I'm Doug Anmuth, J.P. Morgan's internet analyst. We're pleased to have with us today Nerdy's Founder, Chairman, and CEO, Chuck Cohn, and CFO Jason Pello. Nerdy is a leading online live learning platform, leveraging technology to connect learners with experts. The company generates revenue from providing personalized, curated learning experiences across a wide range of subjects and formats, including Learning Memberships, one-on-one instruction, small groups, large groups, even adaptive self-study. Chuck founded Nerdy in 2007. He was previously an investment banker and worked in private equity. Jason has been CFO since October of 2020, and previously served as VP of corporate finance at Save A Lot. Welcome, Chuck and Jason.
Thank you for having us, Doug.
Thanks, Doug.
All right. let's see. Kicking off, over the past year ago, year or so, really important business model transition that you've made toward memberships. hoping you can talk through the rationale in terms of, the transition and how you're coming out now on the other side.
We just announced some pretty exciting results in the first quarter, and they relate to this shift from what had been our historical model for selling products on our platform, which was oriented around selling packages, kind of groups of hours, groups of classes. We announced about a year ago that we're undertaking this journey to shift all these different products that were discreetly sold into this all-access pass that we call Learning Memberships, where people have, on a recurring basis in subscription format, access to tutoring and classes and asynchronous content and videos and other forms of modalities of learning.
What we had seen prior to that that kind of caused us to do that, make this big business model shift, was that when we could bring together all of these different modalities, we saw fundamentally different customer behavior. People used us for more subjects. They were able to learn across a multitude of different learning formats, and they ultimately stuck around longer. We're now to the point where Learning Memberships is now the vast majority of the business, so it's about 3/4 of our recognized revenue in the first quarter. About 90% of new customers joining the platform on the consumer side are joining in Learning Memberships. You know, we're on track to probably double lifetime value, holding customer acquisition costs constant, and we have happier customers, higher gross margins.
It's a much simpler business to operate, and, we're getting a lot of leverage as a result of it, and it was a big contributor to what allowed for us to become adjusted EBITDA and free cash flow positive in the first quarter.
Okay, great. Maybe just to dig in there a little bit more, I think you've talked about having all consumer revenue, basically on membership by the end of the year.
Mm-hmm.
Just dig in a little bit more on what we're seeing in terms of the changes on LTV and conversion rates and retention as well for membership versus kind of the old packaged model.
Yeah, sure. Because of the way it's sold and packaged, consumers are just getting a greater value proposition out of the product offering. Beyond one-to-one tutoring, as Chuck mentioned, they have access to 250 live classes a week, adaptive assessments, and a whole host of other learning modalities that's driven at the six-month point revenue LTVs that are 50% higher under the new model versus the old model. I only use the six-month mark because that's really the duration of the experience we've got.
When you look at those cohorts, which we've included in our shareholder letter, you would see that, you know, the steepness of them on a relative basis to the package model will indicate that, you know, after 1 year, you should expect, you know, a two to 3x LTV ratio compared to the prior package model, which we feel really good about. You know, higher revenue, better gross margins, higher levels of retention and engagement, and then certainly just a simplified business model, which we feel really good about from a scaling perspective.
Okay, great. How do you think about, kind of retargeting, maybe reactivating, some of the legacy package members who may have churned off or just as you had some degree of noise, you know, through the, through the transition? How do you think about that, especially as you focus on back to school, you know, kind of a few months out?
I can take this one.
Okay.
In the package world, the relationship with the customer is kind of defined. Somebody's buying you for one subject and a certain number of hours, and implicitly, when they run out, it was up to us to kind of convince them to continue. In this new all-access model, one, you know, it's a subscription, it's recurring in nature by default. Maybe more importantly, like, the consumer psychology has totally changed 'cause the product is being framed from the perspective of supporting somebody across all academic subjects, all modalities, oriented to multi-multiple academic school years, and then multiple students per family, where you're encouraged to add multiple kids, you know, within the same family.
People come in kind of expecting to use it for a long period of time. That results in, like, fundamental separation of cohorts on, you know, over the first several months, where you just see that people in the package model would fall out of habit, and then people in the new model kind of stay in habit, which is then reflected in much higher lifetime values. We shared in our earnings call that it was about 85% of active Learning Membership customers were new to the platform. This is really resonating with new customers. We've been able to shift the proportion of people that think we're a great value by almost 100%.
From, like, 50% of people that thought we were a great value in the package world to now, like, 90%, which is something that we think is really compelling and indicative of Just, a product that is resonating, winning, causing us to have strong, like, new customer additions.
As you think about back to school, there's all these people that had been in the package model and, you know, given that kind of different consumer psychology fell out of habit, and every year we have a big opportunity to re-sign them up. You know, we would expect that we're able to then target those customers, as we've often done, oriented around the school year and add them to Learning Memberships.
Okay, great. You've recently mentioned that May is tracking in line with expectations. Just to clarify, is this Membership gross adds, and can you unpack, I guess, if these are still mostly new users?
Yeah. You know, we had a great first quarter. The engagement and the new customer addition trends that we saw on a year-over-year basis continued through April and May. We feel really good about the retention levels that we're seeing and the engagement. We think that that'll continue, you know, as we move throughout the summer.
We, you know, like I guess every business look at new customer additions on a year-over-year basis trended over time to account for seasonality and the fact that those were, like, getting better, you know, is something that, you know, we feel really good about, and it's the result of, like, specific pieces of work that we've done that cause for the product to resonate more and ultimately result in, you know, higher conversion.
Got it. Okay. If you think about Membership, about a third of active members are on monthly plans. You see reduced friction during the sign-up process and kind of requires less commitment. How has this informed your pricing strategy for Membership as you look further out?
When we initially went to market, you know, keep in mind this is a product we launched a year ago, and we kept it pretty simple. There are, you know, three, six, 12-month contracts. One of the things that we tested our way into was offering month-to-month, recognizing that some people had, like, a shorter duration need or there was perceived complexity associated with a long-term contract. What's been kind of exciting, is that as you, like, looked at the combined answer both month-to-month and then, like, contract customers, you end up with people kind of self-selecting into different buckets. The total answer from a retention perspective looks the same, but you're able to then sign up way more customers on the front end because there's no kind of perceived complexity associated with contracts.
We've seen, like, conversion go up at the top of the funnel. We feel good about the retention trends there. You're able to charge, 10% to 15% more for month-to-month customers than is the case for, you know, longer term contract customers.
Okay, great. All right, let's shift gears a little bit, talk about the institutional business. You launched VTS in almost two years ago, August of 2021. Maybe you can just help us understand, I mean, first just describe that business a little bit for people who may not be as familiar. Also the shift that you're seeing in terms of selling strategy toward top-down and district-wide solutions, as you're going toward large districts.
Sure. Our historical business was entirely consumer. We built up all these different capabilities that we believed made us uniquely qualified to deliver live learning at scale in a way that could be sold into K-12 school districts. It allowed for district administrators to achieve goals, help students that would otherwise have been difficult for them to do in the past. Nobody else has the same level of experience delivering live learning at scale through a software-based platform like this. What we did was we created, you know, at first, a product called High- Dosage Tutoring that we used, that was kind of oriented toward selling into students that needed significant remediation.
We've kind of evolved that product portfolio over the course of the past year or so by adding in two new SaaS products that are district-wide in nature that, you know, instead of being sold at the kind of school level, are oriented instead toward, you know, a more top-down and strategic conversation and partnership with superintendents and other district leaders. That coincided with Anthony Salcito, who was the head of Microsoft's worldwide education business globally, joining us and helping us kind of take some of these amazing product capabilities we had and then, you know, simplify them for school district administrators in a way that lent itself to really big partnerships and implementation. We had our Teacher Assigned product go live in January.
That was a roughly $5 million software subscription that entitles teachers throughout a large school district to be able to prescribe tutoring to any student that needs it, when they need it. As simple as that is to describe, it's never been done before. That's kind of caused us to totally shift the orientation to these, like, broad, larger partnerships involving bundled services, as opposed to kind of the initial focus on smaller schools and smaller contract sites.
Maybe just to put a little bit of sizing to that business. We started Varsity Tutors for Schools about 18 months ago. Last year, in 2022, it delivered $19 million of revenue, and we've guided that it'll be about 15% of total revenue in 2023, which is $28 million or $29 million, and it would represent 50% growth year-over-year. We're seeing nice product market fit there, especially with the new Teacher Assigned product, and we feel like that business is scaling well.
What has surprised you most over these last, you know, 18, almost 24 months, just on the VTS product journey? Maybe you could also just talk about differentiation versus peers?
Sure. One is that, you know, the same kind of journey we underwent on the consumer side with the benefit of bundled services ultimately causing way more, you know, perceived value and utilization across a variety of different modalities and ultimately leading to stickier relationships. That same thing has happened on the school side as well. Then one of the things that we've seen in the institutional market is that there are a lot of chat-based providers that initially had significant success.
You know, our superpower is delivering live, relationship-based, recurring, sessions, which has always been proven to be highly effective, sticky, and, you know, also very difficult to do, and something that we've been able to do well through the application of software, you know, that from the kind of consumer's end looks easy, but from the back end is logistically complicated. We're now seeing the market shift back to appreciating live, and that's something that we think accrues to our advantage over time.
The only thing I'd add is, you know, we're unique in that we've got three different products to sell into schools, so High- Dosage Tutoring, Teacher Assigned and On Demand, and we can bundle those together to serve the entire student population and the diversity of needs of each of those students. I think that that's a unique opportunity set in the market today.
The one other thing I'd add is, you know, we initially rolled out some of our AI products on the consumer side first, given that there's kind of lower complexity and considerations to take into account. We've been having school district partners actively ask us for access to those, which has been really exciting.
Okay, great. The deadline to commit ARP funding expires in September of 2024. How would you expect school districts just to think about locking up contracts that run into 2025, 2026, for example?
Sure. For those that aren't aware, the American Rescue Plan ascribed $24 billion to help remediate COVID learning loss. About a quarter of those funds have been spent to date. The residual 75% or so needs to be spent by September of 2024. Beyond just that ARP money, there's also Title I funding, which is evergreen, and in the most recent spending bill was about $19 billion. The other source of funding that we see school districts utilize is just their own operating budgets. Because the products we sell are utilized during the day in the classroom, they're eligible for those funds. Given the teacher shortages that all school districts are experiencing, the majority of school districts are running budget surpluses.
As we think about the guide, it wouldn't infer a pull forward of those funds, but we're starting to certainly have conversations with school districts around multiple year deals because they need to deploy those funds before September of 2024. I guess I would consider that upside.
Okay. I mean, is there any risk, just as you think about large districts and how they could face funding pressures without government help?
We feel pretty good about the opportunity here. I mean, you know, this product, like, should exist, right? Every teacher in every school district should be able to assign tutoring to any student that needs it. It gives the teachers immense leverage. We're hearing about it improving teacher retention, it being used to attract teachers to school districts, you know, that have it. That's incredibly powerful, where in many states, like there are entire states that have 10% vacancy rates. 10% of the open spots for teachers can't be filled. This is a way where schools can provide teachers with leverage, put them at the center of the relationship and give them control.
In doing so, not only are they helping the students and the teachers, but all these additional funding sources, like normal operating funds and Title I funds, also become available. Given that these products solve, like, the top two problems that school districts are facing, and given the novelty of what we've built here and how powerful we think it is, like, it lends itself to a big opportunity kind of regardless of funding.
Okay, great. Let's shift gears a little bit. Kind of putting together consumer and VTS, just thinking about the strategy going forward, what is the playbook to really scaling both businesses?
One of the things that's really nice about having these all access models is that you can then bundle in additional solutions.
Mm-hmm.
That add incremental value, solve specific customer pain points, whether it's on the consumer side or on the institutional side, and then benefit from both higher conversion, right? Somebody being more likely to select your product or ultimately higher retention lifetime value. As we've added additional products to both Learning Memberships, where we started off first as effectively just a tutoring subscription, then added in 250 live classes a week and computer adaptive assessments and an AI Tutor and lesson plan generators driven by AI and a number of other products as well, you've just seen, like, the perceived value go up consistently, and we've seen conversion and funnel go up. We've seen retention go up, and the same sort of dynamics proving true on institutions. There's a wide variety of other customer needs that we can solve.
The only thing I'd add is, you know, from a addressable market perspective, certainly we've got 33,000 active members on the consumer side to end the first quarter. There's on average, at any given time, 50 million students in the United States alone. There's significant opportunity to continue to extend this to different audiences, different subjects, over the coming years. Certainly, I think on the school side, the same situation exists where schools are much more willing post-COVID to utilize online providers from outside the school's walls to support students and teachers inside the school's walls, within the school day.
Okay. you added Codeverse last year.
Mm-hmm.
To broaden out content offerings, enter new verticals. How do you balance organic growth versus M&A over time? What kind of assets could be interesting, if you were to think more about acquisitions?
Sure. Codeverse is a kids coding platform that gamifies, you know, learning how to code. This was just an example of two capabilities we didn't have, coding or gamification. It was an opportunistic acquisition. The company had raised substantial amounts in the venture capital world and had run out of money, and we purchased them for about a quarter million dollars and acqui-hired their whole team. Great deal for us. Two quarters after purchasing them, we've rolled it into Learning Memberships to drive higher engagement, higher retention. It's not sold on a standalone basis, but it gave us those additional capabilities to enhance the Learning Membership. Feel good about that acquisition. I would say on a go-forward basis, our focus would be on the organic growth that we talked about and the ability to scale Learning Memberships.
Of our Varsity Tutors for Schools side of the house, we believe we've got, you know, the right team from an engineering and product perspective to continue to deliver against those initiatives. That would be where the focus is over the, I'll call it, coming several years.
Great. You've shifted focus, to providing more monetization opportunities to the highest performing experts. That should lower expert acquisition costs and increase retention. How do you get comfortable with concentrating the expert network more and actually reducing expert headcount?
One of the things that we've done is, as we've leaned into machine learning, matching algorithms over the last, you know, four or five years, kind of bringing that, you know, to a like almost 100% of all the volume going through the platform, the best people on average tend to be good at a number of things. The leaning into the ML algos ultimately causes us to allocate more of the new students joining the platform to people who are likely to generate the highest lifetime value. That in turn causes those top experts to stick around longer, have a better experience, have more recurring earning potential.
As you think about that dynamic, it's still relatively early, where people are only working, call it four, five hours a week on average, and we think we can continue to get more and more efficient. One of the areas where we've been getting leverage is on using machine learning to predict the probability that somebody comes onto the platform but doesn't kind of stick it out or take students. Simply by getting smarter about probabilistic models, you're able to, like, not bring on people where you would've incurred cost and they wouldn't have taken any learners on the platform. We're gonna be able to continue to get, like, meaningful leverage there over the course of the year without, you know, any sort of significant increase in concentration or anything like that.
I would just add from a historical perspective, certainly because it's all online and most of tutoring happens after school and after work, we believe we have access to, like, the entire educated workforce in the United States to serve as tutors. We've never had any challenges scaling to meet demand. We think that this is, as Chuck mentioned, an opportunity to actually concentrate the tutoring into the best tutors that we have and drive higher engagement in LTV.
Again, we're only at, you know, a couple hours a week, and there's an opportunity to shift that up a little bit.
Okay. even as you think about, the membership, acceleration...
Mm-hmm.
Strong growth in membership and healthy BTS contract demand, still feel good about expert supply, basically?
Yep.
Absolutely.
Okay. All right, let's shift gears a little bit, talk about AI. Certainly hot topic across online education. How has generative AI, large language models impacted customer acquisition or retention? Are you seeing any changes to what you talked about at earnings in early May?
All of our new customer addition trends on a year-over-year basis look, you know, strong, same with all the engagement trends look strong. We feel great about the momentum in the business. As it relates to generative AI, you know, we've been applying machine learning since probably 2016, 2017. Started off with using it to inform the kind of expert learner match and getting smarter and smarter at that level of personalization. Using it for computer adaptive testing. Starting to use it to just remove costs from the business, being smarter about propensity modeling and when we engage.
We've had a lot of infrastructure in place to instrument all the data, capture it, and then as a vertically integrated model, there's all these different opportunities to enhance the customer journey throughout that ultimately accrue to our benefit because the customer gets a better experience, and so you have higher lifetime value, holding costs constant, then that drives leverage. That's been a big form of growth over the last few years. With generative AI specifically, we have an AI Tutor, which is kind of used as a homework help Q&A. Like, so you could think of it as quick access to questions and answers, and it's been, you know, a small mid-single digit form of engagement on the platform that's been incremental to all the other different modalities that you could leverage.
We also have an AI Lesson Plan Generator where we pre-generate content that is specific to an individual student across 3,000+ subjects and kind of unlimited, you know, other combinations and complexities that can be taken into account, like the complexity level, the student's interest, you know, other things, and the tutor can edit and regenerate that in real time. That's something historically where we could have a highly relevant lesson plan where we saw an extra, call it 10%-20% lifetime value, but prior to generative AI, we could never figure how to scale that across so many different subjects and combinations. There's other areas where we're using some of the new technologies to automate costs out of the business where, you know, we think there's an opportunity to get more and more efficient.
We've already had some significant wins year to date and, you know, are continuing to invest in, you know, streamlined processes that ultimately will allow for us to have a more scalable and efficient business model. Okay. How do you think about investment needs for AI generally speaking?
Sure. The investments we've made to date, you know, have produced, you know, both top line revenue growth, better engagement with customers and therefore LTV extension, but it's also allowed us to drive increased levels of process automation through the company. As we're adding AI resources, you know, almost in real time, they're also automating tasks within the company that allow us to scale beyond certainly where we are today at a reduced cost structure. The IRR from our perspective is substantial, and the payback periods are quite short. The level investments we're making in AI are modest, but they're paying back quite quickly.
Because we've been, you know, applying this technology practically for many years and had, like, good infrastructure in place for, you know, building services that teams throughout the company could leverage, as well as storing the data and then, you know, back testing models, we're able to move pretty quickly on getting leverage out of them and then, you know, getting automation wins. There's kind of a healthy portfolio of things that would, like, make the product more compelling and drive top of funnel, things that would make it more compelling and drive retention, and then, you know, process automation and cost elimination. Every employee in the company has access to inline tools related to GPT-4 and, you know, use that throughout their day, and they're highly encouraged to do so.
About 30% of our code is now being written by AI-related programs on our engineering team. We would expect to just continue to lean in there. While that isn't proprietary, we do wanna be offensive as it relates to leveraging all these technologies to kind of run ahead.
Okay, great. Let's talk more about financials, margins and OpEx. You guided to substantial improvements.
Mm-hmm.
In EBITDA margins for this year. Maybe you could just talk about how you get there. What drives the leverage?
Sure. Year-over-year, we've guided toward a 1,900 basis point improvement in EBITDA margins. How do you get there? We already started to see it in the first quarter. We had about nearly 1,700 basis points of improvement. On the full year, you should expect about 300 basis point improvement in gross margins, about 800 basis points of leverage within the sales and marketing line item, another 800 basis points of leverage in G&A as we continue to scale without any requisite fixed cost increases there. You know, it's early. We're starting to prove it out. We always thought that Learning Memberships would allow us to simplify the business and drive increased levels of growth and profitability.
You started to see the fruits of that pull through to the P&L in the first quarter. We expect that to continue throughout the rest of this year.
Okay, great. consumer margins-.
Mm-hmm.
In particular, have expanded each of the last two years. I guess, how do we think about further expansion here as the revenue mix, increasingly, becomes driven by membership?
Sure. In 2021, we had 66% gross margins. That increased to 69% last year. This year, we've guided to 71%-72%. Essentially what you're seeing there is as we mix shift towards a higher proportion of revenues coming from Learning Memberships, that accretion taking place. The way we're able to capture that is, you know, Learning Memberships, because they include all the modalities and learning tools, they have a higher perceived value from a customer's perspective. Certainly, some of the tools, you know, have a near zero marginal cost. An adaptive self-assessment, that capability's already been built. It's run on an automated basis and has zero cost.
You know, running 250 live classes in any given week isn't a substantial amount of cost for us, but the perceived value, because of the breadth of the offering, whether it's academic and enrichment subjects, is got high perceived value from customers. All of that in combination allows us to drive gross margin accretion.
Okay. Let's see. The membership shift and then also the revised VTS sales process that we talked about.
Mm-hmm.
That's gonna drive sales and marketing leverage in 2023 and beyond. I guess, is that the right way to think about sales and marketing spend going forward? You know, how do you think about it kind of as you look out beyond 2023?
Sure. I think, you know, a couple things there. About a year ago, we decided to walk away from more transactional, we called them à la carte or class customers, because we combined those offerings into the Learning Membership. We gained some efficiencies there because we weren't targeting those lower value customers. That represented about maybe 30% of the customer base, but only 7% of the total revenues. You're seeing some efficiencies there. Then, you know, certainly in the first quarter, we saw 1,700 basis points of sales and marketing leverage. That's a combination of the increased targeting that I mentioned on the one hand, but then the LTV extension is really the other side of it.
Each incremental month that a customer stays on the platform, you're seeing 75%-80% of those contribution $ just flow through to the bottom line because you've already incurred the sales and marketing costs. We think we can continue to drive leverage there. Certainly, you know, you can envision a situation where your LTV to CACs get into the, you know, three, four, five, six range. If that starts to happen, we'll redeploy $ because we think that that trade-off there from a growth perspective is gonna be absolutely worth it. You know, certainly our Chief Marketing Officer, Adam Weber, is looking forward to that happening.
Yeah. I mean, maybe the more simple way I would've said is, you know, we're adding new customers on a year-over-year basis, so there's growth. You know, we're trending towards 2x LTV holding CACs constant. As a result, you're just, like, layer caking high lifetime value recurring customers on top of one another, which just naturally, without any theatrics, drives pretty significant, you know, sequential quarterly growth on a year-over-year basis each quarter this year.
Got it. Okay. Maybe you can talk a little bit about the competitive environment, perhaps across, you know, both businesses and, I guess just give us a sense of how you think that plays out in future years.
On the consumer side, there's 5,000 mom-and-pop tutoring companies, hundreds of professional testing companies, lots of enrichment centers of various sorts.
Mm-hmm.
Language learning companies, you know, estimated to be, call it one million to two million independent tutors in the United States. It's a, you know, big fragmented market, you know, $15 billion-$30 billion domestic TAM, depending on which statistic you look at. The way the consumer thinks about it is on a subject-by-subject basis. And we're the only company that's doing live online learning at scale and bringing kind of all these different subjects and all these different learning formats to bear that kind of surround, you know, the live superpower we have. It's very fragmented. I can't say that, you know, anything in the last one or two years is all that interesting or different than, you know, our journey.
We're continuing to enhance our product at such a rate that we're pretty internally focused on meeting, you know, additional customer needs that exist as opposed to externally focused. Within schools, I think the school administrators and just the market in general is becoming more discerning. Where as we've gone through like a year or two of different participants being in the market and some of them not living up to expectations, we feel like that ultimately accrues to our benefit, given that we could deliver at scale in a way that, you know, others can't. There's participants, but some have come and gone, and we think ultimately, you know, the fact that doing live at scale with software in a way that makes it look easy for the customer is really hard. We think that accrues to our advantage.
Okay, great. Last question, just wrapping up, what might investors be missing about Nerdy?
I would say that we just finished, you know, a, what I expect to be a grand slam business model evolution, where we significantly increased lifetime values, enhanced gross margins, simplified our business, removed, you know, a dramatic amount of cost that was associated with the old model that this new model doesn't have, and that we're in a position to, you know, accelerate sequentially throughout the year. Maybe separately, that we became, you know, adjusted EBITDA and free cash flow positive in the first quarter, three quarters earlier than previously expected, and that, you know, we're in a position now where we can continue to, like, enhance both products for consumers and institutional way that we think, can allow us to kind of run ahead.
Okay. All right, great. We're gonna leave it there. Thank you, Chuck. Thank you, Jason.
Thank you for hosting.
Thanks, guys.