NRDY - NYSE

Good afternoon. Thank you for attending the Nerdy First Quarter 2023 Earnings Call. My name is Elisa, and I will be your moderator for today's call. All lines will be muted during the presentation portion of the call, with an opportunity for questions and answers at the end. [Operator Instructions] I would now like to pass the conference over to your host, Charles Cohn with Nerdy. Mr. Cohen, you may proceed.

Good afternoon, and thank you for joining us for Nerdy's First Quarter 2023 Earnings Call. With me are Charles Cohn, Founder, Chairman, and Chief Executive Officer of nerdy and Jason Pello, Chief Financial Officer. Before I turn the call over to Chuck, I'll remind everyone that this discussion will contain forward-looking statements, including, but not limited to, expectations with respect to Nery's future financial and operating results, strategy, opportunities, plans, and outlook. These forward-looking statements involve significant risks and uncertainties that could cause actual results to differ materially from expected results. Any forward-looking statements are made as of today's date, and Nerdy does not undertake or accept any obligation to publicly release any updates or revisions to any forward-looking statements to reflect any change in expectations or any change in events, conditions, or circumstances on which any such statement is based. Please refer to the disclaimers in today's shareholder letter announcing Nerdy's first-quarter results and the company's filings with the SEC for a discussion of the risks. Not all of the financial measures that we will discuss today are prepared in accordance with GAAP. Please refer to today's shareholder letter for reconciliations of these non-GAAP measures. With that, let me turn the call over to Chuck. Chuck?

Thanks, TJ, and thank you to everyone who has joined us today. One year ago, we unveiled an ambitious plan to evolve our products and revenue model toward long-term, recurring, always-on relationships with our customers. We created a new subscription and recurring revenue products, including learning memberships for consumers and teachers signed and on-demand products for institutional customers that were built specifically to address the ongoing support we believe both types of customers need it and desired. We shared that we believe these new models would provide a superior platform for innovation by allowing us to bring together multiple different product capabilities we have developed into a comprehensive all-access offering that enable learners to receive the help they need across multiple learning formats, thousands of subjects at multiple academic calendar years. In addition to allowing us to provide a better and more personalized experience to learners, the new operating model would be far more efficient to operate, allowing us to drive operating leverage, simplify our sales model and shift additional resources toward net new innovation, including the application of AI for HI for artificial intelligence for human interaction. To get to this evolved state, we shared that this new model would require trading off revenue recognition in the short term because our package model had a more front-loaded revenue recognition than the learning membership model where subscription revenue is recognized linearly over time. We expected that by the start of the second quarter of 2023, the cumulative build of recurring revenue from learning membership customers would cause us to return to growth in our consumer business as well as for the total company, but now with the product suite and revenue model, we believe would position us for higher levels of growth, profitability, and predictability in the years to come. We stated that we expected our new business model to deliver substantial operating efficiencies and that we anticipated achieving adjusted EBITDA profitability by the end of 2023. I am pleased to share that in the first quarter, we exited the J-curve business model transition to subscription and returned to year-over-year growth, delivering $49.2 million of revenue, which was above our guidance range of $45 million to $47 million. Wording membership subscriptions accounted for 60% of total company recognized revenue, up from nearly 0% in the first quarter last year, demonstrating strong product market fit. On the institutional side of the business, we delivered record revenue of $8.5 million, an increase of 32% year-over-year, representing 17% of total recognized revenue in the first quarter. In combination, our consumer subscription and institutionally contracted revenue accounted for 77% of total company revenue recognized in the quarter. a dramatic increase over the past year, which we believe speaks to the power of our platform-oriented approach to growth and how we can efficiently go to market with all-access solutions that could provide more value to customers. The rapid business model evolution learning memberships and recurring revenue and our continued application of AI to the customer experience and operational processes drove continued growth and improvement in customer lifetime values relative to our package model as well as operating efficiency improvements. Together, they were key contributors to our strong operating results and improved profitability. We are pleased to share that we achieved adjusted EBITDA profitability in the first quarter, 9 months earlier that our target of the fourth quarter of 2023, delivering nearly 1,700 basis points of improvement year-over-year and $1.4 million of adjusted EBITDA. The improvement in adjusted EBITDA represents an annualized improvement of approximately $32 million. We also realized $6.8 million of positive operating cash flow and $5.8 million of free cash flow in the quarter. I'm pleased to share that we made substantial progress on accelerating the use of generative AI around our business, including launching 2 new customer-facing products as well as accelerating the use of AI and machine learning to drive substantial operating efficiencies and internal productivity improvements. As we shared with you in our prospectus 2 years ago, when we made our intention to become a publicly listed company known, we've long believed that AI can fundamentally transform how people learn. We've been applying AI to our business, products, and operational processes for over 6 years, and AI has been foundational to our ability to improve quality, enhance personalization and decrease the cost of our offerings. AI powers our ability to identify the highest quality experts assess Werner's foundational knowledge to help ensure the right expert learner match and drive operational efficiency among many other use cases. In our prospectus, we also described the proprietary technology infrastructure we're building as AI for HI or artificial intelligence for human interaction and outline the core foundational capabilities we apply to live learning to enhance the interaction in ways that were not previously possible. Through the application of AI, we aim to provide experts and learners with superpowers that transform live online learning. We credit our orientation around AI for HI for allowing us to reach the milestone of having recently delivered our 10 millionth hour of live one-on-one tutoring on our live learning platform and more broadly for allowing us to make good on delivering high-quality relationship-based live online learning available at scale. We've used this data in the insights we garner from having instrumented every part of the learning journey to drive real-world practical value in our business. We credit AI for enhancing our customer lifetime values, helping us identify what each student does or doesn't know about a particular subject, when to reach out to a customer to drive engagement and retention, and how to make operational processes far more efficient. We believe we stand to benefit tremendously from the latest advancements in generative AI and further drive revenue growth and cost reduction through its application. What has been maybe most exciting is seeing what our internal teams have been able to accomplish over the last 90 days and the overall piece of innovation internally. As internal access to generative AI tools has expanded, we are seeing it enhance our team's quality of work, the speed it takes to complete set work, and open up new possibilities for products and process improvement in a way that previously would not have been feasible or that would have been cost prohibitive. To illustrate the speed of innovation, I'll cover some of the progress we've made leveraging AI in just the last quarter. We've continued to improve our expert render matching algorithms. As a reminder, we first started applying machine learning matching algorithms 6-plus years ago to begin to detect patterns that no human possibly could to better inform the match between a learner and an expert. And with thousands of experts available for a given learner, taking a technology-first approach to programmatically identifying patterns that were predictive of better learning experiences and outcomes has proven highly effective. Today, we simultaneously test competing machine learning algorithms until one is named the statistical Victor and flipped to 100% of the volume for a given segment, and then the process repeats. And as we capture and better leverage data to inform the learner expert match, the quality of the match will continue to improve for both learners and experts in turn leading to a far better learning experience, ultimately driving better customer satisfaction, better learner outcomes, and ultimately, higher customer lifetime value. In the first quarter, we also further expanded learning formats and content. Our growth flywheel, which we first shared publicly in January of 2021, reflects additional learning formats beyond 1:1 as a key contributor to what attracts new learners to the platform. These additional learning formats, combined with relative content to the subject being learned create personalized learning experience that drive engagement and retention of learners on the platform. During the first quarter, we leveraged generative AI to launch 2 previously announced products, AI-enabled chat tutoring and AI lesson plant generator, both of which we believe will be further accelerated to our growth flywheel. Additionally, we have steadily enhanced the availability of asynchronous content on the platform in areas like self-study and computer adaptive diagnostic testing, which has driven higher levels of engagement and customer satisfaction. However, due to the unlimited possibilities of subject and age complexity and the nuances between school curriculum, it just wasn't feasible nor possible to have rich levels of content in every single subject that someone could conceivably want to learn on the platform. Base to advances in generative AI that has now changed. We stand to be a huge beneficiary of being able to infuse high-quality, hyper-personalized content that has historically been expensive and time-consuming to develop into every learning experience. We believe this hyper-personalization will allow us to further meet the needs of our learners on a recurring basis over time. We're also seeing significant improvements to productivity and operating efficiency through the application of generative AI. Today, approximately 30% of our software code is being written by AI. All of our employees have access to in-line generative AI capabilities like GPD 4 and are encouraged and expected to use it in their work. And we're now using generative AI to more efficiently solve customer support interactions and automate operational processes, including now broadly leveraging AI-powered support parts across learner-facing, expert-facing, and even internally-facing interactions, and we expect to see further wins on driving both conversion and retention as well as improvements in operational efficiency as a result of continued investments in generative AI across the business. Let's move on to learning memberships, which are scaling ahead of expectations. We continue to see substantial evidence that validates our belief that the learning membership model leads to more attractive unit-level economics, longer duration and higher lifetime value customer relationships, higher gross margin, and a more scalable and efficient operating model. We are also able to provide an improved and more comprehensive learning experience for learners and more consistent earning potential for experts. We remain convinced that this all-access always-on business model serves as a better platform for innovation and growth. It allows us to better capture and then apply our proprietary data across product interactions over time to drive deeper personalization for learners across many different learning formats and subjects. And we're able to easily incorporate new products that add more value to the learning membership experience over time. Orders have a better experience on the platform, engage more frequently and for longer periods of time, and were ultimately rewarded in the form of higher customer lifetime values. Learning Membership revenue continued to grow at a rapid pace during the first quarter and reached an annualized run rate of approximately $143 million as of March 31, an increase from $87 million as of year-end and nearly $0 in the first quarter of 2022. Active members grew to nearly 33,000 as of March 31, up from approximately 20,000 as of year-end. During the first quarter, we expanded learning memberships to new customer audiences by fully transitioning all purchases by existing package customers in the learning memberships as well as moving all of our new test prep audience customers into learning memberships. As we look ahead, we plan to transition the professional audience to learning memberships by the end of the year, which would represent a transition to 100% of new customers to our consumer business to always on recurring revenue products. This past quarter, we introduced month-to-month learning memberships, which are driving higher levels of conversion by alleviating friction in the member experience while increasing the average monthly subscription fee and accelerating the marketing payback period. We also enhanced the value provided in learning memberships by providing unlimited access to 2 new products. Our AI-enabled chat tutoring enables learners to receive help from an AI tutor and also involves a live human tutor with a click of a button. After piloting this capability in early Q1 and receiving positive feedback and engagement data, we've recently expanded access to all owners on the platform. The primary use case so far involves quick Q&A and homework health in between live recurring one-on-one tutoring sessions. It's a good example of how generative AI has enhanced our ability to build a product, in this case, one encompassing Q&A and homework help that historically would have required substantial investments in content but now can be done for effectively no cost and served up to the customer in line at the right moment in the learning journey to keep them learning efficiently and effectively. The second product we added to learning memberships was our AI lesson plant generator. We went from idea to minimum viable product to a fully built and value-added capability deployed across our entire platform and available to all experts as of late April. With this product, we use generative AI to pre-generate lesson plans, including practice problems and other curriculum content in advance of tutoring sessions. The lesson plan generator is embedded in the user interface as a dynamic and editable pain that is ever-present during live tutoring sessions. We consider the ability to create hyper-relevant hyper-personalized content spanning any subject, any age level that is personalized for the unique needs of the specific learner to be an example of the sort of superpower we have made available to experts and learners that previously would have either been impossible or cost prohibitive. With the addition of these new products, we continue to grow the percentage of learning membership customers engaging in a nontutoring format during the first quarter to over 27%, the highest of any quarter yet. The multi-format engagement has historically been highly correlated to lifetime value extension. Looking ahead, we're working to make it easier for learners to more fully engage with our learning membership by improving discovery in an all-new member portal. This will include personalized AI-generated learning recommendations that predict and suggest the next product interaction across learning formats and subjects that is most likely to drive engagement and customer value. As we head into the slower summer months, we've created compelling content for learning members to keep learning over the summer through increased engagement with academic, college prep, and enrichment subjects. Turning our attention to our institutional business of Varsity Tutors for schools, enhancements to our product suite of high dosage tutoring, teacher aside and on-demand, coupled with prior investments in Versus for Schools sales and go-to-market resulted in record institutional revenue of $8.5 million in the first quarter, an increase of 32% year-over-year and representing 17% of total revenue in the first quarter. Varsitutors for schools executed a record 97 contracts totaling $6.3 million of bookings during the first quarter. Marcy Tutors for schools engagement trends, including our new teacher-assigned products, significantly exceeded our expectations and provide us with confidence that the solutions we have built have a strong product market fit and are well suited for meeting the needs of school district partners, teachers, and students and helping students learn in an unprecedented scale. Teacher Assign continues to deliver against our vision for delivering personalized live learning at a district-wide scale while providing unparalleled support in agency for educators. These high levels of engagement are occurring across a wide variety of grade levels of subjects, and teacher feedback has been enthusiastic that they love it. In particular, teachers see teachers signed as co-teacher in the classroom, empowering them to help more students. Each features unique insights of individual students, including the student's understanding of the classroom curriculum are incorporated in the tutoring sessions. These strong results and continued momentum to start the year give us increased confidence that [indiscernible] for schools is well positioned to provide solutions that administrators, teachers, and students are seeking to support their evolving needs. In closing, live human instruction that inspires and motivates when coupled with AI is enhancing the state of learning. With recent advances in generative AI, the ability to deliver personalized live instruction at scale for all students is within reach. We're proud of our progress to date growing our learning membership count to 33,000 active members. However, with more than 50 million students in the United States alone were just getting started. We look forward to remaining at the forefront of product innovation and enhancing our ability to meet the needs of both consumer and institutional learners. With that, I'll hand the call over to Jason to discuss the financials in more detail. Jason?

Thanks, Chuck, and good afternoon, everyone. I'm pleased to be speaking with you today about another strong quarter for Nerdy. We previously shared that we expected our evolution towards toning memberships would lead to longer duration and higher lifetime value customer relationships, enhanced gross margin, provide for more attractive unit-level economics and drive higher levels of growth and profitability. I'm pleased to report that during the first quarter, our team's hard work on the evolution to recurring revenue offerings pulled through, delivering a return to growth and positive adjusted EBITDA of $1.4 million, a nearly 1,700 basis point improvement year-over-year and more than 9 months ahead of our stated target. Looking ahead, we expect to yield additional efficiency improvements through scaling learning memberships, driving additional automation and self-service features, and the continued application of AI throughout our business. Turning to Q1 results. In the first quarter, we delivered revenue of $49.2 million, results that were above our guidance range of $45 million to $47 million. These positive results reflect the continued growth in active memberships, which totaled nearly 33,000 as of March 31, up from 20,000 at year-end. Learning memberships revenue grew to $29.7 million during the quarter and represented 60% of consumer revenues in the quarter, up from nearly 0% in the first quarter last year, demonstrating a strong product market fit. Our institutional business delivered record revenue of $8.5 million, representing 17% of total revenue during the first quarter, and delivered bookings of $6.3 million. On a combined basis, learning memberships and institutional revenues delivered 77% of total revenue, a substantial change from just 2 years ago when the business was a 100% package business and 0% institutional revenue business. Moving down the P&L. Gross profit of $33.9 million for the first quarter represented an increase of 3% compared to the same period last year. Gross margins of 68.9% for the first quarter were approximately 90 basis points lower than 69.8% in the same period last year. The increase in gross profit was driven by gross margin expansion across our consumer audience, which was offset by higher-than-anticipated engagement with our new products in our institutional business. As we evolve towards a greater mix of learning membership revenue, we expect consumer gross margin to expand throughout 2023. Sales and marketing expenses on a GAAP basis were $15.6 million in the first quarter, a decrease of $7.4 million compared to the same period in 2022. Non-GAAP sales and marketing expenses, excluding noncash stock-based compensation, were $14.7 million or 30% of revenue in the first quarter. This compares to 47% of revenue in the same period of last year and approximately 1,700 basis point improvement year-over-year. Sales and marketing spend and efficiency improvements were driven by the transition to learning memberships, including the continued expansion of lifetime value, our focus on optimizing the level of marketing spend, and a more efficient operating model in our consumer business. We also delivered substantial Bars tutors for Schools revenue growth, yielding efficiencies from prior investments in the institutional sales and go-to-market organization. As learning memberships become a greater percentage of total revenue, and the institutional business continues to scale, we expect to yield durable sales and marketing improvements. G&A on a GAAP basis was $29.7 million in the first quarter, a decrease of $800,000 compared to the same period in 2022. Non-GAAP G&A expenses, excluding noncash stock-based compensation, were $19.5 million or 40% of revenue in the first quarter. This compares to 41% of revenue in the same period of last year, an approximately 100 basis point improvement year-over-year. Combined with our ongoing efforts in automation, self-service capabilities, and the application of AI, we've been able to generate operating efficiencies and remove significant costs from the business. As noted, we reported adjusted EBITDA of $1.4 million, a nearly 1,700 basis point year-over-year improvement in the first quarter, beating the guidance range we provided of an adjusted EBITDA loss of $3 million to break even. Cash provided by operating activities was positive $6.8 million in the first quarter of 2023 compared to cash used in operating activities of $0.9 million in the prior year period, resulting in cash and cash equivalent balances increasing by $5.8 million during the quarter ended March 31. With no debt and $96.5 million of cash on our balance sheet, we believe Nerdy has ample liquidity to fund the business and pursue growth initiatives. Turning to our business outlook. Today, we're providing second-quarter and updated full-year 2023 guidance. We saw positive new customer addition and engagement trends in the first quarter that have continued into April and May and compared favorably to our normal seasonality and internal expectations. For the second quarter and full year, we expect revenue growth will be driven by the continued evolution towards recurring revenue streams, the corresponding build in the number of learning membership subscribers, and higher institutional revenues. Our positive momentum provides us with increased visibility into confidence in our expectation that we will deliver sequential year-over-year revenue growth each quarter as we move throughout 2023. For the second quarter of 2023, we expect revenue in the range of $45 million to $47 million. For the full year 2023, we are raising our revenue targets to $193 million to $200 million, representing 21% growth at the midpoint versus our 2022 revenue of $162.7 million. For the full year, revenue guidance reflects our decision to shift 100% of the consumer business to learning memberships by the end of the year, including the remaining professional audience. Revenue guidance reflects normal summer seasonality, including anticipated lower levels of new customer acquisition, consumption, and learning membership retention during the summer months when K-12 schools and universities are out of session. Additionally, revenue guidance reflects a higher level of high-dosage tutoring program utilization by school districts in the spring semester and a return to the normal seasonal pattern of starting new implementations in the fall when school starts, thus slightly shifting revenue into the first 2 quarters versus our prior expectation of consistent use throughout the summer. Our adjusted EBITDA guidance for both the second quarter and full year reflects the continuing benefits from our recurring revenue products, which focus on long-term relationships with higher-value customers and improving consumer gross margin profile and operating efficiencies stemming from our continued shift to recurring revenue business models. For the second quarter of 2023, we expect a non-GAAP adjusted EBITDA loss in the range of $3 million to break even. For the full year 2023, we are raising our targets to an adjusted EBITDA loss in the range of $7 million to breakeven. Full-year adjusted EBITDA guidance reflects the impact of normal summer seasonality and higher variable costs in the third quarter as we ramp into the back-to-school selling season, followed by a return to positive adjusted EBITDA in the fourth quarter consistent with prior guidance. Thank you again for your time. And with that, I'll turn the call back over to Chuck.

Thanks, Jason, and thanks again to all of you for joining us today. As always, we appreciate your interest in Nerdy and look forward to continuing the dialogue during this exciting time for the company. With that, I'll turn it over to the operator for Q&A. Operator?

[Operator Instructions] Our first question comes from the line of Ryan MacDonald with Needham.

Congrats on a great quarter. Maybe I just want to start first on the institutional business. Clearly, some really great momentum there with the record number of deals. I would just like to understand sort of how you foresee the demand environment on the institutional side progressing as we kind of get into the heart of the selling season here in 2Q and 3Q. And then as we look at the total contract value or the value of the deal count, obviously, it's down on a sequential basis. Just curious what you're seeing in terms of deal trends in terms of sizing around a number of modules or offerings being adopted or the size of deals in the pipeline.

Thanks, Ryan. This is Chuck. As you know, we landed our first teacher-signed contract. The momentum right out of the gate is super strong, and the feedback has been really encouraging, and engagement was much higher than we anticipated, which is excellent to see demonstrate a strong market fit, and we can now go out and kind of talk and demonstrate just how powerful why tutoring could be when teachers across the entire school district armor to be able to help students at district-wide scale the way that's highly individualized and takes into account their unique knowledge of that particular student and the curriculum that's being taught on the glass. So that was a huge accomplishment there. We obviously had a strong quarter in totality in terms of revenue as well. And as we head into summer and then back to school, what we're seeing is that the pipelines are continuing to grow. These are very large and very lumpy deals and conversations that we're having. So our last teacher-side contract, as you know, was approximately $5 million per year as a SaaS license that entitles the school district to district-wide tutoring that teachers can prescribe. And so these conversations involve many different stakeholders. They tend to be oriented around the start of the solar. So given the kind of magnitude involved, we wouldn't want to kick these off for just a few weeks or a few months because you really want teacher support, future buying where we kind of are very focused on making sure that seaters actually understand how to leverage the tools, and so all of that then lends itself to a series of opportunities that would start at the beginning of the new year. So we feel good about the pipeline value that's growing. We feel good about the bundled offerings and how together we're able to provide immense value and serve a multitude of district needs. And then as we head into the summer back school, I think we feel good about how many of the conversations we're having set us up for significant scale and opportunity this back-to-school season.

Super helpful. Great. I appreciate the commentary there. And then maybe just on the learning memberships, awesome to see the continued progress there. I guess I'd just be curious, as you continue to grow that member count, what you're seeing in terms of the makeup of those members, whether they were previous nervosity Tutors customers previously? Or if the learning memberships are sort of bringing net new learners to the platform.

Sure. Yes, we feel great about the fact that the business has grown about 63%, I believe, is the number from the fourth quarter to the end of the first quarter. So great progress there. In the totality of the active members, ballpark 75% were new to nerdy, and about 25% were legacy, folks who have been using us historically in a package model. And we're seeing in the field to all audiences. So there's kind of a good distribution among student ages, and we're seeing it resonate for different like high levels of subject usage and K5 that are a little bit different than middle school or high school or college or beyond. But in general, this idea of a comprehensive offering with live at the center that is been surrounded by classes and brings forms of pacing content and different modalities engaging, we think, really, really resonates here. So feel great about the progress and the offerings and engagement differ a little bit by age. But in general, we're seeing it resonate across all audiences, and it's just a more efficient go-to-market and one that makes it really easy to add additional product capabilities whether it's an AI tutor or some other form of concept like when we integrated Covis, we think it's a very easy and effective way that our customers understand to get more value over time.

Yes, Ryan, the only thing I'd add there, and we mentioned this on our last call, during the first quarter, we transitioned successfully the entire existing customer base within our academic subjects, which we feel really good about. We also transitioned the test prep audience. And then in our commentary, we shared that we're accelerating the transition of the professional business, given the positive feedback we've seen from customers as it relates to higher conversion, higher engaged higher retention. And then from a business perspective, it just allows us to drastically simplify the operating structure, which we feel really good about. Yes. And during the back-to-school season, we'll have an opportunity to reintroduce learning memberships for customers who would use package models in the past that currently are inactive.

Our next question comes from the line of Doug Anmuth with JPMorgan.

It's Bryan Smilek on for Doug. I guess just to start on the membership model. Can you just talk about the pricing strategy and structure over time? For example, should we expect a tiered strategy as the membership strategy proliferates over time? And then, I guess, just on the investment cadence, can you talk about any needs around AI and generative AI holistically going forward?

Sure. Thanks, Brian. Good questions. So we started off with a very simple kind of plain vanilla approach in learning memberships where we wanted to just effectively have one tier of offerings that would allow for us to go to market effectively last, call it, May and June, as we started leaning in the memberships aggressively. And initially, that was largely just one-on-one tutoring. And then over the course of the next couple of months in the fall, we started introducing all of these additional burden formats that provided more value including live classes in hundreds of subjects every week, we have a variety of different sincerest computer adaptive testing. We have covers in video game creation. And we started adding in other modalities like the AI Juno and chat-based uterine and a couple of other things as well. And so that then allows for us to have all these different ingredients or LEGO box that is a really, really compelling offering for our customers that can allow for us to support them over any academic calendar year across a variety of different subjects and then allowing them to learn how they want, when they want to live in the quarter stone. So in general, there's different amounts of one-on-one tutoring that you can get with your package. And one of the things that we've experimented with that has driven both conversion and retention was changing up the frequency. So you could appeal to people that are maybe a little bit more academically focused and are very concentrated in their efforts versus, say, a more casual burner in an area like foreign languages for adults. And so in totality, we are trying to take each of these underlying product capabilities and then find the content and the frequency that then allow us to really appeal to a given segment. So we've seen tremendous conversion wins through holding on that lever year-to-date. And I would expect that we continue to optimize those for given audiences over time. And then pairing that with your second question, the amazing thing here with generative AI is we are now able to create hyper-personalized content, we're effectively free for a given learner in a given subject across thousands of subjects and all different levels of complexity. So if you think about live at the center, the superpower we have as a company, high-quality live learning at scale and then surrounding it these other forms of content that can drive engagement that are highly additive that we could serve up in real-time in line. There's all sorts of different ways that generative AI can allow for you to get the exact right piece of content in the moment that you need it so you can continue on your vergence -- so we've talked about the sine have a comprehensive [indiscernible] and our ability to really live up to that vision and fill in any potential content holes that existed just became dramatically easier. So we are super excited about our ability to leverage generated by AI to create really comprehensive, really personalized experiences. And we're actively working on it today. We've announced a couple of those products this far, but we have many more in flight. I would just echo that we're going to deploy increased levels of capital against AI and engineering head in the near term to maintain a lead we've established an IST innovative speed and drive significant cost optimization as we move throughout the rest of this year, and we look forward to reporting results in the coming quarters.

Our next question comes from the line of Brett Knoblauch with Cantor Fitzgerald.

Congrats on the quarter. I guess could you maybe just give us a reminder on how the seasonality of the business is going to work now with the relationship model and how that might differ from the package model? And I guess, as we look at the second quarter, what should we be expecting from a membership count perspective in the change? And also, I guess you talked a bit about the institutional seasonality kind of gearing up for the back-to-school season and the start in the third quarter. Should we be expecting maybe institutional revenue to sequentially decline in the second quarter before kind of ramping back up in the back half of the year?

Yes. Great question, Brett. Yes. So consistent with prior years and seasonality trends, Q2 revenue is going to be lower than Q1. And as we previously discussed, we haven't yet gone through a transition of leading memberships from one scale to the next at scale. So we're being cautious with respect to learning membership retention and new customer acquisition during the summer months. So our Q2 guidance reflects typical seasonality, which is the result of a school calendar consumer purchasing and consumption patterns. I think it's important to also acknowledge the significant pull forward that we're seeing from busters for school's high-dosage tutoring business. We also have a portion of our business that's on the legacy package that we expect to sequentially decline period-over-period as we end the summer months. So net-net, we feel good about the anticipated levels of retention that are included in our forecast and guide, but we are cautiously optimistic there. And then from an active learner count, as of June 30, that would infer about 26,000 active members as we move through the summer, in which would then reaccelerate in the back-to-school period, consistent with historical norms and well into the fourth quarter. So we feel good about those trends that we're seeing, especially given, as Chuck mentioned on the call, the high levels of engagement, new ads in both Q1, April, and through May to date. Yes. So we had really strong engagement on the school side. We did a great job on the implementations. And we had previously assumed some of that revenue would kind of be keen to better across the first 3 quarters. We're just now assuming that it's a little more condensed in the first 2, but we feel really good about the engagement there and that how it goes for renewals kind of this back-to-school season.

Perfect. And I may make sure that I kind of understood you guys a bit correctly on the gross margin decline. So that was largely due to the institutional mix in that being such a large quarter in the institutional product because that's maybe not at scale enough where the gross margin is going to be kind of similar to the rest of the business?

Yes, that's right. So if you dissect the business into 2 parts, on the consumer side, we saw continued gross margin accretion, especially as we continue to mix shift towards a higher proportion of learning memberships. And then on the VarseTutors for the school side, if you remember, our new teacher-assigned product provides all access and unlimited support for the entire student population with the implementation that took place during the first quarter into the second quarter, we saw substantial levels of engagement which, frankly, we feel great about that there's product-market fit there but it did impress gross margins about 150 basis points relative to our expectations during the quarter, although we believe it sets us up strongly for expanding that product to additional school districts on the one hand and driving increased retention over time. And I think the important thing to remember is that while the revenue is recognized literally on some of these BTS SaaS products, what actually happens from an engagement perspective is the consumption goes up during the school year, particularly a period like January, February, March where it's all in the school year. And then as you go into the summer, you actually have probe levels of utilization, higher gross margin. So we think it's a great product overall. And then as Jason pointed out, the fact that the engagement was so high on its brand-new product. It's a really, really encouraging sign for other opportunities out there. And then more membership continues to drive significant gross margin accretion.

Our next question comes from the line of Eric Sheridan with Goldman Sachs.

Maybe 2 follow-ups on some of the comments that I've already made, just so we better understand. As you move towards the exit velocity this year and the rate of membership you're talking about, can you just help us unpack a little bit of how we should be thinking about sales and marketing efficiencies, not only in this year but as an exit velocity into next year? And how do you think about either harvesting some of those efficiencies from sales and marketing as you get bigger in memberships versus possibly accelerating membership adoption and investing them back in the business? That would be topic 1. And then Topic 2, you called out deploying capital behind AI. Can you help us better understand how much of these investments might be transient in 2023 to reposition the platform for what might be a different steady state or steady state-run rate of investments in AI over the long term? And any help you can in terms of comparing those 2 buckets.

Thanks, Eric. Good questions. So as it relates to sales and marketing efficiency, we drove 1,700 bps of sales and marketing efficiency year-over-year in the first quarter. And the reason we were able to do that was a couple. First, we have the new learning memberships model, which is significantly extending lifetime value to the shareholder loader. You can actually see just how much LTV has gone up on average relative to the old package model, but at a time and it's continuing to increase. Secondly, we focused our marketing efforts on appealing to customers who are looking for recurring support over multiple academic subjects, and multiple academic caliber years where we could serve them in a comprehensive way. And so by doing so, in fact, you're not focusing on what we previously described as, call it, the 7% of hyper transactional net bookings and you're really optimizing for the 93% that we think are the customers that want to be supported with tutoring and a comprehensive set of solutions like we have. Right now, we've been targeting marketing payback periods that are very efficient, call it, 6 months or less payback periods. And as we get through the summer and see how we're looking on those LTV curves as we head into the full year, there's going to be an opportunity to potentially change that payback a little bit and see if we're so excited about the Milo economics and have enough data, then we can lean in a little bit more aggressively. So the other factor here is not just getting more sales and marketing efficiency, obviously, but rather accelerating growth through changing that payback period a little bit and then acquiring a much larger number of users. So we're actively focused on things that we think will allow for us to do that, not baked into the plan, but we're focused on driving improvements to conversion, making our product more appealing, expanding the amount of content and offerings that are embedded within it and then separately driving retention over time. And through those 2 vectors, conversion, retention, those are the things that also allow for us to fundamentally open up the spigot on marketing in a very efficacious way. So I think we feel really good about the model today and the new customer trends. The new customer trends were pretty consistent throughout the first months of the year and are doing well in April and May, but there's definitely going to be an opportunity as we see some of that data to potentially move it a little bit further to drive higher levels of revenue growth. And then the second part, yes, on the capital side, I think the way I think about the deployments in AI, a lot of the infrastructure has been built. A lot of data has been captured and we've had more than 10 million hours of live instruction on the platform. Every part of that customer journey is instrumented. The wording itself occurs in a video session on the platform that then has computer adaptive testing and Q&A and other forms of content that you could engage in, all of which has kind of captured and fed into our data lake. And so a lot of the work has been done and the plumbing has been built over many years. to ensure that we're capturing the right information and then can use it to drive higher levels of personalization and engagement. So we have been adding to the team, but that's relatively modest in the scheme of things. And thus far, based on what I've seen, it's largely been self-funding. So we would expect to make some investments in personnel in particular, over the course of the next quarter or two, but we're getting incredible yields on our efforts and would expect that as we get into the fourth quarter and beyond, it's fully self-funding at that work.

Thank you. Our next question comes from the line of Andrew Boone with JMP Securities. Moving on to our next question from the line of Mario Lu with Barclays.

This is Alex Hughes on for Mario. Just two quick questions, if I can. On the active expert side, it looks like it was down about 10% this past quarter. And you made comments previously that you're focusing on allocating more work to fewer active experts. So just wondering if there's been any incremental changes in terms of how you're identifying and acquiring those experts. And if your current AI products or future AI products shift that philosophy even more? And then the second question, just is there any update on the American rescue plan dollars being deployed to Nerdy services? And if so, what does that look like for the rest of the year?

Great questions, and I can take the first one and then hand it over to Jason for the second one. So we have made a conscious effort to focus on acquiring and retaining top experts to do a disproportionately good job and making sure that they get a great experience in the sense that they're getting kind of consistent earning opportunities over time. So in that regard, earning membership has been terrific for the expert experience because they're meeting with students on a recurring basis, typically, call it, 1 to 2 times a week or the tri-semester schooler in pursuit of a goal or skill. And as a result, they're able to count on the earnings associated with that. And so one of the things that we try to do is rather than more kind of randomly or democratically distribute the different opportunities, we've continued to lean further into our machine learning matching algorithms that then disproportionately send opportunities to talk experts. And by doing that, we're then able to retain them longer, which, of course, that leads to a much better customer experience over time. So there's kind of a compounding flywheel aspect to this. And one of the things that we've also been able to do is also start getting better at predicting who won't be likely to get work on the platform and be as successful and using AI-predictive algorithms to try to predict upfront before somebody even joins the platform, who's likely to be a consistent expert who then drives high levels of customer satisfaction and takes on a number of students versus those that say maybe a little bit more transient, interested in a short amount of work, a smaller amount of work in better concentrate the relationships a little bit more, so we can deliver a better customer experience. Of course, there's costs associated with bringing on experts. So it's terrific for the business to be able to decrease the number of people who might join and not actually work with students. So this is a conscious effort. We feel great about it, and you're seeing it pull through to operating leverage.

Yes. And then just on your question related to Esterfunding, wanted to keep in mind, unlike the B2B space where companies are pulling down third-party spending, all the government funding for schools has already been provided the money in the market and schools are seeking out our solutions like ours to address the student and teacher shortage that we're seeing in the marketplace. So as of January 31, only about 28% of the funding has been spent according to the DEA start DOE and administrators are looking for long-term solutions like ours because of learning loss and teacher shortages or long-term issues. We believe our new prudent per-year programs like teacher-signed and on-demand, allow for long-term durability as the product offerings evolve to meet both those needs and support teachers as co-teachers in the classroom in their normal daily workflows. So just the amount of money still on the market still needs to be spent by September of 2024. But even beyond Esser, because of that support for teachers in their daily workflow, we've got normal operating school budgets, which are all running surpluses. And then beyond that, you've also got Title 1 funding, which was $19 billion in the most recent as bill and those funds are evergreen. So we feel like there's plenty of funding in the market to support continued growth within Vasteras. Yes, we feel really good about the momentum there. And we've obviously evolved our offerings from what had initially just been high dosage tutoring to one that's much more comprehensive and to these new SaaS models of feature signed where the teachers really are at the center of the offering, and we think that's really powerful here. So the fact that schools have been a little bit slower and we're serving and spending in totality, we think but it's all for us it's given us an opportunity to build this exceptional product that is highly relative to the scale. We think we're uniquely qualified to deliver high-quality live at scale in a way that hasn't been done before through our district partners and really help them accomplish things that might otherwise have been more difficult for them to accomplish as it relates to helping students.

Our next question comes from the line of Andrew Boone with JMP Securities.

This is Matt on for Andrew. Just wanted to ask on with Comiga being offered for $20 a month, is there something that you guys are going to have to do with pricing or packaging? Is the competitive set evolving around AI? And then maybe a second one just on institutional. Are schools just asking for anything on AI-based tutoring offerings? Or is this something that you guys are looking at on a product roadmap? Any color there would be super helpful.

Sure. Okay. Good question. So champions tutoring has been around for 20-something years, and it's been $20 a month. And we're excited about incorporating it as a modality offering, but it's one of many, many different ways that you can learn. So as we think about our AI tutor and what it's able to accomplish, it's something that is being used today for call it homework out -- it's being used for Q&A. And it's kind of taking into place all these content resources that historically has existed online. What we're seeing is that people are using it to engage additionally beyond their live recurring face-to-face tutoring session that are in pursuit of gold. And so we see it as another form of content creation and another modality of interaction. So from our perspective, like we're going to be adding all sorts of different ways to engage. And this is one of them that happens to be in the press a lot, we think is exciting, but we're applying AI across a wide variety of different forms of both content personalization, expert learner matching and then also applying it to drive operational efficiencies. And so as we think about schools, I mean, this is a really exciting opportunity. So if you think about all the technology that we've built and our application of AI to drive human interaction and give what we describe as superpowers to tutors and now educators, this is an area where we're actively seeing schools ask about it and at really how can we allow for their titers to drive higher levels of personalization in their data. So as an example, the idea of an IPR individualized education plan has been tossed around for decades is something that, in an ideal world, schools will be able to provide to each and every student, but they can't. They can only do it for a small subset of students. Now thanks to generative AI, that's the sort of thing where you could actually do it at scale in a way the hyper-personalized and aligns to both the state standards within a given state and the specific curriculum requirements of a given school and that would never have been possible. So we're super excited about our ability to take some of the products we're building and then extend them into schools in ways that are kind of wrapped around by and provide additional capabilities to teachers that give them leverage and help educators solve problems.

There are no additional questions waiting at this time. So I would like to pass the conference back over to the management team for closing remarks.

We'd just like to thank everyone for their time on the call today. As you can tell, we feel really good about the business while the transition, the learning memberships, and the expansion of Varsity Tutors for Schools as well as the application of AI throughout our business to drive continued growth and cost efficiency.