NRDY - NYSE

Good afternoon, and thank you for joining us for Nerdy's Fourth Quarter 2024 Earnings Call. With me are Chuck 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 Nerdy'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 fourth 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.

Thanks, T.J., and thank you to everyone for joining us today. In 2024, we unified our consumer and institutional offerings, improved our marketplace technology, and enhanced our all access subscription offerings. We had a strong fourth quarter, where we delivered both revenue and adjusted EBITDA above the high-end of guidance and we're entering 2025 with a stronger platform and are innovating at a faster pace. Our vertically-integrated quality controlled marketplace manages the entire learning journey, vetting, matching, scheduling, adaptive testing, live tutoring and continuous improvement directly on platform. This closed-loop system drives quality and retention. And in 2025, AI will optimize every step, from personalized lessons and practice problems to real time insights, live instruction and program reporting. Our goal is to make our learning experience the default choice for both learners and institutions. With over ten million hours of one-on-one tutoring recorded to date, we're in a position to benefit from AI driven enhancements that will allow us to elevate engagement and drive growth and profitability improvements. We are introducing several AI-powered products to our platform, that allow us to better provide learner and expert support throughout the learning journey. First, our new AI session playback tool that leverages generative AI to turn each tutoring session into actionable insights for learners, parents, and educators. Our platform now automatically transcribes and summarizes every session, highlighting key concepts and areas of strength or weakness and links directly to the relevant sections of the recorded video. For consumers, these insights help learners track progress and they give parents a clear view of their investments value. The initial signal is positive with tutoring session usage higher for those exposed to AI session summaries with a 99% positive feedback rate from parents and students. For institutions, teachers gain data-driven insights to refine instruction or interventions, while administrators benefit from transparent reporting. Real time feedback from these summaries also drives compounding improvements across our vertically-integrated platform, allowing us to refine service delivery at a pace not previously possible. As we move throughout the year, we will deepen our AI capabilities for institutions with dynamic exit ticket generation, leverage transcripts to generate homework after each session, and create lesson plans in advance of the next one. We will also add an advanced predictive analytics suite, including cohort level analysis and reporting tools, aiding district leaders in identifying at risk students and providing them clear visibility into program efficacy. We recently released our next generation AI lesson plan and practice problem generators to make creating robust customized standards aligned lesson content in seconds seamless. These tools are now available for both experts for tutoring and with our paid institutional product to teachers and initial signals been positive. This year, we will continue to build AI products for teachers and schools that make access to our platform more compelling and increase the value of the AI and software tools provided on the platform. Many of these tools also drive efficiency and benefit us operationally. For instance, by automating content creation, we're reducing operational costs and improving profitability. We recently introduced Tutor Copilot, a product that provides AI driven support that elevates real time instruction, enabling every tutor to deliver personalized high quality sessions. It's embedded in our live learning platform and Tutor Copilot offers instructional suggestions. It generates lessons and practice content, and it helps deepen student engagement. This product's an example of how we integrate AI into the learning experience in a way that enhances rather than replaces the personal touch that learners value. By reducing prep time and streamlining lesson planning, it ensures more time is spent on interactive live instruction. Looking ahead to the rest of 2025, we'll continue to enhance our AI-driven offerings to continue to refine and elevate the tutor and student experience. This year, we're planning to upgrade our AI tutor currently available for consumers and then make it available broadly to institutions. This will empower students and on demand support between live sessions, creating a cohesive learning ecosystem across both consumer and institutional audiences. We'll also refine our expert learner matching algorithm, incorporating more factors and data for even greater precision and improve match quality and ultimately customer lifetime value. Additionally, we're in the early stages of agentifying processes to increase quality and reduce costs. For instance, if a tutor isn't the right fit after a session, our AI can initiate and complete a replacement within minutes of a session completing, minimizing customer frustration and ensuring exceptional experience. Once we perfect Tutor Copilot, we'll adapt those learnings to launch a Teacher Copilot for our paid institutional platform, further embedding our tools and classrooms. For consumers, we will focus on driving retention and extending lifetime value through a higher quality and more personalized offering, building on our strong Q4 momentum. Over the past quarter, it's apparent that our progress on foundational marketplace aspects like vetting, matching and scheduling and more has positioned us to now shift greater focus toward AI-driven innovation. These AI products and tools are now advanced enough to drive significant improvements with less effort, while allowing us to compound many small wins across the entire learning journey. Now looking back on 2024, we made significant progress against the three primary themes we laid out last year, including scaling the winning product for every learner. We unified our consumer and institutional platforms to deliver a modern, intuitive and personalized learning experience. This streamlined approach enables faster feature releases, complemented by improved on boarding and activation for new learners. As a result, consumer engagement rose 26% year-over-year in the fourth quarter, and new customer retention improved. Key investments included My Learning Hub and subject portals, which enable easier resource discovery and progress tracking. We streamlined on boarding for a simpler path to getting started with a tutor, and we improved tutor matching to pair learners with the right tutor more effectively. We also emphasized learning memberships that focused on weekly tutoring sessions and that's led to quicker first sessions, more frequent tutoring engagement, increased engagement with non-tutoring tools and higher average revenue per member per month, as well as stronger retention. We expect these changes will drive a return to growth, enhance unit economics and support our path to profitability in 2025. Over the past year, we invested in our institutional go to market strategy and platform to scale varsity tutors for schools, and we expanded the number of learners we can impact. We productized and launched access to our platform, ultimately introducing offerings to more than 5 million students across 1,100 school districts. These efforts increased awareness in our company at a larger scale, built trust with K-12 institutions and positioned us as a preferred tutoring partner for districts implementing paid tutoring programs. In the fourth quarter, 43% of paid contracts and 36% of total bookings value came from school district partners who initially partnered with Varsity Tutors for Schools via access to our platform and then subsequently converted to our paid offerings. On a go forward basis, we'll be shifting our focus toward paid access to the Varsity Tutors platform for any new customers given the significant AI-led product enhancements that are being introduced in 2025. This past year, we made substantial improvements to modernize several components of our marketplace infrastructure, including scheduling, invoicing, substitution, and other logistics capabilities that we believe will allow us to provide a best in class customer experience. These improvements simplified operations, reduced tutor and marketplace operations costs and simultaneously improved the customer experience. We believe the recent advances in AI, provide us with the opportunity to drive further levels of productivity improvement, including through automation of key processes, that will allow us to improve both the customer experience and operational consistency, while also removing substantial costs. In fact, we believe headcount peaked during the back-to-school period in 2024, and we now anticipate that the productivity improvements driven by AI will allow us to scale without hiring in the majority of areas on a go forward basis. In closing, we're energized by the rapid pace of innovation that is positioning us to achieve significant progress towards our vision of AI for HI, where artificial intelligence for human interaction and what it can mean for our company. We look forward to providing you with additional updates as we embrace these new AI technologies and fulfill our mission to transform how people learn by meeting the evolving needs of learners in any subject, anywhere and at any time. With that, I will turn the call over to Jason to discuss the financials in more detail. Jason?

Thanks, Chuck, and good afternoon, everyone. As Chuck mentioned, we made significant progress against the three primary goals we laid out for the year. In the fourth quarter, Nerdy delivered revenue of $48 million, above our guidance range of $44 million to $47 million, which represented a decrease of 13% year-over-year from $55.1 million during the same period in 2023. Revenue declined primarily due to lower institutional revenue coupled with lower ARPM and active members in our consumer business. Consumer Learning Membership subscription revenue was $39.2 million, representing 82% of total company revenue. As of yearend, active members in ARPM were 37,500 members and $302 respectively, which resulted in an annualized run rate of approximately $136 million from learning memberships at quarter end. Our institutional business delivered revenue of $6.8 million and represented 14% of total company revenue. Varsity Tutors for Schools executed 91 contracts, yielding $4.6 million of bookings. During the quarter, we enabled access to the Varsity Tutors platform for an additional 600,000 students, bringing the total to 5 million students at over 1,100 school districts as of yearend. And as Chuck mentioned, this strategy to introduce school districts to the platform and ultimately convert them to our fee-based offerings produced 43% of paid contracts and 36% of total bookings value in the fourth quarter. As discussed in our last earnings call, since the start of the fourth quarter, we have moderated our investment in the institutional business to align with a more normalized sales cycle post ESSER and given near-term funding uncertainties. We believe a significant opportunity exists in the institutional space and that the product enhancements we are planning to make for the Unified platform will drive growth in future periods, simplify operations and positively impact financial results in 2025. Moving down to P&L. Gross profit of $31.9 million in the fourth quarter was lower by 19% year-over-year. Gross margin was 66.6% in the fourth quarter compared to gross margin of 71.3% during the same period in 2023. The decrease in gross margin was primarily due to lower ARPM coupled with higher utilization of tutoring sessions across learning memberships in our consumer business. We also implemented new expert incentives during the fourth quarter, that we believe will drive further engagement with our platform and customer retention improvements. Essentially for each consecutive session with a learner, experts receive an extra $1 per session up to $40 per session. This change is already leading to improvements in time to first session, the number of sessions in the first 30 days, a reduction in tutor replacement rates, and improvements to retention. Sales and marketing expenses for the quarter on a GAAP basis were $18.4 million, a decrease of $0.4 million from $18.8 million in the same period in 2023. Non-GAAP sales and marketing expenses, excluding non-cash stock-based compensation, were $17.8 million compared to $18.2 million in the same period in 2023. The decrease in sales and marketing expenses was primarily driven by consumer marketing efficiency gains, where we saw customer acquisition costs decrease by $1.6 million or 10% year-over-year in the fourth quarter. General and administrative expenses for the quarter on a GAAP basis were $29.9 million, a decrease of $0.8 million from $30.7 million in the same period last year. Non-GAAP G&A, excluding non-cash stock-based compensation expenses, was $21.6 million compared to $19.8 million in the same period in 2023. Included in G&A costs were product development costs of $10.4 million. We believe our investments in product development and our platform oriented approach to growth have allowed us to launch and continuously improve our suite of subscription products, simplify operations and enhance our unified platform. As Chuck mentioned, we believe that recent advances in AI provide us with the opportunity to drive further levels of productivity improvement. Through the execution of several internal AI-enabled productivity initiatives, we have been able to reduce corporate headcount by approximately 15% during the first quarter of 2025. The total annualized cash savings from these initiatives is approximately $6 million. We look forward to providing you with additional updates on our progress to increase automation, as we embrace these new technologies and streamline operations. Non-GAAP adjusted EBITDA loss of $5.5 million for the three months ended December 31, 2024, was above our guidance of negative $7 million to negative $10 million and compared to positive non-GAAP adjusted EBITDA of $3 million in the same period during 2023. Non-GAAP adjusted EBITDA improvements relative to guidance were primarily driven by higher revenues coupled with benefits from AI-enabled efficiency and operating leverage improvements, partially offset by lower gross margin, due to higher utilization across learning memberships in our consumer business. Compared to last year, the non-GAAP adjusted EBITDA loss was lower, primarily due to lower revenues and gross margin coupled with investments in the Varsity Tutors for Schools sales organization and product development to drive innovation and support our growth. As of 12/31, the company's principal sources of liquidity were cash and cash equivalents of $52.5 million and we have zero debt. We believe our strong balance sheet provides us with ample liquidity to operate against our plan and pursue growth initiatives. Turning to our business outlook. We are providing first quarter and full year revenue and adjusted EBITDA guidance for 2025. For the first quarter, we expect consumer revenues will be positively impacted by improvements in ARPM due to the mix shift towards higher frequency learning memberships and price increases, enacted in our consumer business. We also expect improvements to the user experience, including our earlier mentioned AI products and the implementation of new expert incentives will drive further retention improvements. For the full year, we expect a return to growth in consumer revenues, as the faster pace of product innovation and operational improvement initiatives pull through, leading to accelerating consumer revenue growth each quarter, as we move throughout 2025. Institutional revenue reflects the flow through of lower 2024 bookings into the first half of 2025, coupled with a cautious government funding environment. For the first quarter of 2025, we expect revenue in the range of $45 million to $47 million. For the full year, we expect revenue in the range of $190 million to $200 million. Turning to adjusted EBITDA guidance. For the first quarter, we expect recent expert incentives coupled with higher utilization in both our consumer and institutional businesses to result in lower gross margin. As we move throughout the year, we expect price increases for new consumer customers enacted during the first quarter will yield sequential quarterly improvements to gross margin. Full year adjusted EBITDA improvements reflect a return to consumer revenue growth coupled with benefits from AI-enabled productivity and operating leverage improvements, partially offset by these tutor incentives. For the first quarter of 2025, we expect adjusted EBITDA in a range of negative $6 million to negative $8 million and for the full year, we expect adjusted EBITDA in a range of negative $8 million to negative $18 million. We also expect to be adjusted EBITDA and cash flow positive in the fourth quarter of 2025. This would result us ending the year with no debt and cash in the range of $35 million to $40 million, which we believe provides us with ample liquidity to fund the business and pursue growth initiatives. In closing, thank you again for your time and for your continued interest in our company. With that, I'll turn it over to the operator for Q&A.

Our first question is from Andrew Boone with Citizens JMP. Your line is now open.

Hi. This is Brianna on for Andrew. Thanks so much for taking my question. As we look towards the guide, can you help us bridge the assumptions on the active learning numbers and ARPM for 1Q and just help us parse through the dynamic there? And then, when looking towards the full year revenue guide based on 1Q, this implies an acceleration throughout the year. Can you help us understand the visibility into the full year and confidence on that outlook? And then, second question, just what drove the pricing decision for new customers in 1Q? And how should we think about pricing potentially weighing in on member growth for 2025?

Absolutely. This is Chuck. I'll try to answer all of this. So we feel really good about Q4. We beat our own expectations by a significant amount on both revenue and EBITDA. We also made good progress on a number of important initiatives related to strengthening the underlying marketplace infrastructure that allows for us to be more efficient and deliver higher quality services consistently. So we felt like the underlying foundation was strengthened and we can start being a little bit more offensive, as it relates to some of the product innovations that we're shipping. So, a couple of the things that were particularly exciting about the quarter was a lot of the work that we talked about last time on consumer product and activation started pulling through to higher retention, and we saw that elevate both sequentially and year-over-year throughout the quarter and strengthened. And so, we launched tutor incentives, which was something that, historically, we have really not optimized very much and we saw significant changes in behavior that we got really excited about. So, tutors taking, call it, more than 3x as much work when they join the platform and then also driving significant utilization that we know pulls through to retention. So entering the start to the year, we saw higher level -- we decided to roll this out broadly given the really, really positive engagement it was driving, and then how that ultimately pulls through to retention. And to start the year, we've seen retention continue to elevate year-over-year, which is something we're really excited about. So that has an impact on gross margin in Q1, but ultimately, we think pays for itself a few months out with the kind of crossover line of contribution profit per customer being, something that as a result of being able to enlist and motivate the tutors on the platform to drive retention. We're pretty excited about how that can ultimately pull through. So we feel like the guide is appropriate and we haven't baked in all that much elevated retention, but the trends continue to improve and we really feel good about kind of the start of the year and where this could go as we continue to test. So, separately, we increased prices about 20% in totality, which we thought was appropriate given some of the value that we've been adding to the platform and the way that we've enhanced it, including, but not limited to both the tutor incentives and the quality improvements that we've made in the underlying marketplace infrastructure where we're matching customers on average to a much higher quality and better match for them, as well as then some of the ongoing product and personalization capabilities that we've rolled out, like the GenAI summarization, like the other content and tools, like Tutor Copilot. There's investments that we made behind them, and they're providing real value, and the customers are indicating that they actually value these. And so, we think that the pricing increases are appropriate and warranted and we feel good about those. And then, as we move throughout the year, you're basically blending into on average a higher ARPM base throughout the year and that kind of combined with, what we think are good customer acquisition trends, positive in February on acquired revenue, for instance, combined with that elevated retention, those are things that we feel good about, as we move throughout the year and will drive that acceleration.

Super helpful. Thank you so much.

Our next question is from Ross Sandler with Barclays. Your line is now open.

Hi, guys. It's Ross Sandler on for Ross Sandler. Chuck, how is the new AI products impacting retention in the consumer business? And then second one is sort of a follow-up, but Jason or Chuck, either one, could you just explain the gross margin and the higher utilization dynamic that happened in the fourth quarter? And it sounds like that's going to improve across the next few quarters. Could you just kind of explain what's going on and then why it's improving? Thank you.

Sure. So, there's a couple of different ways. So, it allows for us to better assess quality, improve match quality. We know those things both pull through to better retention. There's features that more directly impact a day-to-day customer experience like the Tutor Copilot, so tutors are able to show up to sessions and be more prepared in general, same with lesson plan generation and practice tools. Those are all things that allow for, on average, the sessions to be more productive, more personalized, more relevant, and also for experts themselves to have higher effective hourly pay, because they spend less time prepping but are able to actually deliver higher quality life lessons. And you take that and you combine it with some of the features like the GenAI summarization that actually allow for parents to see the value in particular of their kids' lessons and very, very quickly, let's say a 60 to 90 minute lesson instead of having to watch the whole thing, which can be tedious, immediately see the value and better understand what's been accomplished. And that's something where the engagement is really high. It's pulling through to customers, making it farther along in their journey from the first week to the second week to the third week to the fourth week to the fifth week, and so on. And so, that's something where, we're really excited about how both the product that we launched as well as subsequent improvements we'll be able to make that could get really, really cool, in terms of being able to see some of the insights that come with it. Those are things that are just allowing us for allowing us to communicate and demonstrate the outcomes and value that much more efficiently to consumers. They're also able to, watch it and jump to the right spot more efficiently. There's a whole host of different ways that, we think that particular product can be really, really compelling. But the whole idea is that, it's less work, you get more value, it's easier, it's more personalized and we kind of remove friction at every point in a way that really delights people.

Yes. And then I'll speak to the gross margin and the utilization. I just said, like to start by saying we love seeing higher utilization on the platform, higher levels of activation, higher levels of first time to first session, the number of sessions in month one and month two, across new customer cohorts based on all the UX work that we did over the course of the last two quarters is really starting to pull through. We always have seen historically that higher levels of engagement and utilization, drive higher levels of retention and we're seeing that. We noted it in the letter and on the call that it's up 26%. That does impact gross margins, although we think it benefits contribution profit dollars over the long run, because customers will stay with us for a longer period of time when they're having a great experience on the platform. The other thing on gross margin is, the tutor incentives that we are investing in are across all customers. And so, there's a significant impact in the first quarter, to that gross margin. But as the new prices take hold for new customers and blend in over the course of the year, you'll see sequential improvements in gross margin, as we move across the year. So, I've got gross margin of 60% in Q1, 64% in Q2, and in the back half, we're in the 66% range as those price improvements pull through. And just from a funnel metrics perspective, we feel good about those price improvements that we've incorporated this past month in February, ARPM is over $400 and we think that the customer remains healthy.

Our next question is from Bryan Smilek with JPMorgan. Your line is now open.

Hi. This is Niraj on from Bryan. Just wanted to understand on VTS, are you starting to see returns from the localization of the sales force? And more broadly, can you give us an update on the competitive landscape?

Would you mind repeating your question, please?

Yes. on Varsity Tutors, are you starting to see returns from the localization of the sales force? And more broadly, can you give us an update on the competitive landscape?

Certainly. So, one of the things that, we've done over the course of the past year is build out our all access product, so that we can actually offer subscription offerings that allow for us to make available all those different solutions, products and schools that, in large part, we built first of our consumer business but have broad applicability within VTS. And so one of the things that, you know, we've done is now take recently a number of these different GenAI capabilities and include them in an offering that will now be paid subscription offering, that we will be selling into school districts and available for many of the 1,000 school districts that signed up for the free offering a year ago, as well as any net new customers, that being kind of the default product that they're opted into with the ability to pair that with high dosage tutoring offerings. And we think it's really compelling, and particularly as you think about the insights components, that ability to look at it over cohorts of students, be able to get predictive insights into when intervention's appropriate, is something that we think is going to be really, really compelling. So kind of given some of the uncertainty surrounding government funding right now, one of the things that we did throughout Q4 and Q1 was that, reduce that go to market team function to kind of right size for the near-term uncertainty, which we think is appropriate and is reflected in the guide. And we're going to let the product enhancements that have recently rollout and will continue to rollout drive a lot of that growth. We think the product is about to get, like, way more compelling and already has. And it puts us in a position where we have a highly-differentiated next generation AI-enhanced tutoring platform and intervention platform that school districts will find really compelling and some of the initial signals quite positive. So the competition has been. There were a number of competitors that went, under or reduced size in Q4 and many of these were in the kind of chat-based tutoring world, not in live like we do. But we think we're in a position to be a great partner for school districts over time and that the enhancements to the platform, many of which we're building for consumer and that are extensible to institutional, are going to be really, really compelling. So, we've taken a conservative stance in the cost structure given some of the uncertainty, but we think that we've preserved the innovation capacity, and again as some of these product enhancements hit, have a very compelling upside potential. So, we're excited about that.

We have a question from Jason Tilton with Canaccord Genuity. Your line is now open.

Good afternoon. Thanks for taking the question. Following up on one of the earlier questions, it sounds like the early signs from these new AI tools have been really encouraging. I'm wondering how much of a benefit have you contemplated from these tools, within your fiscal 2025 outlook, both on the revenue side and on the expense side?

On the expense side, we're kind of fully reflecting the costs in our plan that we would expect to incur. We're also expecting and including in the plan heavy levels of utilization, just to be conservative, as it relates to gross margin. And we think that, we've taken line of sight forecast on the retention side that we feel good about, but there's the potential to outperform as some of those product enhancements hit and we're excited about the extent to which they're resonating. So, we are not, in the plan baking on a lot of upside and are encouraged to start the year with the fact that retention is up quite a bit year over year and improving, but there's a potential that should we continue to execute that that could exceed and be a good guide.

The only thing I'd add is a lot of the cost associated with AI internally will drive substantial customer activation, and process automation, which we think will be self funding over the course of the year.

We have a question from Alex [indiscernible] with Raymond James. Your line is now open.

Hi, Jessica on for Alex. So kind of follow-up on all these questions and we're really excited hearing about all the AI initiatives you guys have. As you're thinking about your growth expectations, what would you rank kind of, as the key drivers for returning to growth or just like executing with customers? Or, you think it's like the increasing ARPU you've talked about or you think it's your increasing asset member space or just what factors how would you rank them? How do you see that all shaking out with the growth expectations?

Yes. So we feel good about the kind of top-of-funnel trends and the fact that we're heading in the right direction, our products are resonating more, conversion related metrics are improving and that looks healthy and improving. The rate of retention in relative terms is a bigger driver, because it's up and accelerating more year-over-year. And that combined with higher prices, higher ARPM, then blends up over time. So you have higher paying customers, who are getting better retained over time, which drives the return to growth. So, we feel good about that dynamic and then the relationship between the product enhancements and the continued work that we're doing and that then pulling through to retention and growth.

And then, Jessica, just to add to that, those drivers Chuck mentioned, retention, acquisition funnel metrics, those all continue to expand in the guide, as we move throughout the year. And so, sequentially within the consumer business, you'll see continued growth each quarter, as we move throughout the year. That also takes into account the seasonality of our business. And then, on the institutional side, we had lower bookings in the second half of 2024, which we talked about on the last call during the back-to-school season. That will carry into the first half of the year, where we'll also be negative on the institutional side. And then, as we move into the back-to-school period that business will return to growth.

Got it. Super helpful. Really quick follow-up as well. When you say, I think your headcount have peaked during the back-to-school period in 2024 and you're saying all this AI productivity. Are you planning on just, like, keeping a stable headcount going forward, or do are you displaying natural attrition, to do with work while you still get efficiency from AI?

Probably the latter. I mean, we're seeing meaningful improvements in team productivity. And across most teams, we wouldn't expect to backfill kind of natural attrition. And we've set expectations there internally. And we feel like it's appropriate and healthy, and we're giving our teams the tools and capabilities they need to be super productive. And, we feel good about that kind of general trend. And, as folks contract out of the organization, and we think we'll just get progressively more efficient throughout the course of the year on both the variable and fixed side.

Yes. I think it's important to note, like, we'll be able to continue to scale without commensurate levels of new variable headcount on a go forward basis. And then, maybe just to touch on the fixed headcount change that we made during the last two weeks, that's been completed. We've preserved all of our innovation capacity and we feel really good about the go forward team, as we move throughout 2025.

Our next question is from Yi Fu Lee with Cantor Fitzgerald. Your line is now open.

Thank you for taking my question and congrats on the stabilizing finish to 2024. So my first question is multipart. It's in regards to the AI-driven products, whether it be summary insights, lesson plans or copilot. I was wondering if Chuck or Jason, you could comment on how do you balance between like the human interaction side versus using too much automation to drive out the human aspect of it? And then can you comment on the back-end technology? How do you build this? What is the technology back-end you use to build this product? And lastly, for this particular question is the go-to-market. Do you intend to like invest more to train folks to, I guess, sell these new products? And then, I have one more follow-up with Jason on the financial side.

Sure. So, we've long ago, I think in 2017 for the first time, I believe we got the trademark in 2020 for AI for HI, artificial intelligence through human interaction, and the concept of giving superpowers to both experts and learners in a way that actually enhances the experience. So our guiding North Star here is that, product enhancements, whether it be AI or something else, need to add value to the experience. In many cases, we're able to better personalize the experience. You're getting hyper-customized content and recommendations. It's removing friction, and you're getting a more relevant tutor paired with you. We're able to identify and vet in a scalable way increasingly higher quality tutors. We're able to serve up hyper personalized lesson plans, based upon what you actually studied and met with on the previous time that removes friction, increases the tutor's effective hourly rate and reduces time spent prepping or creating problems in sessions so that the actual like instructional time goes up. So we think, like, these are all collectively positive and significant enhancements where you're not trading off one for the other, that they're multiplicative and that we're able to allow for humans to do what humans do best with technology and AI surrounding them. There's more than 10 million hours of live tutoring that have happened on our platform. And unlike a lot of other marketplaces out there, the entire experience happens on platform in that actual session. And so, we're able to use our learnings from that to then personalize the experience in a way that, we think is hyper relevant and leverage custom LLMs that allow for us to get kind of incremental personalization, and recommendations in ways that are highly additive. And we think that that vertical integration of the marketplace is really unique. And we're responsible for guaranteeing quality, but then benefit from the lifetime value extension. And that's a dynamic that we think is really powerful and is set to compound in the months and years ahead, as we basically improve every step of that kind of customer journey and optimize it. We think that can really sink and all those kind of 5%, 10%, 20% little improvements in different steps add up to something that can be really, really meaningful. And the hit rate, is such, like, there's a lot of work that you asked about the underlying infrastructure that's been put in place. We first started using machine learning for matching algorithms back in 2017 and put in place a lot of the underlying infrastructure for both the data engineering side and warehousing as well as then how we ultimately train models. And we're in a position, where we can run quickly to take advantage of some of the latest technologies. So there's a whole host of different technologies that we use, but that kind of underlying data infrastructure allows for us to personalize efficiently and move quickly. And we feel like we're now about to hit our groove in that regard.

The only thing I'd add to that from a go-to-market perspective is these AI tools support not only students and experts, but they also support teachers and administrators at schools. So, we're covering off the entirety of the ecosystem from a user base, which we think helps to drive that go-to-market motion across both the consumer base and the institutional base. One thing that we are going to do as we move into 2025, as it relates to platform access, we will have a paid tier, because of all the advancements we'll have within AI, as it relates to the institutional space. So, that will be new, but otherwise, we think that it just overall enhances the value proposition that we can provide to all of our customers on the one side of the marketplace and then experts certainly on the other side.

Yes. And you asked about training on the institutional side for our sales organization. Certainly, some is required and I don't want to trivialize that, but many of these are very intuitive customer-facing tools, where the value that they provide is easily understood. So, the summarization, the insights, you can actually see the session, you can read it, you understand intuitively what's happening. How it's produced on the back-end, frankly, doesn't really matter from the customer's perspective, but they're able to experience these products and immediately see the value, which we are hopeful will make them far more sellable, sooner.

Okay. And then, Jason, just on the follow-up...

We're excited about how this is going to enhance productivity, will offer better us to get better program efficiency for schools and generally just provide way more robust reporting and value and build deeper relationships with districts. And one of the great things is we had to build many of these tools for our consumer business, first, we're now able to then leverage them in the institutional business. So we'll build it real quickly.

Extremely clear. So it's more like internal organic. You guys built this right before the OpenAI, trades, so it's more internal. Got it. And then, Jason, in terms of the financial model, I want to ask a question. You guys gave a lot of commentary on the outlook on the prepared remarks. I was wondering if you could give us a little bit more guidance on the split between institution and consumer. We just want to get the model right in terms of seasonality. It sounds like consumer is going to grow a little bit for the 2025 quarter-over-quarter, but it sounds like from your commentary that, the institutional side is going to be a little pressure. We just want to know the split, that's what we can model properly. And also, Jason, you also mentioned that obviously using AI, you guys get more operating leverage, right? I was wondering, I know it's too early, you already got to EBITDA negative for this year. I was wondering when will this AI, help you to achieve faster breakeven time? And that's it for me. Thank you very much, guys.

Yes, good questions. Thanks Yi. On the mix shift between institutional and consumer, we would expect institutional to be about 15% this year of the total. And that business, given the uncertainty of government-funding that exists in the market today, we have that business largely flat from a bookings perspective year-over-year. And then when you think about the GAAP revenue and how that flows through, I mentioned this a little bit earlier on the call, the first half will be down and then you'll start to see some sequential improvement, as we move into the back-to-school period. And then, maybe just shifting to AI and profitability, I think you see it in the fixed headcount side, where we don't feel like we need to backfill roles as attrition occurs. We'll have very limited open new roles in the company. And all of our engineers, as an example, one of them told me the other day a project that would generally take them 60 hours actually took him only six hours. So, the level of improvements in productivity, whether it's depending on which LLM they're using is substantial and those just accrue to the company's benefit from a profitability perspective.