All right, great. Thanks so much, everyone, for joining us. My name's Adam Hotchkiss, and I cover the emerging software space here at Goldman. Really excited to have Eugene Levin, President of Semrush, and Brian Mulroy, CFO. Thanks so much for being here.
Thanks for having us.
Of course. You know, I guess Eugene and Brian, feel free to chime in as well. For those in the audience less familiar with Semrush, just maybe give a brief overview of what Semrush does and what you're trying to build as a company.
Sure, absolutely. I can do that. So just a quick high-level overview. We're a digital marketing company. Our focus is on what we call top-of-funnel digital marketing, which is basically all the channels that are important to business owners and marketers to generate traffic and awareness about their brands and ultimately convert that traffic to paying customers. We have seven core channels we talk about and focus on. One is AI and brands and businesses trying to figure out how to navigate this new AI landscape and figure out how to monitor their brand's perception, overall mention, and overall share of voice within this new AI landscape, and then shape it to make sure that it's actually favorable and accurate based on how they want to be representing their products, brand, and services out there in the digital landscape. Second is traditional search.
There's still 15 billion searches a day on Google, and marketers and business owners want to make sure that their content is ranking and visible and they're getting the appropriate recognition out there based on their overall business needs. And then there's other social media, local marketing, content marketing, brand marketing, and we have a data and intelligence layers that allows them to gain important insights and actionable insights about their brand online. Overall, from a financial perspective, we achieved $435 million in annual recurring revenue, and our guidance for 2025 is revenue growth of 18% and expanded profitability. So I know we'll dive into a lot more in terms of our go-to-market and overall product strategy, but that's an overview of the business.
Great. Before I begin, maybe Eugene and Brian just give brief backgrounds on yourself would be helpful to start.
Yeah, sure. I'm Eugene. I've been with the company for about 10 years. I'm President at Semrush, and so have seen it all. When I joined, we were a small shop, maybe about $16 million ARR or so, so we grew the business more than 25x , which I'm very proud about. And before that, I was in venture capital, so did everything from seed stage to late stage, pre-IPO, technology, software, consumer internet, pretty much everything.
Brian Mulroy, I'm the CFO here at Semrush. I've been here for two and a half years, and I've been in software technology and particularly finance for pretty much my entire 25-year career. Prior to Semrush, I was at Microsoft, Nuance Communications near Boston, and other software technology companies over the years.
Fantastic. I think a great place to start would be the changing search landscape. Obviously, SEO is one of the core things that you do as a company. You've talked about doing AI optimization in terms of helping folks get visibility into how the large language models and things like ChatGPT are affecting their brands and businesses. Maybe just let's take a step back and look at the high level. What, in your mind, is happening in the search landscape? How is Semrush addressing it?
So the way we think about this is that AI adds new use cases that didn't really exist before AI, such as multimodal search, more complex queries. You can ask AI things that Google in pre-AI era would struggle with. And so as people learn about those new use cases, they use AI search increasingly frequently. And the total combination of, let's say, traditional search use cases and AI search use cases is much greater than all the use cases that existed before. And so as a result, pie gets bigger. So more and more search interactions happen today, and it will keep growing until AI search adoption reaches, you know, let's say, 100% or something close to 100%. And at the same time, brands need to figure out how they appear in AI search. What does it tell about them? Is it beneficial? Does it drive purchases or not?
And while this new discipline is very similar to traditional search engine optimization in terms of research and measurement, it also has a lot of differences in terms of how you influence the output of the model. So for traditional search, the goal was to rank your own website, and most of the work that you would do would be technical optimization of your website and content creation, and you would optimize primarily content published on your website. And then, of course, you would work on the authority, but you would try to build authority for the website, not necessarily for brand in a broader sense. And then for AI optimization, it's a bit different because you're not necessarily trying to make sure that AI chooses your content in every instance.
What you're trying to do is figure out what content influences the output of AI prompt and work with those sources to make sure they have correct and favorable representation of your brand. So it's about building authority, but not just for your website, building authority for your entire brand and your products. And that includes working with influencers, with media, managing reviews, managing your overall online reputation, and of course, publishing great content that people want to share.
Okay, that's great. And how do you do that in an AI world, in a setting like OpenAI, ChatGPT, or the LLM outputs on Microsoft or Google relative to the way we all were used to using search and seeing sponsored pages and organic pages?
So right now, most of the output is all organic. We think eventually, of course, Google will introduce a monetization model for Google AI mode, so there will be some sponsored links there. We don't really know how exactly this UX is going to look, but there will be options for brands to promote themselves by paying money. For ChatGPT, they may actually choose to not do this anytime soon. They're making a lot of money through a premium subscription model, and so I don't see why they would need to urgently build an advertising arm of their business. So their results might remain purely organic for a foreseeable future. And I think that's something that brands need to understand. You cannot really go and buy yourself visibility in AI search, at least not today.
And so you need to think about organic and start building your authority today because if you don't do this today, someone else will do it tomorrow. And at the end of the day, AI cannot mention an infinite number of brands in their reply.
Okay, that's really helpful. And next one for you, Brian, onto key performance. I know you lowered guidance, and you mentioned decisions around resource reallocation at market rather than demand issues as causing some of the changes in the revenue guidance. So maybe just talk about the quarter a little bit and what drove the change in the guidance.
Yeah, our guidance for 2025 is 18% growth, so still very strong, durable growth for Semrush. We're focused. Our priorities and growth vectors for now are focused on AI and capitalizing on that opportunity. We see it as a once-in-a-lifetime generational shift where, as Eugene mentioned, marketers need to think differently and focus on more than just the traditional channels that they've been relying on, so we're investing very heavily in AI. Enterprise is a huge opportunity for us. We noticed over the years that a lot of enterprises were piecing together point solutions, and no one had a standard digital marketing platform that allowed them to focus on all aspects of digital marketing, and we've invested in our product portfolio to create that AI-first enterprise comprehensive platform that allows marketers and businesses to displace a lot of that point solution fragmentation.
We're investing very heavily in the product and the overall go-to-market that elevates us from being a transactional player to one that's more of a trusted advisor and partner to figure out their needs and make sure that they're capturing value from our expansive enterprise platform. There's some parts of the business that are a little softer than others. Our enterprise business overall is growing over 30%, has really good, strong net revenue retention, and we expect that will continue to perform really well over the years. On the flip side, there's a cohort of freelancers where, because of macro dynamics, the performance there is a little softer than it's been in the past. What we're seeing is there was a surge in demand post-pandemic.
Everyone was taking advantage of the Great Resignation and starting these small, basically hanging a shingle and starting small agencies, and the supply of freelancers just outpaced the demand, and we're seeing a reversion to the mean trend there. Even with that, though, we still continue to deliver really strong growth, and our focus and priority growth factors are on AI enterprise from a product and a go-to-market perspective.
It's a really good segue on the first point there to the enterprise piece. Roughly 260 enterprise customers with the new enterprise product out of the 9,000 potential accounts you've identified. It's about 3% penetration. How do you go about with your current resourcing and profitability objectives, targeting the right enterprises to go after next? And then how do you think about, I know you've talked about this ACV uplift. I think you said $60,000 the last time we spoke. How do you think about what that opportunity looks like longer term?
Definitely. It's a great question. Prior to us embarking on this enterprise product and go-to-market journey, Semrush, through really low touch, no touch type of go-to-market motion, had acquired 9,000 enterprises. We're in directly, 40% of the Fortune 500, and then through agencies, pretty much all of the Fortune 500 accounts are using Semrush either directly or through those agencies. And we saw that as a big opportunity. So we launched a year ago an enterprise version of our platform, and in the last year, it's the third fastest product. The first two are actually new AI products that we've launched even more recently. But the enterprise product has been performing well, building momentum, and our goal is to convert all 9,000 plus all the enterprise-sized accounts out there that have yet to use Semrush and to capture them as a Semrush enterprise customer.
About 60% of those who have adopted it so far are existing Semrush customers that we've converted from our core platform to the enterprise. 40%, though, are brand new logos that are new to Semrush. So it's a big growth factor for us and something that will allow us to drive durable growth into the long term.
Brian, or Eugene, either of you, what does that function really look like? When you had brought on 9,000 customers sort of through the low touch model, was that just a couple of employees? Because I often hear when I talk to folks that come from the agencies, right? There are folks who are perennial Semrush users, and so when they move to an enterprise, they may bring that experience with them. Do you often see that way where the low touch model brought you to maybe a couple of employees within at the lower levels of an organization, and your goal is to move it higher up into the organizational level? How does that mechanically look for you?
Yeah, it's a mix of both. It is certainly a large cohort of customers, whether enterprise-sized or even SMBs, who find Semrush, realize there's a tremendous amount of value in our platform, and are able to generate a good return on investment, and many of them come through with our own top-of-funnel online awareness activities, do a trial, register, and then convert to a paid customer with a credit card. Over the years, though, as we've expanded our platform and, of course, our influence in these accounts, many of them, even before we had enterprise, had already evolved to using our full portfolio and, in certain cases, paying us over $100,000 a year in annual recurring revenue. In those cases, we've already gone through the process to negotiate contracts, work through their procurements and legal, and, of course, establish a relationship with their digital marketing leader team.
We're kind of, we feel like we're already ahead and have already established a relationship in the contracting structure and are using this enterprise portfolio investment to just accelerate and expand our footprint in a lot of the sophisticated enterprise accounts.
Okay, very helpful. Let's move on to AI Toolkit. I think you had mentioned two products are growing faster. This is sort of explaining that. Fastest growing product in company history reached $3 million ARR within months of launch. And when you combine enterprise and AI, you've said that you expect that ARR to double to $50 million in ARR by year-end versus where it was in the second quarter.
That's right.
What gives you the confidence there?
Yeah, just to kind of map out numbers. A year ago, we didn't have any enterprise products and no AI-first products that help companies to essentially navigate through and understand their online position and the shape that within the new AI search landscape. So nearly zero a year ago. By the end of the second quarter, the combination of our AI products, two of which were just launched in the first and second quarter towards the end, so only a couple of weeks or a couple of months in, and our enterprise portfolio had already achieved about $25 million in annual recurring revenue, so combined across those two focus areas. By the end of the year, we expect $50 million, and it's a function of really three things. One is we're seeing really strong interest and demands for AI and our enterprise portfolio.
I think brands are recognizing to compete in this new landscape that they need to think differently and to leverage actionable data and intelligence workflow and technology that gives them the visibility and ability to shape their brands in this new AI-first world. So there's a lot of interest. We're seeing a lot of momentum building. And then, of course, in the enterprise play, we've been investing in the products. We launched a second AI, sorry, enterprise product in June and then a third one in July of this year. So we're expanding the overall enterprise capabilities and the average ARR potential in that particular growth vector. And also investing in the sales organization, which is continuing to ramp up, and we're seeing traditional second-half loaded type linearity and seasonality with the sales organization.
It's a combination of really strong demand and interest, new products that are in the market, and then scaling up our enterprise go-to-market that gives us confidence that we can continue that strong momentum and double the size of those products by the end of the year.
Okay, that's helpful. And who do you often most compete with in enterprise? And then what unique advantages do you have when you think about your AI tools versus some other folks in the market?
Yeah, I think in enterprise, the main competition is usually with something homegrown, so essentially convincing people to abandon old ways and embrace kind of new platform approach. One of the best examples that I have is we have a client which is a very, very big corporation, and as we were onboarding them, they sent us their report that they wanted us to replicate in Semrush enterprise platform, and it was a report that would go pretty much all the way up to CEO level, of course, in a processed way with charts and everything, but it would be based on the same Excel spreadsheet that we received, and this Excel spreadsheet had over 100 tabs. Each tab would have 23 columns. Each column would need to be populated manually. Like you would need to go and download something from somewhere and then manually paste it.
And you would have to do this every month at least. And then there would be some formulas. For example, they would have custom definitions for their share of voice metric, and they would have custom segmentation. And often you would make an error either when you copy-paste something or in the formula. And so the whole research, sorry, the whole report is wrong, and it may take you months to find out that it was actually wrong. And sometimes you have been reporting wrong numbers for a while.
And this is the best example that I have to show how bad sometimes status quo is in those large corporations and how much they need to move from this approach to a true enterprise platform that does all those things automatically, that updates in real time, where you have certain mechanisms to make sure there are no errors, there's certain quality control, and so on. And then in terms of complexity of those setups, sometimes a spreadsheet is not the worst thing we've seen. Sometimes you have some kind of custom code written on Python that pulls data from somewhere, and then the guy who wrote that code is not with the company anymore. And so there is another person who tries to maintain it but struggles to make any changes because this person is not an engineer.
And then sometimes they maybe migrated from one BI system to another BI system. And then in the process, they had critical dashboards that, let's say, C-level consumers, and they migrated those dashboards. But for other dashboards, they didn't have resources. So they have one type of reporting in one BI system and another type of reporting in another system, and they're completely inconsistent between each other. And so those are realities of SEO in large corporations and how they do that. And of course, it impedes the decision-making process. And there are certain things that they need to do that they cannot do with this kind of data set of things like, for example, internal linking or programmatic SEO or content creation at scale, or even understanding what has happened with your website during one of the algorithm updates.
And so our Enterprise SEO platform solves it all, and now it also solves the same kind of problems for AI optimization. It's just with AI optimization, there hasn't been enough time to build all this ridiculous legacy. But in terms of competition, I think that's really what we are fighting against. The point solutions in Enterprise SEO, they've never been very competitive because most of the businesses would rather use Semrush PLG version and supplement it with their homegrown code than use point solution Enterprise SEO.
Okay, that's helpful. Why didn't something like this exist then? I feel like whenever I hear about these case studies of such inefficiency within a business, it feels like within the SaaS world, a lot of that has been solved at this point. Why do you think that hadn't been solved before your product?
Right. So like I said, in enterprise SEO, there are players who offer their products, but they have much weaker data. And so a lot of large enterprises, they looked at it and said, "We can add a lot of functionality through custom development internally, but we cannot solve that data problem that a lot of those other companies have." So because Semrush always had superior data, they would use Semrush and then try to build something around it. And that's why it's so easy for us to go and replace this homegrown kind of patchwork because we're saying it's all based on Semrush data anyway. We just have much better tooling around this data now that works out of the box and that we will maintain.
And so you can allocate your R&D and data science resources in places where they would have better ROI than maintaining kind of clunky reporting infrastructure or data analytics infrastructure. So it's a very compelling value proposition.
Yeah. I want to get to the financial questions, but I just thought of one that I think to follow up on that. We've heard a lot about the death of software in the last couple of months. And I would love to just have you give, and I think this goes to your data mode as well and maybe explaining what the mode around your data is. What does Semrush look like in a post-generative AI world, and what gives you the confidence that your platform won't be displaced by some of these AI-native applications that are coming out?
Yeah, I think you're absolutely right. The data that we collect and the way we process it is the most efficient way to do this. And AI can present information differently, or AI can maybe simplify analysis. For example, we're building a lot of workflows where things that previously would take hours to combine into report now take minutes. So we definitely leverage an AI to do that. But in terms of fundamental data, there is still a cost of collecting and storing data, and our infrastructure is just designed to do this in the most efficient way. On top of that, there are network effects with metrics such as traffic estimates or search volume estimates, where they're accurate because we have access to source of truth data that our customers share with us.
And so then we can combine our signals from publicly available information from the internet and from our panels with the source of truth data to calculate metrics such as search volume estimates and traffic estimates. And the more source of truth data we have, the more accurate our estimates become. And the more accurate our estimates are, the more attractive this data is for customers who use the product, share even more of their data, which we use to make our metrics even better, kind of reinforcing the flywheel effect. And again, there's nothing that AI does to that, right? So while I understand the sentiment, and I think in terms of interfaces, interfaces I think become indeed less valuable in the future. And the fact that you can simply type what you want and get the result is very powerful and disruptive.
However, Semrush product is not really interface heavy. Most of the value that we provide is data, and interface is really just one of the ways to get this information. And we're okay if you're using our interface. We're okay if you're using our API. We're okay if you're using MCP, which we launched today. We just want you to use the best data to drive your marketing decisions. And the interface is really a secondary consideration here.
Okay. That's really helpful. Brian, on the enterprise and AI product portfolio as it stands today, as you think about moving towards your target ARR per customer longer term, do you already have the necessary product components that you need, particularly on the enterprise side, to be able to get there? Or do you foresee inorganic and organic product launches needed to do that?
We do now, yeah. In Analyst Day, and even in the last four quarters, we've been talking about average ARR for the enterprise cohort expanding to over $100,000, and the Enterprise SEO product that we launched in June of 2024, so about a year ago, got us to $ 60,000. The rest of the way there came in June of this year where we launched our Enterprise AIO product, basically an enterprise version of everything we've been talking about tonight. It's a product that allows you to monitor, analyze, and then shape your brand inside these new AI digital marketing channels, and then in July, we launched Enterprise Site Intelligence, so combined across those three products, there's more to come. There's, of course, other non-enterprise components of our portfolio that enterprises absolutely get value out of and continue to use.
But all in, we're already up over $ 100,000 in average ARR for the cohort of customers that are adopting this enterprise platform.
Fantastic. And then on operating margins, you're maintaining 12% operating margins. I know there are incremental investments going in, but you're balancing that with growth and profitability. Just more broadly, as you start to think steady state for this business, how do you think about the balance between growth and profitability?
Yeah, it's something we have to think about every single day, that's for sure, and we're a growth company and continuing to invest and continuing to make sure that our company is positioned to maintain our strong market-leading position and, of course, continue to deliver durable growth, which as of 2025 is 18%. We're also able to take advantage of incredible synergies and efficiencies throughout the business. Eugene mentioned our data collection and ability to scale those data assets is the best out there. That, along with our ability to scale the business across a very wide cohort of customers, we have 116,000 customers and continue to grow and scale our business. It gives us a structural advantage. So our gross margins are really strong. They continue to improve.
It gives us a structural advantage to reinvest back into the R&D, sales, and marketing capabilities of the business, but also to generate expanded profitability. Even within our operations of the business, we're continuing to take advantage of AI internally. We're automating a significant number of customer engagements from both a customer success onboarding and customer support perspective. Our R&D teams are using it to bring products to market faster and to be more efficient in the work that they're doing. We're leveraging it in hundreds of ways throughout the organization. The combination of our scale, the structural advantage that our gross margin provides, and our ability to gain efficiencies in part through AI continues to allow us to invest in the business and maintain a growth priority, but also to expand our margins. That's something that we'll continue to do.
Okay, great. Any questions in the audience? Brian, on capital allocation, how are you balancing your thought process between opportunistic M&A, share repurchases, and investment?
Yeah, Semrush made it easy on me, so we have $260 million in cash, so it's about as flexible as you can get. It's, of course, our first priority to invest organically in a business to maintain that AI and enterprise edge and still have plenty left over to do M&A. We're focused on close adjacencies and things that allow us to accelerate our strategic imperatives and priorities forward, so we'll continue to focus on both organic investments and inorganic investments as they make sense. We even have, because we have $260 million, we're afforded the opportunity at a time where we feel our valuation is attractive and where Semrush needs to convey conviction about the AI and enterprise opportunity that we can even return capital to shareholders through the share repurchase.
So we're doing all three and feel like we have plenty of firepower to invest in the business and enable the share repurchase as well.
Fantastic. Last question for me, for both of you. Number one, what is the one thing you're most focused on over the next 6-12 months? And then as you look further out, 3-5 years, what are you most excited about from an opportunity perspective for Semrush?
Sure. Look, our vision as a company is to fuel every digital marketing journey. It's been the vision of the company since we started. As digital marketing continues to evolve and marketers get more sophisticated and need more intelligence and workflow to do their best work, we've kept pace and launched new products to market that fuel their journey regardless of what channel is important to them. In the three, five years, we're going to continue with that vision and continue to keep pace with the evolving digital marketing landscape and expanding sophistication of marketers. In the next 6 - 12 months, the digital marketing channel is important and new to marketers is all about AI.
Our focus and priority number one is to make sure that Semrush is enabling marketers and that we're essentially serving as the AI or LLM optimization layer that's important for marketers and business owners to succeed in this new landscape.
Okay. Eugene?
I'm really excited about AI. So we've talked a lot about AI optimization, how we're helping our customers to be visible in this brave new world. At the same time, I'm equally excited about all the AI features that we are implementing in our product, AI workflow automation that we are adding to the portfolio of products. And yeah, my goal for the next five years is that things that would take you hours or maybe would not even be possible will now be not just possible, but would be something that you can do within minutes if you can just imagine them.
Wow. Very inspiring. Appreciate it. Thanks so much for your time, Eugene and Brian.
All right. Thanks, everyone.
Thank you.
Appreciate it.
All right. Thanks. Great job.