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

All right. My name is Josh Baer, software analyst here at Morgan Stanley, and we have Ken Hahn, CFO of Coursera. For research disclosures, please see the website, Morgan Stanley research disclosures or ask for your -- ask your sales representative for those disclosures, something like that. So thanks for joining us, Ken. Really excited about this session.

Wanted to start off at high level and like how you're thinking about 2024. Guidance, and you gave some specific guidance for each of the 3 segments, and that's kind of how we'll structure our section. But it's another year of 20% growth in Consumer, and that's, I think, 5 straight years of 20%-plus growth and then around 10% growth for Degrees and Enterprise. Consumer growth comes off of outperformance. Degrees and Enterprise comes off of more kind of in-line performance. So hoping you could review some of the broad performance across these 3 segments that's been driving this divergence here.

Sure. Absolutely. So the Consumer business has continued to go as -- exactly as you referenced. We've been really pleased with that. I'd say since the IPO, which we're coming up on the 3-year mark, Consumer I'd say, has exceeded our expectations as well where -- and even from the beginning, it's the same effects. We've had just great success with the specializations that we offer, which, for those of you who are unaware, are these essentially college-like courses for a semester, where it might be 3 courses to make up a specialization where you take somebody with no required education and you teach them a skill from scratch to enable them to have a, say, $60,000- to $80,000-a-year job depending where you are, which part of the country. And those have proved wildly successful. Interestingly, and we're coming up on the year, we did our first survey of the completers. And 1/3 of them -- it was just a survey. We didn't actually go look at the specific data to validate, but people self-report pretty well. 1/3 of them said that a significant career outcome because of that specialization, which given it costs you a couple of hundred dollars is pretty amazing to have that kind of payback. And so that was either a raise or a new job or -- so anyway, the story around Consumer has been the specializations and essentially the rate of change that COVID caused, right? As things moved on, it created a lot more change, business moved faster, people had to adapt. It wasn't so much the, you're stuck at home and how do you fill your time. I'm sure there was an element of that, too. But the reason it's persisted, we believe, is because at the end of the day, change is what drives the need for education. Hopefully, we see the same thing with AI adoption coming up, and it will probably cause a lot of displacement, a lot of change. And so we'll see how that plays into this year. But again, the story of Consumer has been the story. And then the other thing we've done there is we introduced, it's now been older, a couple of years ago the subscription product, so Coursera Plus, which you can think of it as a pricing and packaging move, which typically are pretty good. It provides more value to the consumer. And especially when people are looking at different specializations and thinking -- trying to choose between them, giving them the choice to be able to figure that out is something that's worked out really well for the learner and well for us from a revenue standpoint because it's more predictable and you tend to get longer paying periods from the consumer. So that's Consumer, continues to plug on. Enterprise. Enterprise, we divide into 3 verticals. The largest and when we went public, really the primary one was Coursera for Business. So we have Coursera for Business, which is roughly half the revenue we discussed on the last call in Enterprise. Coursera for Campus and Coursera for Government, which split the other half of it roughly equal. We don't disclose it specifically, but we've given that general guidance. C for B has been more muted. It's been subject to a lot of budget pressure throughout the last year or 2, which I don't think really abated, and Q4 was still a tough time economically in businesses. And so that's -- it's sold into L&D departments primarily, not always, and those departments have not had -- there's been budget cuts a lot of times, and so that's been a harder market for us. It's pretty competitive. We have some good players in that space, which is great because it supports a robust end user market for a while. Coursera for Government and Coursera for Campus is where we've grown there. Coursera for Government looks the same as Coursera for Business. It's a different, obviously, procurement cycle, but we play particularly well there based on the attributes of the product, which for us is really good brand. That's where we started and it's a result of the way our content engine works and kind of the job-ready specialization. So those two things are primary buying considerations around Coursera for Government. So we've had amazing results there over time, and that continues to do well. It's a little lumpier. They're bigger deals across the enterprise system. It's sold like enterprise software. They're multiyear contracts, 6 figures, sometimes 7 figures. Particularly in government, they're bigger. It's the State of California, it's New York, it's whole countries. So we continue to do well just because it's what they're looking for, both, again, from a brand and content standpoint as well as a job-ready or a job utilization standpoint. And then Coursera for Campus, where we compete somewhat uniquely. It's a bit of an adjunct and will become more important over time with our Degrees business, but that's selling to campuses the same product. It's access for students and for the faculty to use the whole platform. By the way, you can see all that content if you -- for free if you go on our website if you want a certificate with that or you want to offer it for credit, which is the important use case around Degrees -- I'm sorry, around Coursera for Campus, which again, it plays in Degrees with a lot of the same partners. That -- when it's for credit, the utilization is kind of off the charts. And if you step back and think about that, what that is, it's students being able to do these certificates for credit with part of their degree, which they can then put on their resume that shows kind of a job relevant aspect. And then the schools like that as well because they're being forced to be more outcomes-focused, right? And so when it's for students for credit in C for C, it does very, very well. And that's a relatively new product just a couple of years ago. I don't know, Josh, if you remember how we used to talk about it, but we'd say, "Hey, it's substantial." It's in the millions of dollars now and it's growing well, but we're still sorting product market fit because we wanted to be completely transparent about it. We don't say that anymore. We've figured out the fit, and we think it's a very strong part of our business where we compete almost uniquely. It's not where other people go. So that's the Enterprise segment. And then Degrees. Degrees is more a matter of still sorting product market fit. Things move slowly there. We're making real progress. There's a bit of a transition in that business, and it's the smallest part of the business by far. It's 50 of the 750, and it's been relatively stagnant the last couple of years. It is -- it's our largest market of the markets. And so we continue to talk about it even though it's a little bit painful with slower progress there. But ultimately, that's enabling degrees online and the use of our content for those degrees. And the most important use case around that is when -- on the high-value degrees, it's bang for the buck for the consumer essentially. But it's offering those degrees in a fashion that's more scalable. UC Boulder is a great example. We did our initial cohorts there. We're having good success. And that's a degree that costs less than $20,000. It's $14,750 for a 2-year degree on the data side versus it's $60,000 anywhere else. And so we work with them to lower that price and offer it on a bigger scale. That is the model we're moving to. And so there's a bit of a transition period where we have the older, larger because the tuition is much higher. And they're great clients. We're not moving away from them by any chance -- by any means. But the newer ones are slower to come on, and so we had slower growth there. Again, we think there's nothing that's changed from our outlook around the market being massive and we think the strategy is on. But implementing that and replicating it is the next focus.

Okay. Great. Great overview, and I'll have a lot of follow-ups across all those different businesses. But just thinking about the -- like back to the relative growth rates, like we're seeing another year of increasing mix of Consumer. You have these long-term targets out there. Like do we need to see a mix shift back towards Enterprise and maybe Enterprise and Degree at some point? And like what do you need to do to get there?

Yes. For the longer-term models the -- and to step back one step, there's some structural margin considerations. It's just based on our rev share for content. The Consumer business, which is by far our largest business, is contractually roughly 50 -- the standard contracts are 50% rev share. So we share half of our revenue, of course, with our content partners. The Enterprise business is a 70% margin business. We get paid more. Again, it's just contractual because we service these Enterprise. We have an Enterprise sales force. And so that's how it's working. And then Degree is 100% margin because there's no content involved. It's the customer's degree, that school's degree and it's a service charge essentially based on a percentage of tuition. The student is their customer. They are our customer. So yes, to get to -- and we're talking about EBITDA margins over the longer term, we do need a higher mix of gross margin. And since we've gone public, I'd say, in the long-term buildout, there's -- we've had more success relatively on the Consumer side than we have on the Enterprise or Degree side and so it's slowed that progress. We continue, of course, though, to move. And we announced last quarter plans for the year, our projection for the year to be 400 basis points EBITDA positive and for free cash flow to be in line with that number. And so we've made steady progress, which is our commitment when we went public that the expectation should be we're not maximizing for profit nor growth. We're going to grow to the degree we can, and the commitment is to improve our EBITDA margin every year. How much will depend on our growth prospects as we do planning. But I'm sorry to get in a now roundabout way. So that's the mix, and that's the context, of course, of your question. That's why you asked it. But we do need a bigger mix over time of Enterprise and Degrees. And we do think the opportunity in Degrees is -- it's our largest opportunity, frankly, and we do expect larger growth over time there.

Okay. Great. Couple of questions on Consumer and the really strong guidance. You mentioned the Consumer subscription having maybe a little bit more visibility. Like hoping you could unpack a little bit what gives you confidence in such a strong guidance on the Consumer segment and maybe incorporate how big is Consumer subscription? How much of that revenue is coming from newly launched programs versus some existing programs that are more in the bag?

So in any given quarter, the majority of the Consumer revenue is existing content because you can only introduce content so quickly. Where you start to see real contribution from content is content introduced a couple of quarters before because it ramps for a little bit while, and it's harder to measure it. So the vast majority is existing content for the year, but certainly for any particular quarter. And then the -- a lot of it's about the performance on the marketing front and top of funnel and just the broader consumer demand. So that's the piece that's less predictable. But the piece that's fairly predictable is the content portion of it.

Maybe the marketing piece is an interesting place to go. We've seen such durability of growth, but sales and marketing expenses, I think it's been flat or down over the last year or so. Like how is that dynamic possible? Is it that marketing spend is becoming more efficient? Or are you reallocating from other places?

It's a little bit of both is the answer. Part of it is we've become a lot more efficient. But we've also intentionally expanding our marketing efforts with better capability. And so this is paid marketing, right? It's clicks. And where we can -- we won't do it unless there's a good return. There's real discipline within the company because, otherwise, if you do that, it's -- the growth isn't real. It goes away as soon you stop spending irrationally. And so -- but we've built that capability over time to become a lot more sophisticated. And I think there's a lot more room to go, but that's been particularly helpful, but it's been more efficient. So to -- on one hand, we're becoming more efficient, so we're spending less for the same result. At the same time, we're also expanding that, which adds a little bit to the cost. But it is -- and then just some overall scaling just because we're becoming bigger.

Perfect. Let's shift to Enterprise. And I think that's where the -- talking about competition is maybe most interesting. So I guess if you could outline the current state of the competitive environment with the focus on Enterprise, probably Coursera for Business but also the other segments within Enterprise, who are you running into? Where do you win? Where might a company select another vendor?

Sure. So we -- I'll start just to remove them, make it easier. C for C, there's essentially no competition. Those are unique relationships, and our offering's unique. Government, we'll compete against everybody in the space, but we tend to do very well in those. Again, the focus tends to be branded content and employability. So it's the usual suspects with C and C for B, but we just tend to do particularly well. On the C for B side, there's LinkedIn, there's Udemy. There are some other players. They're the primary competitors. LinkedIn, we don't see so much. It would just like -- it's very different from a pricing standpoint. Usually, the way I think about it is if we're in the same deal as LinkedIn, one of us shouldn't be there. It's the wrong spec. Udemy sits between the two, I would argue. And they're really strong competitors. They have good product. It's a well-run company, which I think we can use more of in edtech. So we are supportive, I guess, is what I would say. But that's the real competition. A lot of it's -- and it depends region by region. But that's the -- yes, that's how I'd summarize it.

Got it. And could you talk a little bit about differentiation in business definitely...

Yes. A lot of it is the approach. So where we don't do as well, which is the bigger part of the market, is L&D because they're very price-focused, right? And their budgets are somewhat discretionary, and there's been a lot of pressure, and we've talked about this on the calls when things first slowing down broadly economically going back 1.5 years ago. So a lot of it, it's budget. It's a bit of a discretionary budget. When companies are pulling back tightly, they often deprioritize training. You can't do that for long. It's not helpful, but people do that in the short term. Companies do that in the short term. So that's one of the areas where there's been some headwind. The competitive -- where it's maybe we do better competitively, it's when it's deeper learning. It's not the quick 5 minute. There are certain competitors out there who will talk about no one has more than 5 minutes to learn in a day. Well, you're not learning real skills if you can only take 5 minutes. So when things are deeper around data science, around engineering, where you're taking real classes as we talked about, the specializations, which is like a college course, that doesn't get done in 5 minutes, right, in forever. And so those where there's a business outcome and the requirement to have that learning, we tend to do better. And it's a function, I think, of both our emphasis as well as the content engine. So some of the content engine's not necessarily better or worse out there. Some will take essentially user-generated content or subject matter expert content and bubble it up to the top, the best thing. So they tend to be able to respond very rapidly with shorter courses, shorter content, which plays very well in the L&D department. On the longer form that, our form, where we go and we sign up partners and curate it, it takes longer to have new topics, but it's a deeper learning. And so where there's especially budgets on the operating side, where people need real skills, we tend to compete well there.

Got it. And with that deeper learning, like the demand for your type of deeper learning in mind, does that lend well to any specific verticals or departments within a company? Or is it just about the company culture?

It can be a little bit of both, but the departments tend to be the techy departments. Where we're good on the content side is data and -- a lot of it hails from our background and the initial schools we signed up and the initial focus we had. We were founded by 2 comp sci professors at Stanford. And so it explains a lot of our content. It also happens to be content that delivers a lot of value, right? And so it's not accidental that those have continued. We've added on since then, of course. There's been more of an emphasis in health care, but like the core tech is pretty important to us. That's where we do well. So we do well with those departments.

Got it. Most of our conversation in the Enterprise so far has been around content, but there's also a platform to think about, like broader set of offerings that can be more strategic, in my view, to your company. So like how often are you brought in for your content versus the platform? And are you seeing any changes in that mix?

It's often a content -- it's a great question. It's often a content purchase. But with that content, it's hard to actually tell the reason. They are buying a platform. And increasingly, that's becoming part of the conversation. How does AI work? Can I introduce my own courses? So we're making a lot of that easier. Course Builder we've announced, which lets companies build their own courses. That's in the large corporate environment, what's of great interest to many of these L&D departments who want to put the branding and want to put the specialization relevant to their particular company on the platform. So that's something interestingly with AI that we are promoting more because we have a product that supports it. The secondary piece is Coach, which is part of the platform that's useful on the Consumer side as well. But it's essentially -- it's AI generated based on the content of that course, which can, on the course side, can help generate questions on Course Builder but on the Coach side can help you get through various learnings, essentially increase the value of the content of the learnings. And when you do that, there's more value to split. So it's an interesting transition period where it used to be content, content, content. There'd be a question of depth of content versus I want new content versus last year. Now there are increasing conversations around platform and capability.

Yes. Let's stay on that, platform, skills, AI. Any good customer examples where a customer is really using Coursera in a strategic way to skill or reskill workforces aligned to digital transformation or AI initiatives? Any context to kind of talk about that future of being like the key tool that companies are using to drive their skilling initiatives?

It's -- what I'd say is it's more, there are examples, but it's more the AI piece particularly is more future-based. And again, it's a little bit based on the nature of the content engine is what we're developing, is we're developing role-specific in a deeper form so that can be driven through -- down throughout a company and where -- as opposed to just the shorter-form course. So we don't have as much available today. We're in the middle of sales processes now where that is a very important part of what we're selling. And we're well along the way, so we can show people early on what we're doing. So I'd say a little bit more -- we'll see more in the coming quarters, the relatively near term in this year.

Got it. And to kind of round off this conversation on Enterprise, I think we all want to see stronger growth and stronger net retention rate. Besides macro and some of the pressures on budgets, what else needs to occur to accelerate growth and get to where you want to be in that segment?

Yes. I do think part of it is the platform changes we're having is going to be pretty helpful just from a stickiness standpoint because learners will -- at the end of the day, if you can teach people and they're real skills, there's value there. So I think the platform is a big portion of that. There is a secondary piece, which is translations, which is also being product enabled around AI. So our capabilities are there, and the content offerings in different languages are through the roof. So we've gone from $10,000 a course if you have a big C for G deal that would -- we used to be able to agree with 7 figures or if they're well into the 7 figures, we do custom essentially. A translation of courses into a different language for a government makes sense. With AI, we're able to do that broadly, which is helpful on the Consumer front, but particularly on the Enterprise front, where you have these global platforms. It's really important to be able to have it in multiple different languages. So a lot of it is platform-oriented.

Great. And translations is an interesting topic, maybe just to hit on international for a second. Being able to go broadly to the rest of the world with the full catalog of translated courses, like how has that changed your go-to-market internationally? And like what should we expect to see? What kind of conversations are you having?

It's -- so it's earlier on, although we have all the product now, and we're seeing real pick -- I think we're going to see greater growth internationally than we are domestically because of that, because we're enabling it. There's a secondary piece that enables it, though, on the Consumer side, particularly on the payments piece. We're exporting -- we're expanding our payments capability. We didn't -- we have a big implementation right now where we're starting to see some positive effect, but really the enablement of consumers to pay in this international market. So there's some enablement there as well.

Got it. Spend a couple of minutes on Degrees.

Yes.

As we said, relatively stagnant over the last few years, still a massive market and an important piece, thinking about longer term. Maybe if you could review your current strategy around Degrees and how that's evolved over the last couple of years?

Sure. So where we started with Degrees, go back a couple of years ago, we were selling -- it was -- and we've been in the business for a while. It was a decent -- again, it hasn't grown much. So it was a decent-sized business then. And -- but they were -- it was -- a lot of the customers span to pre-pandemic, and they were more expensive degrees, very good quality. And relatively innovative schools because they were choosing to go online with these important master's degrees and they're well-ranked programs, right? I won't go into the details of them, but good programs. And -- but they were based on -- which is how the degrees industry used to work, they were based on these much more expensive tuition rates. Where we've shifted to pretty hard over the last year or 2 is that, that model doesn't work great for most people. Given the types of degrees we had, it still worked. It was cost effective. And the idea was you needed to enable working adults, because most of these are master's programs, to not have to live in region to take a course. So you could expand your reach, both monetarily and just from an education standpoint. Where we've shifted and pushed pretty hard is really fixing the value and the ease for the consumer. And so there's been -- and traditionally, the topic, the subject matter were more technology-oriented. It continues to be so because there tends to be the best bang for the buck there. Those are -- those degrees tend to increase earnings power. But the secondary piece to work on was the pricing. And so we've worked with a number of the partners, and this has all evolved over the course of the past couple of years where there's increased focus on student debt and the general cost of college, where I think the consumer is somewhat dislocated. I think we believe there's been a change in people's belief that they need to go get a degree. In fact, we're even seeing changes in degree requirements for a lot of jobs where companies are saying, "You don't really need a degree to do this job." And so the unit of learning has changed. And at the same time, the degrees, we think, that are more viable need to be offered at a better cost. There's a certain level of university where people will pay $70,000 a year, and it makes sense economically. There's a whole swath below that, that are still -- pricing is still in that range, and it doesn't make sense. So we think there's going to be a pretty big change there. And so enabling, making it easier for the consumer, changing the pricing, which is what we've worked on in the current example that's up live at UC Boulder, and that's worked very well from a price standpoint. At the same time, we're working to make admissions easier and credit for prior learning relevant. So a lot of it is about increasing the funnel so that you can provide students for these courses. And that's what's hard for the schools to do as well. That's where the Consumer part of the business and the specialization works. But the process of getting those accredited and getting the schools is a long process. We're making progress. There's nothing that says it doesn't work. In fact, it's working across the board, but there are slow processes that don't happen frequently. And so that's what we've been working on. And we're making progress again. It's really rate of adoption, not whether we adopt is the view. And we need to replicate and show a couple more examples so that we know for sure, and then everybody believes there's not something special about these particular courses. But again, there's nothing yet that disproves, and we don't think there will be.

Got it. And what you're describing is the pathway degree?

The pathway degree, yes.

Maybe you could provide just a little -- like a more specific example of how that works?

Yes. So a pathway degree is there's entry levels through existing content. So if you were to take one of the specializations, we would work with the school. The way accreditation works is if you have ACE-recommended courses. It works across everybody, but then the schools have to adopt it. So most of our specializations are ACE-approved or ACE-recommended, and then you work with it. So if you have students coming from ACE-recommended course or specialization, which is a broad swath of students for us. It's a lot of volume. And again, this is being able to try to provide volume. And the school will offer credit, sometimes including admissions capability based on the fact that you have passed a couple of these courses. A lot of the -- if you look historically, a lot of admissions and programs schools have intentionally made very difficult as a sign of quality, right? Which is not how most business is done. We'll make it really hard to do business with us so that you'll perceive it as high quality, whereas with the schools who agree with us that this is the right way to do it and many of them are, they'll say, "How about instead you take a couple of courses that are part of the program to show you can pass those to show you can be successful in the program?" Because that's the end state. And so a Pathways program is offered through a pathway, right, one is Pathways with a lot of consumers. It helps you in the admissions process that's not required, but is very helpful and is into a program that makes sense economically. That's the -- those are the core components. And so really a product that makes a lot more sense than many degrees do today for students.

Excellent. So I have several more questions on financials. In a minute, I'll see if there's any questions in the audience. First, I want to talk about something that came up, I believe, just on the last earnings call around a different way to think about content, different content strategy, acquiring certain content assets. Could you provide a little bit of context or what that means and what the economic implications will be?

Sure. And we provided that context around free cash flow discussion forecast for the year, in which we said we thought we'd spend up to about $20 million on content, which is something historically we hadn't done lots of. I think we spent $5 million the year before. And it's because we're seeing more opportunities as the specializations are recognized as a good thing. And what we'll find, we do them with brands who we're familiar with or important. And we're just trying to get the content on platform, and we'll work with a group within the company. And they don't have budget because a lot of them don't have budget to create content. It's not core to what they do. And so instead of that being a hurdle, we'll enable it. Now for that, we'll usually ask for a different revenue share that's a little bit better than the standard. So we get paid back essentially on our investment and usually ask for it to be proprietary although they're -- we don't require things to be proprietary in our core platform and very few dual list. It's maybe a couple percent at best. But we are seeing increasing opportunity especially as the ability to create content becomes cheaper, right? And so we're doing that by AI-enabling Course Builder that we can use internally and work with our new content providers. And so it's more cost effective. So we're happy to pay for it, and we do a full financial analysis on each one. It's not some broad budget program where people go out and spend it. It's looking at the specific expectations around each of those grants, if you want to call them that, but each of those investments of capital we make in the content. So that is a, at scale, relatively new thing. And in fact, we changed our definition of free cash flow to include that, which worked negatively, worked against us, to be very clear. But we wanted to make sure it was a good and thorough measure, and it belonged as part of free cash flow because we're generating revenue from it.

Right. More favorable revenue share, higher gross margins down the road. I guess beyond acquiring what you've just been describing, are there other opportunities that you see for lowering the cost of content? Just as -- thinking about some of these revenue streams as you scale, as your registered learner base gets larger, like is there a potential to deleverage on the content line?

Yes, this is going to be an important couple of years in product development, which essentially just gives us more leverage around all of our content. We talked before about the translations, but the fact that we can translate as cheaply as we can expands the market for the same content. And we're using AI to do that. That will become more sophisticated over time. Course Builder itself and being able to break out the components, both for companies and for individuals and to take pieces of different -- and create custom courses using bits of content is one of the things we're also working on right now. Again, the concept is you just get more revenue, you get more leverage from that content. The Coach, which just helps people get through and learn better, adds value because there's more value there and people will persist longer. And it creates more value for companies as well, even on the Enterprise side. So there's a broad range. I think we're going to get more leverage out of product over the coming couple of years like we haven't seen before. And AI provides a lot of that opportunity.

That's great. So lots of areas of efficiency. And just to ask more directly, I guess, on -- as far as the revenue share with university partners like as -- if your Enterprise business triples, like is there a potential to decrease that revenue share and increase gross margins with existing partners?

Sure. So we haven't addressed any of our -- where we've changed the revenue shares when we're contributing economically. So no, we haven't gone back to any of our partners to try to renegotiate.

Okay. Got it. One more for me now and then questions from you. Very -- a question that comes up a lot from investors, just looking at your balance sheet, ton of cash, you have no debt. And you're basically going out and posting really strong results and like we see where the stock is.

And we're cash flow positive and -- right?

Right. And so is there -- question's obviously on capital allocation, capital return. Is there -- but maybe starting with buybacks. Can you talk a little bit about buyback strategy? Is there potential to do more sooner?

Sure. So we did do a buyback. We didn't buy anything back in the previous quarter. It's price-dependent on the trigger. What we have bought back was at an average of $12.11. So it was -- versus today, it's been pretty good. The -- we thought that was the right thing at the time. Part of it is we had a larger-than-typical market dilution through some grants we had the previous year, which we understood when we did it. But a lot of companies also had issues during the year, and they didn't correct it. But we thought it was important for people to understand that the dilution around those equity programs matter to us. And we did it for all the right reasons to retain the right people to contribute to the company over time. But we wanted to make the statement that we understand there's a market norm here, and we didn't want to go above the market norm. So it was put in place to remove that excess dilution as we thought about it. As it relates to the cash balance today and what we can do, the general belief is that we're early in a very large market, and that optionality is particularly valuable right now. I don't expect that we're going to change the model, as we've just talked about. We've been increasing -- we're cash flow positive, we've been increasing the EBITDA margin over time. We're now positive. I don't see us going backwards and using that to fund losses in the future. That's against what we've stated numerous times. Never say never, but that's not -- that is not the plan. But the -- on the M&A front, I think there's opportunity likely over the next couple of years as this market changes, particularly around AI. And when there's a lot of changes, when there's more variability and option is more valuable is the way we think about it. But it's not -- that's not a religious comment. If there's not the opportunity or if that changes over time, we'd absolutely use it to buy. We've already shown that we're happy to do a buyback. So we'll continue to monitor what we think the opportunity is there to expand the company and grow the top line and add appreciably to what we're doing, both from a growth and bottom line standpoint versus keeping the cash on the balance sheet.

Okay. Perfect. We're actually out of time. Ken, really appreciate the conversation. This is great. Thank you very much.

Thanks, Josh.