Coursera, Inc. (COUR) Earnings Call Transcript & Summary

Good afternoon, everyone. My name is Ryan MacDonald. I lead Needham's EdTech research efforts here at the firm. Thank you all for joining us today for this virtual fireside chat between myself and Coursera's CEO, Greg Hart. We've got about 40 minutes to go through a wide variety of topics around the business and outlook and product strategy and all the good stuff. And so we'll be jumping into that. But if you do have questions for Greg, you can send me an e-mail Ryan -- [email protected], and we'll save a few minutes at the end to get those asked and answered. But with that, we'll jump right in. And Greg, thank you so much for joining us today.

Thank you very much, Ryan. It's a pleasure to be here. Excited for the conversation.

Absolutely, me as well. So since stepping into the role in this past February, why don't you sort of give investors sort of an overview of sort of what surprised you most about Coursera and the education industry more broadly? And are there certain aspects that give you reinforcement to your conviction and the opportunity when you joined the company?

Great question. Maybe I'll start just by talking a little bit about what attracted me to Coursera because I think it's relevant to your question. And part of it is that I saw a lot of -- even though I never worked in EdTech, I saw a lot of similarities between the experience that I had and the business challenges that Coursera was facing and the opportunity to do a better job of serving the customer through a lot of techniques and experience that I had exposure to. So my background is rooted in building and scaling technology-driven solutions for customer problems that serve tens, if not hundreds of millions of customers depending on the product and the prior point in my career. And so while I think higher ed is a new sector, I think a lot of the customer problems that we are trying to solve at Coursera have a lot of similarities to things I've done in the past in my time at Amazon. And so the approach that we took at Amazon was to try and always focus on the customer and famously work backwards from that to try and figure out a better way to deliver a service or a solution for that customer through technology. And I think that same principle really applies here at Coursera. And so our customers, our learners are seeking really flexible, trusted and job-relevant education. And they are doing that to drive growth in their career. 86% of the learners who come to Coursera are coming to advance their career. And I think we have an opportunity to help them do a better job of finding the right content to help them accomplish their career goals and then also helping them go through that experience in a way that delivers them a tangible skill they can take into the workforce. And then when I think about our other customer sets, both our university and industry content creator partners, they obviously are the ones who enable that learning for our end consumers and for our enterprise partners for that matter as well. So they share that same goal, and they're all using our platform to take their world-class content to market, if you will, and deliver personalized learning at a global scale. And I think that there's a huge opportunity to continue to leverage AI to transform that experience. So in the same way that AI is transforming jobs far faster than educational systems can adapt. It also provides an opportunity and a tool that we can use to evolve our learning experience and everything that we do to help our learners achieve those career outcomes through things like Coursera Coach, our AI-driven tutor, through Course Builder, which helps our university and industry partners create content much more quickly and in a way that's much more aligned with the outcomes that our learners want to achieve as well. And so I think there's a real opportunity to continue to take skills that I might have developed and applied in the past at Amazon and apply them within the educational technology space. And then finally, I just think that Coursera is incredibly well positioned to continue to transform the online education space. So we have been a pioneer in that space since our very beginning with Andrew and Daphne co-founding the company out of the MOOC movement that they were pioneers in themselves. And now we've seen that grow to 191 million registered learners around the world, fantastic content and AI-enabled platform. And then, of course, a very healthy balance sheet. All of those are incredible assets that I think we can use to continue to evolve our leadership position within the space.

Yes. That's well said. I mean, look, learning is lifelong. And the great thing about education is content is never static. As innovation comes, you need to require new content for that. And clearly, the business is in a really great position to sort of help foster that continued learning innovation. Maybe sticking on sort of, let's call it, management changes. The company obviously is currently going through a CFO change, and Mike Foley was announced last week as the interim CFO. Can you just talk about the decision to bring in an interim CFO from, I guess, you want to call outside the day-to-day operations of the company? And is Mike being considered for the permanent role? And maybe just provide a bit of an update on the search process or sort of the rationale around that.

Sure. So obviously, as you and your listeners all know, Ken stepped down as our CFO at the end of October after 5.5 years with Coursera very grateful for Ken's help as I got to know the company and the space. And with that transition, obviously, the CFO role is such a critical one for us as a public company. And I want to make sure that we are approaching that with the right mix of urgency and discipline. And so I want to make sure that we have the time to take a very thoughtful approach to that search to find the right candidate. In the meantime, obviously, we announced Mike Foley as our interim CFO. One of the reasons for that is, number one, Mike is a phenomenal leader. He has over 2 decades of finance and operational leadership in global tech and in media. He has a proven track record of scaling companies through rapid growth and transformation. And his prior roles include a really interesting mix of different finance and operational experience at various different companies that I think can be very helpful for us in this interim period. And so I believe that he can help us continue to make progress. Like we can't afford to and nor certainly would I settle for pausing progress in critical areas while we search for a full-time CFO. And so he can help us keep advancing the ball forward in the meantime while we continue with the search. That search remains underway. I am very focused on candidates with public company CFO experience and strong operational leadership. And as with the other roles that I brought into my leadership team this year, so Chief Product Officer, Chief Data Officer, Anthony Salcito, who we just announced as our enterprise GM, I am extremely interested in finding leaders that can round out and elevate the skill set of our entire executive team. And so that's the approach that we're taking. I'm confident that we'll find the right leader who's motivated by our mission, but also ready to help build on the momentum we've demonstrated so far this year.

That's great. That's great. Greg, in your tenure thus far as CEO, you've made a number of adjustments across the business. Can you talk about what areas you're emphasizing in terms of investment heading into '26? And how do you think about the right balance of sort of growth versus margin expansion for the business moving forward?

Well, first of all, I've been really pleased with the fact that we are starting to execute with both renewed focus and at a faster pace. And so I joined in February, as you mentioned, we provided our first guide for the full year 2026 following our Q1 results. And at that time, we were guiding on the revenue side to $720 million to $730 million, so 4% year-over-year growth at the midpoint. And in keeping with the annual EBITDA margin percentage guidance that the company has provided in years past, we provided an annual EBITDA margin of 7% on that call in April. And since then, we've had 2 -- so we grew in Q1 at 6% year-over-year, which was basically the same growth rate that we had grown at in Q4 of last year. And then in Q2, we grew at a 10% rate, and we increased our guide on the EBITDA side. We also increased our guide on the revenue side. And so did the same again in Q3. So grew 10% year-over-year in Q3. So 2 sequential quarters of 10% year-over-year growth. So nice to see that improve from the 6% that we had shown in Q4 and Q1. And then on the EBITDA margin side, we're now guiding to annual EBITDA margin of 8%. And then on the top line, revenue of $750 million to $754 million, so a growth rate of 8% to 9% versus the 4%. So we've expanded or increased both the year-over-year revenue growth rate and the EBITDA margin. And so I'm very, very pleased with those early signs of the impact of that renewed focus and a little bit faster pace of growth. I believe that the framework that the team had used historically is a great framework. I really like that approach of having an annual EBITDA margin guide because it allows you to make investments on a longer time horizon basis versus having to manage to a given quarter in the same way. And I think that drives the right longer-term outcomes for our product and for our business and therefore, for our shareholders as well. A few key growth areas that I've been very focused on to drive customer value. First and foremost, more rapid product innovation. We need to continue to accelerate our development cycles to leverage AI and data to improve the learner experience and continuously enhance our capabilities across all areas of the platform. Second, a faster, more agile content engine. So content is the fuel of our business. We need to keep expanding our catalog of branded career-focused content. We now have more than 12,000 courses that increased at 44% year-over-year, which is our fastest rate of increase in content in 5 years. And then also continue to leverage our own IP. So we continue to create courses to fill in gaps in our platform and in our catalog but also to use them as test bets for how to continue to improve the learning experience. And so continuing to double down on our content engine overall, so making it easier for our content partners to produce content, easier for us to ingest that, easier for us to optimize that in partnership with them and then doing all of that in a very data-driven way. And then finally, continuing to improve our go-to-market and our funnel. So improving search, discovery, merchandising while also optimizing our enterprise channels and continuing to recognize that on a regional basis, we have the ability to tailor content in a way that really works well for our learners, whether they're consumer learners or enterprise learners in a given region. And so I would say that generally, we're going to continue to focus on driving growth while continuing to deliver on our track record of improving profitability on an EBITDA margin basis annually year-over-year. Obviously, our Coursera produced is one of the levers that we have because there's favorable economics. And so there's a lot of benefits for us from a consumer and enterprise perspective as well as a margin benefit. And then investing in multiple different opportunities to create and deliver more value within our product that really drive better engagement and better -- therefore, better conversion and retention of our learners throughout the funnel. And so I would say that the first priority remains continuing to focus on driving growth because I believe that we are in the very early stages of a massive, massive opportunity at Coursera. There's never been a larger demand for learning than what AI is driving across the world right now. And so we're going to continue to focus on how do we invest to drive maximum long-term shareholder value through reigniting growth.

Yes. No, there's plenty of great initiatives there both on the product side and the go-to-market side. Maybe just a question, I guess, on strength of the macro. I mean, are you seeing right now within your base of consumer customers, in particular, like a strength in the consumer or sort of, let's call it, an ability of the end market to be able to consume the content from an affordability perspective? Are you seeing any pressures on the consumer as you think about how these growth drivers turn or translate to revenue?

Well, on the consumer side, I've been really pleased with the growth that we've shown in the consumer side, 2 really strong quarters of growth from consumer in Q2 and Q3. So we grew consumer at 10% year-over-year in Q2 and then at 13% year-over-year in Q3. And so really pleased with that acceleration versus Q1 and Q4, which were much, much slower from a growth perspective. I would say that, number one, we've done a couple of things to continue to evolve our approach to the consumer business and really be much more focused on those funnel economics. And so we've made a lot of good progress over the last couple of years on sort of top of funnel and really using marketing as a lever to drive growth for the business. I think we have an equal, if not larger opportunity to do the same on conversion retention and then ARPU. We've also evolved our pricing. And so as I think we talked about a little bit on the Q3 call, we rolled out our freemium model over the course of Q3. So all learners have access to the first module for free. And then if they want to continue with the course, they need to become a paid learner. And so that, I think, is a good evolution. And then we also, at the same time, made pricing changes on a global basis in many markets to reduce the price of our offering because -- and we talked a little bit about this on the Q3 call as well, because it really was just not priced in line with purchasing power in a lot of countries around the world. So we lowered the price of Coursera Plus monthly and annual in 60-odd different companies or countries rather around the world. And we've been really pleased with the response that we've seen to that overall. I think there's still a lot of work that we need to do to continue to improve the product experience. But I think there's an immense appetite from consumers. And you see that reflected in the fact that in Q3, we had 7.7 million new registered learners in the quarter, which is our highest number since, I think, Q2 or Q3 of 2020, the height of the pandemic. And so I definitely think there's a lot of appetite from consumers for online education and people are really recognizing the disruption that AI is causing and their need to really become sort of continuous lifelong learners.

Yes, makes sense. Yes. And maybe on that sense, so let's shift the conversation to AI a little bit. As -- from our work as we've kind of evaluated the market and sort of potentially this obviously major change that's being created by AI, there seems to be maybe 2 schools of thought on AI. In one scenario, enterprise organizations adopt AI, lay off staff based on productivity gains and that shrinks the overall workforce and it forces people to reskill themselves in an AI world. The second scenario, enterprise organizations, again, still adopt AI, but more so instead of layoffs, slow the pace of hiring, invest more in their workforce to make them more productive with AI in which the next scenario, enterprise organizations invest more on AI learning content themselves for their employees. I'm curious, what camp do you fall in, in terms of, I guess, in those 2 scenarios? What do you think is more likely? And which scenario in your view, benefits Coursera the most over the next several years?

Well, I mean, I think, first of all, like the history of human progress is sort of irrevocably linked to the history of technological progress. And every single technology shift over time has historically changed the roles that are required. It doesn't mean that there are fewer jobs than there were prior to that. It just means the jobs that are required are different. So you saw that like in the industrial revolution, you saw that more recently as computers became a widespread thing in the '80s across all kinds of businesses. It simply changed the types of roles and the types of skills, therefore, that are required. I think that is playing out now with AI at a faster pace than any of those prior revolutions happened. And the way that AI is really transforming roles is really unprecedented. I mean the World Economic Forum in their Future of Jobs report from earlier this year, they predicted that if the world were 100 people, if the global workforce were 100 people, 59 of them would need retraining by 2030. So that is a huge, huge percentage of the global population that needs to be retrained. And my personal opinion is that, that number is going to go up, not down over time just because of the pace of change that AI is creating. And so that creates huge challenges for all parts of the system, right? It's -- AI is transforming what skills you need for the roles that are out there in the workforce faster than the existing educational systems can adapt. Employers are struggling to find the right AI-ready talent and to upskill their existing workforces. And then learners are overwhelmed by the pace of change and unsure where to begin and how to go about gaining the knowledge they need to be successful. So I think that creates incredible opportunity for Coursera. And the real opportunity is rooted in delivering solutions that bridge that skill and educational gap. And I think that scale, which obviously we have, with the right technological platform, which we have, and then with the right ecosystem and the right content to reinvent how education can be delivered and how learning can be delivered. And I think that the mixture that we have with world-class content, like in a world where AI makes it easier than ever before to create content, the value of branded trusted content really elevates in my opinion. And so obviously, our relationships with some of the preeminent universities in the world, along with some of the preeminent industry partners who are creating content is a huge asset for us. And then we need to keep inventing and investing to invent on the product platform to continue to enable us to take advantage of AI as a tool to deliver far better learning experiences that are more engaging, more interactive and deliver better outcomes for those learners. And ideally, that lead to verified skills that they can demonstrate to the workforce that they have. And I think we are uniquely positioned to execute on all that. So to me, this current moment is an incredibly exciting time for our company. And I think we're in the very early stages of how this is playing out. And so those are a few of the thoughts I have. But my primary thought is we don't all yet fully understand or grock the extent of this transformation that we are going through. We are in such the early innings. I think this is going to be the most disruptive and most transformational technological change that the world has ever seen. Obviously, it builds on top of other large changes, mobile, the Internet, et cetera but it, I think, is going to eclipse all of those, both from a pace at which this happens and from an impact at which it happens. And I think we are looking to see how can we help people navigate through those changes.

Yes. That makes complete sense. One of our other analysts that I am here today or even earlier today, just was hosting a human capital management insights call with one of the largest survey collectors and it really speaks to across the enterprise, mid-market SMB that L&D remains a top investment priority for HR departments at global organizations. And I think what you just said there really speaks to why that is the case. So very much there's a huge opportunity here. Moving into sort of recent announcements. You recently announced a partnership with OpenAI. Can you just talk and provide a bit more details about the relationship? Is this an exclusive partnership for Coursera? Obviously, ChatGPT has quite the dominant lead amongst consumers relative to its peers at the moment. But I think that gap is going to continue to narrow over the next couple of years. And just broadly, how you think about the LLM channel as an overall driver of traffic to Coursera sort of moving forward?

Yes. We're incredibly excited about the partnership and what it can mean for Coursera. It is very much in its early days, obviously. So it's not exclusive. We are the sole online learning partner that they included in their initial launch of the 7 first skills or apps that they integrated with their ChatGPT app API. They reached out to us. And I think they did in part because they recognized that, number one, learning is one of the most common use cases on ChatGPT. I think it's the #4 most common topic that people are asking about. And two, I think they recognize the value of our content, and they see that in some of the citations that ChatGPT will present. And so then I would say that they wanted to find partners who could move quickly. And I think as soon as we started engaging with them, they were really impressed with the pace at which we moved. And I think also with the way that we provided them input and feedback on what they were trying to accomplish. And it's really interesting because it's the early days of this integration that we have with them. And I think the experience has a lot of ways in which it will continue to improve, and many of them are actually like exactly analogous to when I led Alexa. And so when we launched Alexa, we very quickly announced the Alexa Skills Kit, which was an API that enabled any third-party developer to create a skill for Alexa and app. And when we first launched those and rolled them out, they were relatively sort of on rails, and they were a bit clunky, to be honest. And you'd have to say, like Alexa ask Uber to get me a cab. Alexa ask Spotify to play X. And I would say that, that broadly is where this first instantiation with OpenAI is as well. It's still relatively an on-rails experience. And I think where it will become even more magical over time is when -- and I know that they will solve this, given how quickly they develop things, it becomes even more contextual and leverages ChatGPT's long-standing interaction with a user to really understand when they can recommend the right content. And obviously, we want that to continue to be Coursera. We continue to be the only online educational partner. But I expect that they'll keep doing that over time across all of their app ecosystem partners. I think -- we also -- we were founded obviously, by 2 AI experts, Andrew Ng and Daphne Koller were both AI professors at Stanford. And so I think that for us, because we've invested so much in AI from a product perspective from very quickly after the launch of ChatGPT with things like Coursera Coach, which I mentioned earlier, Course Builder, which is our AI-driven sort of instructional design tool that content creators can use to create better courses on Coursera much more readily, AI translations, AI dubbing, academic integrity features. We are going to keep investing on our ends to improve the learning experience when the learner comes through to Coursera. In the same way, we are engaged in ongoing conversations with OpenAI about how they and us together are going to keep evolving that app experience to make it even better and even smoother. So I'm very excited for where this can go for us.

And maybe just like kind of the obligatory question here is like because obviously, these large language model companies have wanted as much content and data as they can to train their models over time. What protections can you put in place to sort of ensure or mitigate the risk of sort of disintermediation from an OpenAI or some of these other LLMs out there. So they're not taking your proprietary content at the end of the day.

Yes. Well, I mean, through this specific integration, it's just a preview. So they don't have access to the entire course. The user can ask questions within ChatGPT about the course that gets surfaced but they can't actually consume the course. If they want to enroll in a course, they have to come through to Coursera and then they'll benefit from our AI features when they do that and the full learning experience. The preview video that's in line in ChatGPT is a short snippet of the full learning experience. And so -- and they cannot train their models on our content. None of the LLMs can. And so I'm not particularly worried about that. I do think that they are going to continue to want their experience to do a better and better job of helping to answer user queries generally but I don't think that they are trying to become a full-on education provider in the same way that Coursera is. I don't think that's what their primary focus is because I think if it were, they wouldn't have decided to reach out to us for this. So I would say that the value of the content that we provide being trusted and verified and from name brand institutions is absolutely a differentiator for us and will help us continue to be both relevant and also a leader in this space on a go-forward basis.

Makes sense. You also announced a different type of partnership with Anthropic recently. Can you just talk about sort of the value that bringing in another sort of LLM partner like Anthropic can have to your broader AI strategy and content strategy? And what should we expect in terms of sort of new content and learning experiences out of the partnership?

Well, yes, I mean, we're incredibly excited about the Anthropic partnership as well. One of the things that is really interesting about it is we see this content is coming on to the platform at a time when AI and learning about AI is the most in-demand skill in our history. So we are seeing 14 enrollments per minute in Gen AI-related content on Coursera. And that number -- that's the average year-to-date in 2025. That number has gone up as the year has progressed. So if we were having this conversation in the spring, it would have been like 11 or 12 enrollments per minute year-to-date, and now we're at 14. And so if you looked at like a trailing 30 days, that number would be even higher than 14. And a year ago, it was 8. 2 years ago, it was 1. So we are just seeing this unbelievable rocket ship of interest in Gen AI learning generally. And so against that backdrop, who better to have create content for the platform than one of the world's preeminent AI research labs. And so we're thrilled to have Anthropic be on the platform. There are 2 aspects to the content that created -- they've created. So one is a specialization. That's a more generalized specialization that has 3 courses in it. That specialization was created by Anthropic in concert with Advancing Women in technology, AWIT. And the goal of that is to really help learners of all experience levels build practical fluency for using AI in their daily work. And it's really meant at helping to level the playing field. And then we also offer 7 different modules in building with the Claude API. And so that is focused on teaching developers how to deploy Claude to build intelligent systems and streamline their workflows. So there are 2 sort of very different types of offerings. One is offered at people of all experience levels. The other is specifically targeted at developers who want to leverage Claude. I am really excited. Obviously, it's super early days for those things given that they just launched last week, but really excited to see how they perform and where that content relationship continues to go.

Awesome. Maybe the last one on AI is sort of your use of AI internally to improve the learning experience. You've done this in a number of ways, but I thought one interesting example was sort of your AI dubbed courses. You mentioned on the third quarter call that AI dubbed courses are going to surpass 1,000 offerings in 5 different languages by the end of this year. Can you just talk about what impact that's having on sort of course enrollments internationally? And how much room do you have for expansion into additional languages with these AI dubbed courses as you go into '26?

Yes. As background, we started years ago using AI before it was Gen AI, it was just sort of machine translation to translate the text of our courses into other languages and to provide subtitles for videos, which is a step in the right direction. And now we have more than 60% of our catalog is translated into 26 different languages. So more than half the world's population can now take courses in their native language, which is fantastic. But subtitling and text translation is not as effective as dubbing. And the beauty of AI dubbing is that the technology is now getting good enough that you don't have the feeling that you're like watching an old B movie that was dubbed. It actually -- the lip movements are synced to the sounds. The accents are realistic and real but they still have the voice of the instructor. And so it's really quite impressive. And so we have about 600 courses right now that have been dubbed into 5 different languages. And by the end of the year, we're going to have more than 1,000. We're going to keep expanding from there. And so right now, the languages are Spanish, French, German, Brazilian, Portuguese. And then we also have, I don't think quite all 1,000, but Indonesian as well, a number of courses. The impact is substantial in the sense that we see relative to dubbing. If you have the option to use dubbing, 91% of the people are picking dubbing over just translation. And then for translation generally, the stats are really positive. And so 85% of our learners report benefits to learning in their preferred language. None of these are really surprising when you think about them. We see 22% faster course completions and 59% of learners say that translations benefit their career. We haven't yet looked at the stats just because the number of courses is much smaller. So we don't have the stats on the dubbing specific pieces for those last 3 data points, but I would bet that it would actually be higher in each of those cases. The main -- I would say the main gate for -- because my goal is over time that everything that we have available on Coursera is available in every language, both text-translated/subtitled and AI dubbed. The main barrier to achieving that is on the text translation side, it's really just cost and time and time is less the issue than cost. And so it's just making sure that we see the right return on doing that. On the dubbing side, the main barrier there is actually whether or not it's actually sufficiently high quality because the -- for dubbing, it's more complicated than just a text translation. And so we're really making sure that we're trying to maintain the right quality bar there. We're actually in the middle of a big effort to dubb a number of -- to both translate and dub a number of courses into Arabic. And the big thing there is making sure that we have the right quality bar on that. And so that's our eventual goal is to get to everything available in every language but that's our longer-term big her audacious goal. And so we're just going to keep making progress against both of those along the way. And as I mentioned...

Go ahead. Sorry.

I was just going to say just quickly on around -- to me, like when I think about how do we make Coursera as relevant to as many learners in as many different countries as possible. So obviously, language is a big piece of it. But also the pricing, which I mentioned earlier, the geopricing, payment methods. So we're continuing to expand the payment methods that we accept in countries around the world. And then finally, our promotional capabilities, making sure that we're able to run promotions in a way that makes sense in any given market, just given what consumers might be used to. And so our goal is really to continue to make Coursera as accessible as possible for all of our learners around the world.

Makes sense. As we think about potential economic benefit, maybe let's just call it on the AI dubbed courses of the translation. Is it something that's typically more of an immediate benefit where you have -- you want to sort of expand with the AI dubbed courses in a new language, a new region and you sort of see the more immediate like tail or I guess, uplift in enrollments? Or is this more of a gradual, hey, NPS starts to improve from the existing learners, word of mouth gets out and then you start to see sort of more of the positive impacts?

I would say it's a mix of both. And so sometimes on the enterprise side of the business, like there can be newly translated content into a language that's relevant for a given enterprise, whether that's a government customer or a business can definitely provide an immediate improvement or ability to win a specific deal. And so some of our road map for translations is certainly driven by where we see demand from our enterprise customers or prospects. On the consumer learner side, I would say, broadly, it tends to be a longer-term benefit that we would see. For an individual learner, it might be immediate and instant. But broadly, I would say it tends to be a longer-term thing that plays out over time.

Yes, makes sense. As we shift to sort of now the Consumer segment, you talked about it earlier about the success you've had this year, the acceleration in the growth rate throughout the year. Obviously, ideally trying to sort of at least try to maintain, if not continue to improve upon that. Can you just talk about like where you think the major levers for continuing that positive trajectory as we head into '26? I imagine AI courses and expansion of that content is going to be a big driver. But you also mentioned localized pricing, better promotional spend as other factors there. Maybe just talk about what you think are the biggest drivers within the Consumer segment moving forward.

Sure. So number one, as I mentioned earlier, we are executing with sort of renewed pace and a growth focus. And so that was reflected in the Consumer revenue growth for Q3 and in the top of funnel on 7.7 million new registered learners in Q3. I think that -- both of those were driven by a combination of different factors, starting to better serve some of our international traffic. So the geo pricing was certainly a part of that. Obviously, the ongoing translation work is a part of that, the payment methods that I mentioned as well. Then I think also we have a more responsive revenue model on the consumer side of the business, where it is faster to demonstrate the benefits of some of the investments that we've made. Just by its nature, consumer is more responsive from a revenue perspective than enterprise would be because of revenue recognition and the pace of those cycles. We are seeing some early improvements in paid conversion, also a little bit on retention. But I think that those continue to improve both of those remains our biggest opportunity. And so we've made some really good progress, I would say, over the last 1.5 years, so certainly predating my time here on top of funnel. and on really driving healthy growth overall at the top of funnel. We haven't made as much progress at a faster rate on conversion and retention. And then as you make improvements to retention, that's obviously increasing your ARPU over time. And so I think we have a lot of opportunity there. We did mention the fact that on the Q3 call that Coursera Plus, our monthly and annual subscription now represents more than half of our consumer revenue. And so that certainly benefits lifetime value for each of our consumer learners on average. and forward-looking visibility into our revenue. And as we make improvements, not just on conversion, but in particular, on retention, that will just sort of supercharge some of that. And so there's a lot of work that's going on right now that will benefit both consumers and enterprise around skills and career outcomes. And so we have just launched in September for our enterprise business, Skills Tracks. We have -- obviously, that content is also available to our consumers. One of the things that we're going to be focused on for 2026 is how do we package that for our consumer learners in a way that they can wrap their head around and that puts them into the right learning path for their specific career objectives. My personal belief is, I think that the market opportunity that we have on the consumer side is going to continue to expand given the transitions that AI is driving and the need for different skills that we see in the workforce. And so we are going to continue to invest on the product side, certainly on our marketing side as well but increasingly on the product side and on our content initiatives to ensure that we are creating fantastic learning experiences that really drive better outcomes for our Consumer learners and help drive more durable growth for us. And then we'll keep experimenting with sort of different new ways to attract learners. So the Coursera integration with ChatGPT is a good example. Some of the things that we're doing on the marketing side with experimental marketing budgets and new channels is another example, and we're going to try and do all of that in a very data-driven way.

Awesome. Well, 40 minutes is cruising by faster than we expected. But before we let you go, I want to touch on enterprise just quickly. Obviously, that business segment has been a bit more in transition as you kind of come on post-COVID contracts in the Coursera for Business or government side. Campus has been obviously a much more attractive opportunity more recently. Can you just talk about how your -- what your approach is to sort of reaccelerating growth within that Enterprise segment as we kind of move forward? And within those 3 buckets where you see the most opportunity in the near term to really drive growth?

Sure. So first of all, we're not pleased with where our NRR has been over the last few quarters. We were at 89% in Q3, down from 93% in Q2, which had been a bit of an uptick in Q2 versus Q1. We've seen mixed trends between the different verticals. You mentioned Coursera for Government, C for G has been particularly challenged. We do see really promising momentum on Coursera for Campus. That was true both in North America and in APAC, in particular. And then C for B, which is our largest of those 3 verticals is -- has been more measured, I would say, and more muted. One of the first ways that we're focused on turning that around is bringing in the right leadership. And so super excited to have Anthony on board. He's now been here for, I don't know, 1.5 months as our enterprise GM. He joins us after 25-odd years in the education space, 20 of those at Microsoft, leading their worldwide education efforts. So deep familiarity with the space. He's a very product-focused leader. And so he's partnering very closely with Patrick on how we continue to evolve all of our enterprise offering. And I expect that he'll bring both renewed operating discipline to the approach that we take with enterprise, sort of permeating throughout all levels of the organization as well as a better focus on how we make sure that we have all the right integrations with our enterprise customers, whether that's with LMSs or LXPs or with enterprises themselves. And then on the product side, one of the things that we started and launched before Anthony joined but I think we'll continue to invest in and build out over time is Skills Tracks. And so there is a clear focus that we hear all the time from our enterprise customers on workforce upskilling and reskilling at scale as technology changes in the job market and therefore, the needs that they have within their organization. Skills Tracks are curated sets of content. We launched our first 4 in September. Those 4 are data, IT, software and product development and then finally, Gen AI. And those are more curated groups of content where in each case, we've leveraged our career graph, which has over 2 million data points that brings in third-party data from organizations like Lightcast and combines that with global skills frameworks so that we have a map of all the occupations in the world, the roles within each of those occupations, the levels of each of those roles the skills that are required for each level and then the courses on Coursera that map to those skills. And so with Skills Tracks, what we're doing is we're taking curated sets of content, and we are giving any employer a map for if you have this role and you want to provide this type of upskilling, here are the courses that do that. And we're building out verified assessments and credentials at the end of that so that the enterprise can be sure that, that employee has learned and mastered that skill, not just that they took the course and got a checkmark for taking the course, but they actually have demonstrated mastery through a verified assessment. And so that verified assessment work stream is in pilot right now for the data Skills Track. We're going to expand it across all the Skills Tracks by the end of Q1. We're going to keep adding more Skills Tracks and more verified assessments as we go. And I think that is going to be an area that we're going to continue to invest in perpetuity effectively because I think it's just highly aligned with what the world clearly needs right now and what we're hearing from our -- both enterprise customers, existing customers and from our prospects as well. So it's still really early days for Skills Tracks but very excited for the potential that, that can create for us over time.

Yes, it's a really attractive opportunity, especially as you think about these enterprise organizations going from developing, let's call it, boardroom level AI strategy to actually then going out and trying to execute on that strategy. A big component of it within your workforce is identifying what skills you have and don't have and then sort of being able to then close some of those skills gaps so then you can actually execute on the strategy effectively. It sounds like Skills Track is really going to help to address that issue.

I think so. That's the plan. Obviously, that will play out over time. And so our near-term priority is just really improving our execution within the enterprise space, which obviously Anthony will be incredibly helpful in driving. And then he'll also influence certainly all of that road map as we continue to expand the Skills Tracks.

Awesome. All right. Well, we're up on time. I think I've just missed 4 calls from my wife tell me I got to go home and make a turkey and clean the house. So Greg, we're going to let you go. And thank you, everyone, for joining us today. Hope everyone has a great holiday rest of the week, and we'll leave it there. Thanks, Greg.

Thank you, Ryan. Have a fantastic Thanksgiving to everyone. Take care.

Bye.