'd be a diagnostic tests and a mock lesson plan and transcripts and grades and test scores, and all this stuff. And if they were really good, they'd get access to the platform, be able to work with students over time. But that was very, very labor intensive. And we've put incredible energy into that initial vetting process. And then once somebody's on the platform, we actually use machine learning algorithms to match them with the tutor or expert who's right for them, factoring in over 100 different variables. And that personalization is something that is profoundly important. And there's lots of companies that have thought that the market need was in kind of this hyper transactional last minute request. We have those solutions too. But the hard part is it's all about human connection and a consumer model. Unlike if you think back to your school days, you had good teachers, you had bad teachers, but no matter what, you had to show up to school. And when you were in college, you had good professors and you had bad professors, but no matter what, you had to show up to class. And in a consumer model, you can opt out at any point. And so one of the things that we've tried to do is create this holistic experience that involves personalization technology that adds value, that digitally enhances that experience, and do so in a way that is markedly better than what you get offline, than you could get without the system in place, then the platform in place that really adds to it. And so we think about the holistic customer experience as opposed to thinking that our platform is the user interface and our live learning platform, or a matching algorithm, or the ability to get great tutors or adaptive diagnostic testing system, or the fact that we actually have a customer service team that will pick up the phone when you call and ensure that you get a great experience. So they all come together to create experience that consumers elect to continue using. Which I think is the big difference.

No, that's. That's a significant difference. It also allows you to build a moat, I would imagine. Right. With effects that are captured then where people want to stay in that network. Let's shift to the district conversation. You alluded to it earlier. Obviously, federal government has put a lot of funding into districts over the last couple of years, more to come over the next year or two. But we've also seen an increasing interest in tutoring for the reason that you said the Benjamin Bloom study, but plenty of others that have found large effect sizes from tutoring. And as people are worried about learning loss and topics of that nature, there's a sense that tutoring might be the only answer to help students get back on track, perhaps get on track for the first time, I'm curious what you're seeing in the marketplace itself. What are the big trends? And do you think it'll be sustainable, this interest in tutoring beyond the current federal funding, or will it drop off after the cliff, if you will?

Sure. So there's been recognition for a long time that tutoring can have a profound impact on students and outcomes. And there have been other big programs in the past at a national level, like no Child Left behind and SES tutoring and a few others. But what's different this time is a couple of things. So, you know, first, over the course of the last couple of years, and I think this was kind of implied in your question, there's been just profound learning loss occurring within schools. And it's really sad and it's devastating. And fortunately, our elected officials have raised it to the point where there's now actually a big focus on it. And there's a recognition that tutoring with a subject matter expert over a current period of time can really help students. So 10 years ago, you couldn't deliver that online. It was very labor intensive. People were in cafeterias. There were, you know, problems with busing students back and forth. And now you're able to do it online in a way that's way more seamless and scalable and efficient. And the entire experience can be recorded so schools actually can know exactly, you know, what's happening. And there's accountability there, you know, that I think is really healthy. And so when we think about this model, we're trying to partner with schools in such a way where we can have a platform based approach that allows them to deploy the solutions that are appropriate for whichever group of students at scale in a way that has been done before and empowers them to be more effective. And so historically, learning was really confined to what happened kind of during the school day. And what we're seeing now is there's this opportunity to let students learn all the time, whether it's before school or after school, or at night or on weekends, and allow schools to deploy solutions to students that don't limit the learning to just the school day. And we think that's really powerful. It's part of this broader concept around the unbundling of services and giving and empowering technology to administrators in a way that's really helpful. One of the things that we think is important is to do it in a way that is really durable, that lends itself to long term relationships with schools that is oriented around student success. It is not kind of short term in nature. And so what we're trying to do is bundle solutions together in a way that allows for flexibility on the district's part, but provides enough leverage that they would never want to part ways with some of these solutions because they're so helpful. So as an example, one of the solutions that we're going to be partnering with schools on this back to school is a teacher prescribed tutoring solution where at scale, all the teachers in a school district could assign tutoring to whoever needs tutoring in their classroom. And as simple as that concept that sounds described, it has never been done before. So you've never been able to give all the teachers the ability to prescribe tutoring and give tutoring to students in their classroom. And given how short staffed schools are, given how hard teachers already work, we think we can be a partner to them and allow them to get immense leverage that then benefits students.

That's actually phenomenal. I want to stay on that for a moment because I don't want that to get swept away. What you just said is essentially, if I'm understanding correctly, teachers have their roster of students, they can essentially click on a student and assign them the opportunity to get tutoring. That's a significantly easier lift than say differentiating instruction where they have to do all the diagnosis, figure out exactly what's challenging the student, find the right instructional material, match it themselves. Here they basically say, hey, I know Michael is struggling with double digit addition or complex fractions or whatever it might be, send it over to varsity tutors and they'll find the right tutor to help him get keep moving. And by the way, you know, I've got soccer practice after school. I can schedule it at the right time for me. Am I hearing that right? Because that's, that's actually a. Sounds hugely complicated, but also hugely simplifying for a teacher's life, if I'm. If I'm getting you correctly.

Yeah, exactly. And the learning no longer needs to be confined to the school day. So the, you know, the teacher could allow for it to occur outside of the traditional school day, you know, in evenings or weekends. And then all of a sudden you free up all this extra capacity for learning that wouldn't have been aided by the school in the same way in the past. And so nothing precludes the teacher from still doing all the same things they did previously to help the student. They just now have a partner that can help them help more students at scale.

That seems like a huge breakthrough from my perspective. Speak to the financial piece of this, that when federal dollars dry up in a couple of years, what will keep this sustainable so that districts don't slide back and leave, frankly, all the research about the benefits of tutoring and making this accessible to students.

Well, you know, first, you know, there's existing funds available. You know, unfortunately, schools are having trouble filling the roles and staffing, which is leaving for budget surpluses that haven't existed in the past, which then leaves room for just within the traditional normal school budget, you know, funds available to pay for tutoring and other solutions that really can help students and provide schools and teachers and administrators productivity and leverage that we think they'll value. The second part is there are substantial funds available. So the American rescue plan had $25 billion in Covid learning loss related funding. Tutoring is being held up as the preeminent solution for that learning loss. And then as of 1231, only about 2% of those funds have been spent. So over the course of the next couple of years, we expect that schools will start deploying those at scale. And you're seeing headlines, you know, every week about states passing specific initiatives oriented around here. And then of course you have additional funds like Title 1 that have been increasing over time that are focused on students that need, you know, a disproportionate amount of help and resources. And so we think there's substantial funding available in perpetuity here. But like any product, including you, think back to our earlier part of the conversation in the consumer world, you know, if you provide immense value and you do a good job and you're a trusted partner, then people will want to work with you on a continued basis, irrespective of funding. If you do a bad job and you're not a trusted partner, then you know, when the funding dries up. And I think you'll see some of the companies that haven't focused on the holistic solution not experience the same level of success. So we're trying very hard to take that long term orientation here.

So there may be a cliff for some companies, in other words, but if you really nail both the purchaser and the user, teachers, et cetera, all their jobs, that you have a good chance of sustaining it points to something else that you've said multiple times here, but I just want to go into it to make all these things happen, both from a holistic matching perspective, from an on demand perspective, for teachers to be able to use this to get their students the right tutoring solution. That's a significant investment in the technology stack itself that you all make and strikes me from the outside as a differentiator compared to some of the other players in the space. Can you just speak about that investment into the technology stack? How do you think about the advantage? You've built a lot of proprietary tools as well, and just, just give us a little bit more behind the, you know, just a little bit more of a look into those investment dollars.

Sure. So, you know, at this point, we're 15 years into, you know, investing in totality, but really, you know, since the Live learning platform launched in 2014 and we started raising outside institutional capital and really trying to hone that experience for the student and make it exceptional, spanning all these different modalities of learning, making sure the entire platform's instrumented, that it's this immersive experience that digitally enhances learning. There's a lot of technology that underpins it. There's also a lot of data and a lot of know how where we've had to kind of test and learn the hard way, but then build. We've been able to build those learnings into the software, into such a way where it becomes replicable over time, at scale, in a way that hasn't been done before. And so one of the things that we've been able to do as we've entered the school market and started partnering with schools is to leverage all the investments that we've made on the consumer side in the platform now into schools. And we have a whole host that we haven't yet deployed. But the big idea here is, is what we call kind of learning platform as a service, where we have an underlying platform where schools can roster all their students and then they can deploy from a wide variety of different products across different audience or subject areas, solutions for the students that need those, that need that help. And so in general, a lot of those benefit both consumers and institutional. And we're prioritizing a lot of investments that are kind of overlap in both areas and make sure that we can deploy and have both, both audiences, both groups in mind. So from our perspective, to the extent that we can have a platform that has the capability to have modularized specialties, products, capabilities, etc. You can kind of use them like Lego blocks to create solutions for a given segment that really resonate with them, so that when we start helping students with special needs, we're able to create a differentiated solution very efficiently that's focused on helping them. We're able to leverage all the different modules and capabilities that we've built, you know, across the platform, ranging from something as simple as like landing page design, to, you know, our adaptive machine learning based diagnostic tests, or our group classes, or our live stream capability, or tutoring or a whole host of other functionality that's helpful.

One phrase you haven't uttered is artificial intelligence or machine learning or things like that, but I know it's something that you think about as well. I confess I've been a skeptic of a lot of the AI that I've seen in education, but it strikes me it might have more use in some of the holistic matching that you all do. How does AI, machine learning, etc. Play into the service that you deliver?

Yeah, so obviously that can mean a bunch of different things. We think about machine learning from the perspective of being able to detect patterns and then allow for us to act on those patterns when a human couldn't detect them on their own. And so examples include leveraging some of those capabilities during the initial application process to identify great tutors at scale and do that in a way that was very consistent over time. So the net effect of looking for some of the patterns with applications and even what they say in recorded video interviews comes back to leveraging technology. Then kind of back test against available data sets that we built up from all the instrumentation and data capture in a way that can really help us do a better job over time. And then similarly, when students are on the platform, we use over 100 different variables to inform in the one to one setting which student they're ultimate, which expert they're ultimately paired with. And then as it relates to diagnostic testing, it's incredibly important to know what you do or don't know. And if you can do that at the concept level and do it in a way where you understand the foundational skills that lead up to that particular concept, some of which might be multiple grades behind, it, then allows us to be much more thoughtful in informing the ultimate tutoring that takes place and providing that information about students and the tutors. And that's something that on the school side is of immense value. The ability to give adaptive machine learning based diagnostic tests to thousands of students at the same time and then be able to provide back to the school or a tutor that information that allows them to be much more informed in the actual teaching.

Are you actually building those assessments yourself or those item banks that you're taking from other places and deploying them in your platform?

No, these are all proprietary.

Gotcha. Well, that's a significant advantage, I would think, in terms of being able to get information and insight into where students misconceptions are to be able to deliver the right tutor then as well.

Yeah, and you can also pair students who have similar foundational knowledge together. So even if you're in a group setting, you can still get the know benefits of one on one tutoring because people need to learn the same material as opposed to what has historically been the case is, you know, kind of a random smattering of people that need help and then you're teaching to the average. Even in a group of five as an example, there's an opportunity to use, you know, data to very efficiently inform how you group students so you get the maximum amount of benefit from personalization possible. A lot of this, you know, AI, ML, whatever, like it comes back to personalizing the experience for somebody's needs and that's how we think about leveraging that tool.

It's a lot of advantages though that you have against that stack. Chuck, I really appreciate the time today continued luck in educating the youth to all the way up to professionals trying to upskill in their careers. Really appreciate it. And for all you watching, if you like conversations like this one, give us a thumbs up on whatever platform you consumed it so we can keep following the Chucks of the world as they go forth and keep revolutionizing education. Thanks so much for joining us.

Thanks Michael.