Chegg, Inc. (CHGG) Q2 2023 Earnings Report, Transcript and Summary

Greetings, and welcome to the Chegg Second Quarter 2023 Earnings Conference Call. At this time, all participants are in a listen-only mode. A brief question-and-answer session will follow the formal presentation. [Operator Instructions] As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Tracey Ford, Vice President of Investor Relations and ESG.

Good afternoon. Thank you for joining Chegg's Second Quarter 2023 Conference Call. On today's call are Dan Rosensweig, Co-Chairperson and CEO; and Andy Brown, Chief Financial Officer. A copy of our earnings press release, along with our investor presentation is available at our Investor Relations website, investor.chegg.com. A replay of this call will also be available on our website. We routinely posted information on our website and intend to make important announcements on our media center website at chegg.com/mediacenter. We encourage you to make use of these resources. Before we begin, I would like to point out that during the course of this call, we will make forward-looking statements regarding future events, including the future financial and operating performance of the company. These forward-looking statements are subject to material risks and uncertainties that could cause actual results to differ materially from those in the forward-looking statements. We caution you to consider the important factors that could cause actual results to differ materially from those in the forward-looking statements. In particular, we refer you to the cautionary language included in today's earnings release and the risk factors described in Chegg's annual report on Form 10-K filed with the Securities and Exchange Commission on February 21, 2023, as well as our other filings with the SEC. Any forward-looking statements that we make today are based on assumptions that we believe to be reasonable as of this date. We undertake no obligation to update these statements as a result of new information or future events. During this call, we will present both GAAP and non-GAAP financial measures. Our GAAP results and GAAP to non-GAAP reconciliations can be found in our earnings press release and the investor slide deck found on our IR website, investor.chegg.com. We also recommend you review the investor data sheet, which is also posted on our IR website. Now I will turn the call over to Dan.

Thank you, Tracey, and welcome, everyone to our 2023 Q2 earnings call. Our team executed well, outperforming guidance for both revenue and adjusted EBITDA. As the second quarter progressed, we saw year-over-year trends for customer acquisition and retention rates improve, which drove the upside in our results. We are entering an exciting new chapter for Chegg catalyzed by the advances in artificial intelligence. To take advantage of these new opportunities, Chegg is rapidly pivoted because we believe that category-defining companies with strong brand loyalty, sought after services and highly valuable data sets can leverage AI to grow and will create outsized returns. It is still early and since we last reported, we gained greater insights into students use and perceptions of AI and how it relates to Chegg. Our recent survey shows student see ChatGPT and Chegg as complementary with very different use cases. The latest wire poll survey states that while Gen Z students are using AI to improve their education, they are not comfortable with the exact information ChatGPT could tell. And it's become clear to us that a simple, high-quality, accurate personal learning assistant is needed and we feel we are uniquely positioned to deliver a world-class personal learning system. We are moving fast and launched the beta version of our initial generative experience in May. Feedback has been very positive. Specifically, our students like our simple user interface, which is conversational, and they have always trusted the quality, accuracy and relevance of our proprietary step-by-step solutions. Our research also shows that 86% of students say that they prefer study help that is reviewed by human subject matter experts, and 85% say they wanted to be personalized to their individual learning needs. So it is no surprise that engagement from our beta testers is extremely high, and they are interacting more with each question and are staying for significantly longer sessions. We appreciate that speed and execution are critical to our success. Our partnership with Scale AI announced today will allow us to accelerate our ability to deliver the new Chegg experience starting in the Fall and rolling out over the course of the next two semesters. The new Chegg will combine the best of generative AI, with Chegg's proprietary high-quality solutions and demonstrated ability to improve student outcomes. They can expect to see a much simpler conversational user interface, personalized learning pathways, more in-depth content and the ability to transform it automatically into innovative study tools such as practice tests, study guides and flash cards. In order to further enhance our competitive moat and lower our costs, we are building our own large language models, which gives us the ability to train them specifically for education. Our LLMs will be trained with our unique data sets and with the help of our 150,000 subject matter experts, we expect to deliver a significantly enhanced and differentiated learning experience for students compared to the generic models that are available today. And this is just the beginning. I want to give you a sense of how big we believe this TAM expanding opportunity can be and how we plan to capture it. We intend to build the largest connected community of learners around the world. With a truly scalable, affordable adaptive learning assistant by combining the tools, pathways and the accuracy that students depend on. Chegg's proven learning taxonomy, along with our deep history of data from schools, classes and professors sets us apart. We have said for years that students challenges go way beyond the academic need and now by leveraging advancements in artificial intelligence, we believe we can make a significant impact on reducing the nearly 40% of students who drop out of the higher education system and the more than 50% that never enter. Increasingly, students are connecting their academic journey with their skills-based needs in order to be employable in today's economy. Chegg is developing integrated skills pathways that will help students assess their current proficiency, identify their gaps and then help them acquire those skills. We are in a great position to do this by leveraging our skills offerings, where we continue to see excellent growth. We also appreciate that students today face a wide variety of personal challenges that can get in the way of graduating on time or at all. We know that if we can connect students to solutions that address some of these issues, such as mental health, food and security and financial barriers, we can improve their chances of finishing their education and thriving. We have created a concept video for you, which illustrates how this may all come together, which you can review within our investor deck posted on our IR website. More than 50% of the world's population is below the age of 30 and they have increasingly turned to online to advantage themselves academically and professionally. Now aided further by the proliferation of AI, the opportunity for Chegg to serve them is bigger than ever. And with that, I will turn it over to Andy. Andy?

Thanks, Dan, and good afternoon everyone. Q2 was a good quarter as we exceeded our revenue and adjusted EBITDA guidance and also delivered strong cash flow. Total revenue was $183 million, driven by subscription services revenue of $166 million. During the quarter, we had approximately 4.8 million subscribers on our platform. Sales and other revenue was $17 million, driven by strong growth in skills, offset primarily by the change in the required materials model, which is now a revenue share. Gross margin of 74% came in slightly higher than expected. This, along with the revenue beat contributed to adjusted EBITDA beating guidance, which came in at $60 million or 33% margin. Free cash flow was $56 million, the result of strong operating performance and higher interest rates with interest income contributing $10.7 million in the quarter, an increase of $8.7 million from last year. We had several items that impacted our GAAP net income for the quarter. These included a gain of $53.8 million from the repurchase of some of our outstanding convertible debt, which was partially offset by a restructuring charge of $5.7 million we announced during the quarter and a loss contingencies of $7 million we accrued related to a previous gain taken on an equity investment. We continue to have a strong balance sheet and drive significant free cash flow. We ended the quarter with $808 million of cash and investments, with total convertible debt outstanding of $773 million at par value, representing $35 million of net cash. As mentioned earlier, we repurchased $427 million of our outstanding convertible debt of $369 million and use some of the net savings to retire 3.4 million shares of our common stock for approximately $35 million. We continue to believe the combination of our operating model, balance sheet and cash flows are among the strongest in the education industry and will allow us to deliver attractive results to our shareholders. As Dan mentioned, we are rapidly realigning our resources around AI efforts, including partnering with scale AI to develop large language models required for students to have a fully generated conversational experience rolling out over the next two semesters. We believe our approach of developing and owning these models versus solely relying on third-party providers will create a truly differentiated and better experience with students at a lower cost. Now moving on to guidance. For Q3, we expect total revenue to be between $151 million and $153 million with subscription services revenue between $135 million and $137 million, gross margin between 68% and 69% and adjusted EBITDA between $34 million and $36 million. It is worth noting that we typically experience seasonally lower revenue and margins in Q3. We also have an elevated level of content depreciation from recently acquired Professor-led material, which is impacting gross margins. We expect the impact of this to moderate in Q4 and margins improve. In closing, we expect the development of AI will allow Chegg to embrace a much larger opportunity over time. We believe there is nobody better equipped to meet the current or future needs of students than Chegg. We have an industry-leading brand for proprietary data, strong operating model and a balance sheet to extend our leadership into the future. With that, I'll turn the call over to the operator for your questions.

At this time, we will be conducting a question-and-answer session. [Operator Instructions] Our first question comes from Doug Anmuth with JPMorgan. Please proceed with your question.

Thanks for taking the questions. Can you just talk about how the new AI experience with Scale AI will differ from CheggMate and kind of the path that you've been going down? And then, I guess second, if you could also just talk about the drivers you think of the 2Q improvements around customer acquisition and retention as you went through the quarter. Is that just distance, you think, from the ChatGPT launch or more of the students recognizing perhaps some of the deficiencies on that side? Thanks.

Yes. No, great question. I'll take the second question first, which is I think it's what you said, which is -- and the research seems to indicate that, which is once students recognize that they do very different things, and that what we do, they can't do and what we do is actually what they need because it's been built directly and specifically for students that they will use ChatGPT, but they're not going to try to use it for the things that they use Chegg for. And our current research seems to indicate that the overwhelming majority of those that use both plan to remain with Chegg. So I think that's very good news, and it's an update on what we've seen, and I think the results seem to suggest that. So that's really good news for us. The second thing -- the first question that you asked, which is how does it differ? Really, what it is, is instead of building a separate product, we have in our ripping Chegg down to its does and completely rebuilding the user experience so that it's available to everybody that uses CS and CSP. So it is the same vision, and that's why we supplied a concept video for people to really get an understanding of just how great that this can be and how different it is from the existing experience, how much more conversational, how much more simple, easy interface, conversational in nature. And one of the really cool things that we'll be able to do differently than anybody else would be able to do is take the 100 million-plus questions that we have and all the data we've been able to collect and create completely personalized learning experiences on a per user basis based on knowing not only the history of that particular student, but others that have gone to that school, that class and with that professor. So that is not something that any generalist AI can do or frankly, anybody else in the education space could do because we have the largest direct-to-consumer list. So that's pretty exciting. And then to take the way they learn best and actually build study tools out of the content they're using without them having to do anything. So I think our ease of use and the difficulty in using ChatGPT, the fact that it's overwhelmingly for writing and not for learning. I think those things were pretty significant in affecting those trends to the positive which is great for us and great for our shareholders. And then instead of it just being CheggMate, it's just going to be Chegg. What Scale AI does is something very different. We continue to work with ChatGPT just as we announced and that's all the conversational nature and the descriptions and explanations and things of that nature, which is what it does well. What Scale AI does is allow us to use what Chegg uniquely has, which is our data, our history, our user information, our content and create very specific and unique learning experiences for each student. So this is the differentiator. We were always planning on building this, working with Scale AI will allow us to do all of the categories we have much faster. So instead of launching with just a few, we will be doing a rolling launch and get to all 26 of them in a much more rapid period of time. So it's a great deal, they're an extraordinarily great partner, they work with ChatGPT. They work with everybody, but we will continue to work with all the relevant generalists that can supply value inside of our system. But what Scale AI does is allow us to leverage what we uniquely have that no one else has and everybody else wants.

That's very helpful. Thank you, Dan.

Yes, appreciate the question.

Let me just fill in here. Dan mentioned earlier in the answer to that question. CS and CSP just so everybody is aware on the call that he's referring to Chegg Study and Chegg Study Path. That's our internal lingo that he was using. Thank you.

Our next question comes from Jeff Silber with BMO Capital Markets. Please proceed with your questions.

Thanks so much. I'm just wondering if you're seeing any different trends in your subscribers between the U.S. and some of the larger markets that you serve overseas?

Well, the summer school is really a U.S. experience more than it is outside the U.S. So most of the data we have over the last few months in terms of actual user behavior is U.S.-based. So I really can't speak to outside the U.S. What I can say for those that we have outside the U.S. that have been testing our new experiences, the results have been very, very similar, which is significantly increased engagement, length of time using it because they're getting access to more information and asking more and more questions that we will have in the system already. So -- but it's too early to give international lens versus the U.S. I think we'll be able to do a lot more of that on the next call because really, the semester picks up in about two weeks. So sort of the last week or two of August through the first two weeks of September is the overwhelming majority of when the good stuff happens.

Okay. That's helpful. I know -- I'm sorry. In the past you've talked about the strong visibility in your business. And I'm just curious, what do you need to see before you think you'll start talking about annual guidance again?

Yes, we ask ourselves that same question, and I'm sure Andy is going to have some perspective on this. But I think for the moment, given the volatility of how people are reacting to information, our lens right now is to focus on what's right ahead of us, which we have really great confidence in, and as we build a few of those, I think it leads us back to where you're asking. What I will just add to what I said earlier is as the semester evolves, third, fourth week of August, first couple of weeks in September, that gives us a really good lens into obviously, to Q4. And since Q1 is a rollover from Q4, I think we're getting closer to that. But I think we're just going to stay with what we're doing now, because it's just a more conservative approach to things. And I think we learned the impact of the volatility of people just, in my opinion over reacting to information that we provided. Andy, I don't know if you have more to add to that.

No, Dan, I think you nailed it. I think -- and Jeff, you're aware, obviously, the season start to pick up, like Dan said, in the next couple of weeks. If we continue to see the trends we've seen over the last few months, that will certainly give us more confidence, but we need a few more periods before we get to that point.

Okay, fair enough. Thanks so much.

Our next question comes from Eric Sheridan with Goldman Sachs. Please proceed with your question.

Thanks so much for asking the -- for allow me to ask two questions, if I could, quickly: One would just be the cost question. So when we talked three months ago, you laid out a strategy around CheggMate and OpenAI. With this new strategy, how different is the potential cost or investment implications for this approach versus the prior approach? And how should we sort of bring that back and reflect that in cadence of either margin or needed to invest in the business? And then I'm just a little bit unclear. So the second one will be a follow-up. When people go back to school over the next month, this will be a solution that builds momentum as you get deeper into the year and the solution you're talking about today would be more implemented towards the end of this calendar year and the beginning of next calendar year? Or will it be deployed over the next one to two months? Just one would be timing and one would be depth of investment. Thank you.

Yes, I'll handle the second one first, and then I'm sure Andy is going to want to handle the depth of investment, which I think you're going to be happy with the answer because we are. But on the second one, consider it just a forever rollout, which is the product is just going to get better every day, let students use it. So Depending on what subjects they use or what they use it, they will see the product evolves. So it really is that way to think about it, which is they'll start to experience certain aspects of the conversational nature and the generative AI kind of content, depending on the subject matter they're doing, they'll start to experience that in the fall. So it's a little bit like rolling thunder and we expect word of mouth, viral nature of it, all the things that have built Chegg over the years to start to spread over that period of time. But I think right now, what has spread is the quality of our content and the accuracy of our content, which are essential for people trying to learn. So I'll let Andy talk about the cost structure and the cost. But I think again, I think you're going to find them to be constructive and positive.

Yes. So when we talked about this on the last call, there was a lot of moving parts. And there still is to some extent, but as we evaluated the multiple options that we had as far as having a fully generative -- conversational generative experience for our students. When we evaluated this, the option of partnering with Scale AI became by far a few things: One is it made us get to market sooner and it was the lowest cost versus completely leveraging third-party technology. And so as we look at this, this is by far the least cost-effective way -- as soon as we can get -- the least cost effective way, yes, the least the most cost effective way and like I said, we've said in the past, we think that this will allow us to maintain or even potentially over time, increase margins and we believe over time will allow us to actually deploy less CapEx as we implement these solutions. So this is -- it was really a big win for us. And as it rolls out over the next two semesters.

Great, thank you.

Our next question comes from Stephen Sheldon with William Blair. Please proceed with your question.

Hey, thanks for taking my questions. I guess on the improvement that you talked about during 2Q and customer acquisition and retention, I would love some more detail there, especially how you're measuring that improvement, if there's any kind of rough quantification you can provide? And also, did you see some of those trends continue into early third quarter, thinking about July?

Yes, so what we're referring to are, obviously, everybody that runs a company should know the levers in their business. Andy has explained subscription math multiple times. And -- so it really is, what is your retention rates and in retention rates, you look at a lot of different variables, which is what's up for renewal, what canceled, what percentage renewed, those kinds of things. So we have been seeing a decrease in cancels, which is excellent. And we have been seeing an increase in people that are up for renewal that renewed. So those metrics are basically we're seeing higher retention rate, which is great. The second one is new accounts, which is really where all of this challenges started, because we used to do -- before COVID, we used to do about 3.2 million new accounts a year, and we peaked at peak COVID about 5.8 million. Now we have remained above 5 million. So we're really so far ahead of where we were before COVID, that the company has just accelerated. So it always surprises me when people don't really understand how big we become compared to what we were just a few years ago. But what we're referring to now is the trend in new account growth. So we were seeing declining new account growth and it was pretty substantial, and that is improving each and every months, including in July. So we're getting closer to our objective of returning to growth.

Very helpful. Good to hear. One quick follow-up. I just noticed that you didn't use CheggMate branding at all in the press release or prepared remarks. Was that intentional? And I guess are you considering changing the potential branding there at all?

Yes. I mean, the point we're trying to make here is it's no longer going to be a separate product, it's Chegg. And when I say Chegg, initially, it's Chegg Study and Chegg Study Pack, Obviously, AI will be integrated into everything, including skills and writing and math. But as we got deeper and as we really understood the depth and the quality and the differentiation that we have versus ChatGPT or barter anybody else, even in the education space in terms of what we could know and the value it could create for students that we made the decision to just make it all of Chegg. And as we think about how additional value gets created, the more value that we create for students, the more that we had, the more sticky that it becomes, the more things we can do for them. And as you watch the video -- the concept video that we put out, you'll see the other areas that we can address. We believe our pricing power, which has always been strong, will even get stronger. And over time, as we roll it out to everybody, we can imagine continuing to increase our ARPU and our yield. So it's a bigger move than the one we had before.

Good to hear. Thank you.

Our next question comes from Josh Baer with Morgan Stanley. Please proceed with your question.

Great. Thank you for the question. Wanted to ask one on competition. I think historically, you had a pretty favorable competitive landscape as the clear leader and maybe students used a few different solutions sort of in a complementary fashion. I'm just wondering how to think about the new competitive landscape, maybe ignoring ChatGPT for a moment? Like how does the future Chegg compared to other education-specific companies that are leveraging similar language models or open AI, APIs thinking about Conmigo or Quizlet or what might come from learning. Like any thoughts on the new competitive landscape as these education vendors use AI as well?

Yes. No, great question. And I think from our perspective, they don't -- we have such a big moat -- and the moat is only going to get stronger, because really, what we're learning is the speed of the computing, NVIDIA, thank goodness for everything that they have done, they really have sort of changed the game here. And then sort of the conversational nature that the analysis that ChatGPT has shown. But at the end of the day, the next set of value is being created by companies that already exists that already have very big brands, incredible loyalty are known for doing something and are able to leverage against their own datasets and their own customers. So nobody in the education space from our perspective has a more relevant or better dataset than we do. We're learning it's running the AI against the data that creates the differentiated experience not the AI itself on its own. Without the data, it's irrelevant, that's why these generalists can't do what we do. And so by keeping it proprietary, and as Andy said, building our own LLMs, we think our moat gets only bigger. The second thing is, it does take capital to invest and none of them really have it. I think we generate more free cash flow than most of them generate revenue in total. So the actual size of these companies is insignificant versus what Chegg has in terms of the business, the datasets and the capabilities to do. What you've seen is mostly just sort of chat bots versus what we're building, and I think you'll be able to compare our concept video as where we're going with what you see from them. I think you'll leave with similar perspective that we do, which is our moat only gets bigger and it was already big.

Thanks Dan, that's helpful. And then one quick one for Andy on CapEx. The lowest quarterly level since 2018, I think. Just wondering how much of that was lower engagement or Q&A from students versus leveraging Gen AI for your own content creation. Essentially, like is this level that we saw sustainable? Or how should we think about CapEx going forward? Thanks.

No, it was fairly seasonally low without a doubt. And as you know, we have -- it's probably the -- is the quarter where we have the fewest students actually in school, right? Some school, while we went -- it seems to have gone well for us. It's still a small period of time. So no, I wouldn't expect that. But we do expect CapEx efficiencies beyond this. We do believe that as we start implementing some of these -- some of our AI solutions that will have a benefit on CapEx as we move into call it into 2024 for the sake of argument at this point.

And so the reason for that, by the way, I think Andy explained it perfectly. The reason for that is the cost of content each particular piece of content should get less for us being able to leverage AI versus everything always being human. So the cost of content we'll actually be able to answer more questions than we've ever answered before at a lower rate on a per question basis and therefore, overall spending. So I think some of the other areas that we were investing in, like professor led content and things of those nature become much less important in this new world and the ability to leverage the data we have with ChatGPT and scale AI to create unique learning paths is really going to be the differentiator. And so that's the reason for what Andy was saying.

Great. Thanks.

Our next question comes from Ryan MacDonald with Needham & Company. Please proceed with your question.

Thanks for taking my questions. I'll have two separate ones in this. On the first one is just a clarification, Dan, in terms of your commentary around retention rates and sort of the new accounts sign-ups improving. Should we expect then second quarter subscriber counts to sort of trough here in 2Q and start to see improvements as we go into the back half of the year? And then separately, just curious to get your thoughts. There were some research out, I think, Stanford and Berkeley, in mid-July about sort of potentially the maybe declining or lack of efficacy from GPT4 on math problems. And just wondering if you're using that at all when you think about the additional rollout of Chegg of leaning into that math use case more and sort of the functionality you have with the math way to further differentiate yourself. Thanks.

Yes, good question. Let me take the second one first. One of the things about the college market that we've learned over the years is, it's very viral in nature in terms of students communicate to one another. And I think that's to our advantage. It was when we were building the company. And I think people got very excited about ChatGPT. And we are too by the way, and AI in terms of helping education. I mean if you're going to apply AI to anything, how great would it be if it could help people with all backgrounds, all walks of lives, have the ability to elevate themselves both with academic support and skills-based support. I think we'd all root for that, and that's what Chegg is building. And I think that's why a lot of people are rooting for our success that would surprise you in terms of the people that have been calling to be interested in and to help. So that message gets out on its own. And it takes one person getting the wrong information to destroy their semester, and so that happened pretty quickly. And I think we're known for our accuracy. So there's really nothing more to lead into it, because we're great and getting better. To be honest with you, I think I forgot the first question, would you mind re-asking it?

Yes, It was around the commentary in one of the previous Q&A around retention rates and that you're seeing sort of fewer cancellations and more subscribers that are up for renewal actually renewing. So as you think about sort of that dynamic, would we expect sort of second quarter subscriber count to be sort of the trough level here and you start to see growth in that metric as we get into the back half of this year?

I think, I'm sure Andy will add to this. From our perspective, I think we're going to stick to the one quarter at a time like we mentioned earlier, because we've had false promises before in terms of what we see, and -- but we -- if current trends continue, and we'll find out if current trends continue as the year goes on, then I think you'll see a very different and a very positive trend that we're all looking for and I'll expect to happen. We do expect to return to growth. From our perspective, we think the launch of the new products will be very valuable to that. We think the fact that the Stanford Research that you pointed to also acknowledge is that it isn't really good for what we do. And what we do is more valuable to students than what they do, which is writing papers, which we don't do. So we expect to be a growth company again. And -- but when that happens, we're just going to have to wait and see, but the trends are moving in the direction that we were hoping for.

Our next question comes from Alex Fuhrman with Craig-Hallum. Please proceed with your question.

Hi, guys. Thanks very much for taking my question. I was hoping you could talk a little bit about your business in emerging markets and specifically India, I think in the past, you said that maybe you hadn't handled the rollout of payments so well in India, can you give us a little bit of an update on the time line to all debit card acceptance in India and when you could start to see that market move the needle?

Yes. The latter one is hard for me to say, and I don't want to predict it yet, because I think in the world that we've lived in, in the last three or four years, making predictions has not really resulted in a good outcome for us or anybody else for that matter. But the first one I can answer very specifically, which is we are rolling out this month the new payment capability of the unified payment platform, which will allow for debit cards. So that's going to start to happen this month. And when we start seeing the impact of that we'll find out over the next few months, but it's going live this month. So that's very exciting for us.

Okay. That's really helpful. Thank you.

Yes. Appreciate the question.

Our next question comes from Brent Thill with Jefferies. Please proceed with your question.

Dan, just on the tone of demand you saw in Q2 and going into the later summer. Can you just bring us up to speed on that and ultimately, anything new in the playbook in the going back to school season here that you're seeing that may be working or contemplating? Thanks.

Yes. Look, to give you a sense to your question of the scale, investor winds down, school starts, we have summer school, and then the school semester really picks up. So the volume will double, then triple and then go up by about 50% again, all within a three and a half to four week period. And that period will start in about two weeks from now. And I think we'll peak around first 10 days of September in terms of what we're expecting in terms of when our real new account volume happens, and we're prepared for that. Now in terms of the go-to-market, we've always been really good at go-to-market. I think you've seen that, no one else has anything near the size that we do with paid subscribers. I mean in the education space, others have tried. No one else has succeeded. As I said, we generate more EBITDA than most companies do in revenue. So we're very good at it. But I think the channels that have been working. Obviously, TikTok as a channel over the last couple of years, that has become much more significant to us in terms of us being able to use clips and influencers who are using our content to teach people, and that's been really effective for us. So I would say that's the only new one on the horizon that we're leading into more than we have in the past.

And sorry, on the tone of demand, the demand continuing to strengthen week-over-week and did you see that going into the beginning of Q3 as well?

Yes. As I mentioned earlier, we began to see positive trends versus what we were expecting, really beginning of June and all through June and through the first month of this quarter, and that that gives us confidence to give the guidance that we have given. Things can change because as I say, the last two weeks of August are as big as the first three weeks of the first two to three weeks of August. So one week will be larger -- one day will be larger than a week in the peak season versus the trough season. So as Andy said to me this weekend, he said, you have to do something, you're not the best at, which is be patient. So I like the rest of you are just going to be patient as we go through it. But what we were hoping to see now, we are seeing, and what we were seeing in terms of the positive trends have continued.

Thanks Dan.

Our next question comes from Brian Peterson with Raymond James. Please proceed with your question.

Hi, this is [indiscernible] on for Brian. As companies inside are exploring new use cases with AI and in the live with your partnership with scale AI, are you considering M&A opportunities in this space? And more broadly, how are you thinking about our overall capital strategy as things stand today? Thanks.

Yes, I'll do the first part and let Andy do the second part on our capital strategy because I think people should see the work that they've done and recognize what he's done with the debt and buying back stock has been really effective for shareholders, which is why we -- I think people were confused that we weren't net cash positive to our debt, but we are. And that's really Andy and his team's great work. In terms of M&A, look, the thing that you want that companies are going to need is the data, and the database, and the customer list, we have all that. So it's -- for us to achieve what we're trying to achieve in this space right now doesn't feel like M&A is the answer because the moat that we have actually gets amplified as a result of AI, because what we can do with what we have will be so much more than anybody can do with what they have today or what they'll be able to do simply, because of the way our data structures have been built. And the over 10 years of understanding student behavior and content and behavior by each school and each class and each professor. These are things -- and the way the questions are crafted informed and even our ability to take them through images, which others can't do, we're slight years ahead of where other people are. So there really isn't an obvious need in the short-term for us to be able to do what we need to do. There's always -- there may always be something that will just accelerate it or speed it up. But right now, that's not a priority. I think Andy has better uses for the cash. So Andy?

Yes. So I mean, when you look at capital allocation, I mean in the ideal world, you'd use your capital to drive and grow the business. Those opportunities, as Dan had mentioned, really don't exist and don't make sense for us at this point in time. But what we have been doing over the last several quarters has been very opportunistic with respect to some of our securities buyback. Last quarter, for example, we retired debt at a sizable discount, almost $54 million. You saw that and we will continue to be, I'll call it, opportunistic and potentially active depending upon the value, but we've got a very strong balance sheet, like Dan said. We're driving significant free cash flow. So we've got the ability and confidence to make those moves should the opportunities exist.

Got it. Thank you.

Our next question comes from Jason Celino with KeyBanc Capital Markets. Please proceed with your question.

Great. Thanks. This is Devin on for Jason today. Thanks for taking our question. Nice beat on the revenue and EBITDA results. I just want to dive in on the subscriber number a little bit more. I think in the quarter, 4.8 million subscribers came in a little bit lighter than expected and also represents a decel there. But any additional color you can provide on what drove the decel in the quarter? Is it mainly on the softer new account side of the house? Or did you see higher expected churn in the quarter?

I'll let Andy give more detail, but we -- that's not our read on what we did. I mean we beat revenue and we beat EBITDA substantially, which is very hard to do on a $17 product in a lower season, particularly as the season end. So I don't know that we felt that we were light at all. And I think what we saw was better than what we expected, so I'm not sure what you're comparing it to, but Andy?

Yes, yes. So first thing is we don't guide to net the paying subscribers to be clear. But it's what I call subscription maps. We're in a period right now, the Q2 was a period where it was. And to some extent Q3, where it's dominated by renewals, which is from prior periods where we had steeper declines in new subscribers. What Dan is talking about is the fact that we're digging ourselves out of that pool, is improving relative to what we expected. And in fact, when you look at the quarter as a whole, both from a, like Dan said, from a retention standpoint and from a new subscriber standpoint, it was better than we expected. But once again, that's -- it was better.

Got it. Understood. And then just one more question for me. I think last quarter, you called out areas like Mexico with really high [indiscernible] interests, but low conversion. I'm just curious if there's any additional initiatives that went in, in the quarter to kind of drive that convert up? Thank you.

Yes. These are really interesting questions here, which is as we have adjusted the entire company to leverage what AI is capable of doing, the prioritization of the countries hasn't changed. So Mexico, the Philippines, Canada, Australia, the U.K., Turkey, places like that, remain the places where our initial efforts are focused on. What has changed to the better is what AI allows us to do is do these things better and faster, in some cases, simultaneously. So for example, AI in terms of being able to do instant translation is much better, for us to be able to add localized content faster and less expensive than we would have done before. So we're really just going to take advantage of what the new capabilities allow us to do on behalf of the studio. And we'll continue to do all the price testing that we've been doing. But during a slow season, it doesn't give off the right signal in terms of knowing really what's going to happen, because particularly, U.S. at high summers, most other places don't.

There are no further questions at this time. I would now like to turn the floor back over to Dan Rosensweig for closing comments.

Yes. Thank you, everybody. Thanks for your questions. We're really proud of the team and the results. As the information and the awareness of AI evolves and its capabilities evolves, we believe that history is a good guide here, which is the players that have really big brands, really unique and proprietary assets that are valuable, can leverage new technology and new capabilities to the advantage of their customers and therefore, the advantage of their business models. And I think you're beginning to see that, whether it's Adobe or Microsoft or Google or ServiceNow or any of these companies that have built really strong relationships and loyalty and expectations of quality with our audiences are going to be able to do things faster and we believe at a lower cost over time and differentiate themselves from their competitors. So we are excited about the momentum that we are seeing right now. We are going to take things one quarter at a time until we get comfortable that this is sustainable. We expect to return to growth. And as we do, you'll see the margins continue to improve. We have remained over 30% margins for the last several years, so we have the capital necessary, the business model to get it right and this proprietary data set that others wish to think you can access to that we are not getting it to them. We are going to use it and build our own elements to our advantage. And so the next six months are going to be super exciting for Chegg, and we're just head down working. So thank you all for joining the call. Enjoy the rest of your summer. Thanks.

This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.