I guess we're ready to go whenever you are.
Let's do it.
All right, great. Thanks, everyone, for joining us today. Appreciate everyone's attendance. My name is Scott Berg. I lead the Enterprise Software and SaaS Research Efforts here at Needham. Today with us, we have Semrush, a company I've covered for, I think, three years now, roughly, plus or minus. We have the company's president, Eugene Levin, and CFO, Brian Mulroy. Thanks for joining us today, gentlemen.
Thanks for having us.
I guess for those less familiar, how about giving you an overview of Semrush?
Sure. I can do that quick, so Semrush is a digital marketing company. In particular, we focus on what we call online awareness, which is basically helping customers leverage all the digital marketing channels to increase traffic and ultimately convert that traffic into paying customers. There are six core digital marketing channels that are of the most important for marketers today. One is search, and we focus on search engine optimization and making sure that they're ranking, able to get data and intelligence about their own assets, their competitors, and what topics they need to be talking about to make sure that they're ranking high within organic search.
A second category is around paid advertising. That's all about optimizing your hard-earned dollars to making sure you're paying for clicks and advertisements in the right platforms and optimizing your spend and the overall return on that spend. Another is in local marketing.
So a lot of companies have a local component to their business, and we want to make sure that we're attracting customers to those specific locations. So doctors, restaurants, and other companies that are relying on that local presence and making sure that they're optimized within GPS and maps and getting that local traffic that's important to them. Brand marketing is important. There's a lot of review websites out there that create a perspective for consumers about a brand or the products that they have. And it's important to be monitoring that and making sure that you're responding to it and driving the right conversation. Social media, of course, is huge these days, and a lot of companies are relying on going viral or at least getting likes and shares and comments about different products or content that they're posting there.
Then finally, content marketing, where we provide a number of tools where companies can build the right content. So six core channels, all interconnected, that create a digital marketing platform. Overall, we have 117,000 paying customers. That spans everything from individual solopreneurs and freelancers right up to Fortune 500 companies, of which we have about 40% that are using our products. We're globally diversified, so we're in 150 countries. What's unique about Semrush is we get a benefit internationally from companies that are looking to promote their products within the U.S. markets and other top-tier markets. So even though they might be in a different tier or a different market, they still have a need for our products to be able to position themselves worldwide and attract consumers to their business. Overall, just from a financial perspective, we're over $400 million in annual recurring revenue.
100% of our business is SaaS recurring technology subscriptions. We have really strong margins. We just reported 12.4% Non-GAAP operating margin at the end of the third quarter, and growth is very strong with ARR up 24% year-over-year as of the end of the third quarter, so quick highlights about what we do and the scenarios in which customers are leveraging our technology and getting value out of our platform.
All right, lots of questions here. I guess I always like to start talking about how you all fit in the broader marketing software space, right? Brian, you just talked about it, digital marketing platform. But I think when the average investor thinks marketing kind of software space, they think of, I don't know, the HubSpots, the ExactTargets type of the world. Kind of how do you fit relative to what they do?
So I think the way markets have developed is a little bit backwards. Because initially, when businesses have started going online, it was very easy to get traffic. You would have a website, and then people would just start showing up and maybe do a little bit of advertising. Most of the time, that's enough. And though the systems that they've needed the most were systems to capture those demands of things like CRM systems, marketing automation. And those are the older categories of software that have developed. And that's why, like you said, people know about HubSpot and people know about ExactTarget of this world. And over time, market developed in a way that everything became way more competitive. So now you are competing with a lot of other players to get this attention.
And to be really visible and get customers online, you need to be visible in all the places where those customers spend time. So social search, watching YouTube videos, reading blogs, or digital media, maybe interacting with your ads and so on. And you want to have a holistic view on how you perform in all those channels relative to competition, where you need to improve, and what exactly would make a difference, where you need to spend more of your time and money. And that's what we do. So we help customers to generate demand across all those key channels. And then once customers have this demand, they can use other marketing automation systems to do a capture and work with existing and loyal customers.
But I think while historically there was probably a greater need in the bottom of the funnel marketing software, I think today there is a much greater need in the top of the funnel marketing software because there are so many things and so many different options you can use for CRMs and marketing automation. And there is only Semrush that can help you with all your digital marketing channels.
There's one very key differentiating factor. So HubSpot, obviously, more CRM, they tend to focus on companies that need to generate leads and then have a human or a sales organization facilitate and nurture that lead through a sales cycle. What we focus on just defining top of funnel is there are companies out there that might have millions of customers. They're selling shoes. They're selling a lower value product at a high volume. They can't rely on a sales cycle and leads. They need to be able to generate awareness, education about their product, make sure that their product is available for purchase online through all these different digital marketing channels.
So we focus on how do you get that top of funnel presence in the most efficient and cost-effective way and make sure you're generating traffic and then converting it into a paying customer, whatever service or product you may have available.
So Brian, you touched on the customers you kind of sell to earlier, a lot of small, a lot of enterprise customers as well. You've had enterprise customers for as long as I've covered the company, but just released a true enterprise kind of platform last summer. How does the functionality differ in that platform? And why would a customer, I don't know, we'll take Disney. It's a customer logo I've seen that you all discussed before. Why would they move from what they were using three years ago to this new enterprise platform?
So I think that's a great starting point. We need to establish what they've been actually using before they could buy Semrush Enterprise. And usually, it's going to be a combination of many different solutions. And often, a big part of this stack is not even going to be marketing specific or search engine marketing specific. It's just going to be a combination of generic knowledge work tools that people use. And they would supplement this and have tools like PLG version or self-service version of Semrush. And they would pull data from us into other systems. But then they would create this totally custom setup to navigate the market environment. And their setup would include things like custom Python code that pulls data from multiple places. And then they need to put it in some kind of database, could be Snowflake or MongoDB or something else.
And then they need to use custom scripts again to run analysis and calculate their segments and custom metrics. And then they would need to visualize it somewhere. So they would use BI system or, in many cases, multiple BI systems. And what we see often is that they started with something like Tableau, and then they wanted to migrate to something like Looker. But they actually didn't finish the project, didn't have enough resources. So some of the older dashboards are in Tableau, and some of the newer dashboards are in Looker. And then when you have this kind of shaky setup, things often break. And when things break, they have to go back to Excel spreadsheets. And we recently had an example of a customer who showed us how they do reporting for their leadership team.
It's an Excel spreadsheet with 100 tabs, and each tab has thousands of lines and multiple different columns. Each column would be populated manually, and someone would need to go into some system, set up filters, download a CSV file, then copy-paste it into this Excel spreadsheet. It's extremely inefficient, error-prone, and ultimately just drives a lot of confusion and missed opportunities. Then what we do differently, of course, we combine all this in one platform that was built specifically with the purpose of helping businesses to gain visibility online and run analysis. Often, it enables them to do things that they could not even do with the setup that they had.
So on top of this old setup being inefficient and costing them a lot of money, we estimate some of them would spend seven figures just maintaining the system and paying data scientists and back-end developers and database managers. So they don't just save money and time, but they also now can do things that they could never do before, things like doing algorithm change analysis. So let's say Google rolls out new algorithm, they need to figure out what has changed, why their rankings have dropped. And previously, it would take them three months. And by the time they're finished with this analysis, Google rolls out a new update, and they have to do the whole thing again, and they didn't have an opportunity to react to that. And with Semrush Enterprise, they push a button, go grab coffee.
By the time they're back, analysis is done, and they know what exactly they need to do to improve their rankings. And we provide the same workflows for other things like internal linking or content optimization or technical website optimization and user experience and many, many other workflows that were not possible with the old setup that they had. So I think that's the key differentiation. We really unlock opportunities that they didn't have before. And of course, from technical implementation, there is a lot of customizable dashboards and a lot of customizable workflows. But I think what they care the most is that they don't have to spend a lot of time and resources maintaining their patchwork. They can focus on their job and what they love and know how to do. And tools enable them to do things they never imagined they could even do.
I know you've only been selling the product for about six months now, roughly. Is there any commonalities or trends across the customers that have either migrated to it or purchased it net new that's worth calling out?
I think we like a lot of examples where people get, let's say, instant value. So I've talked a little bit about, for example, our internal linking workflow. So it's a very sophisticated workflow where we analyze the entire internet to find all the websites that link back to a particular website. And then we look at how the internal linking structure sort of spreads this topical authority across the pages and catalogs. And then we can run machine learning algorithms on this and ask our clients, "OK, tell us what are the most important pages that you need to improve and what are the keywords or topics that you're targeting with those pages?" And then we feed this all into our machine learning AI system. And it gives us back an ideal structure of a website that would help them to achieve their business goals and get more traffic.
What we see, and that was like some of the most rewarding experiences, is once they implement those changes, they get often 15%-20% traffic uplift. For a lot of companies, this can be millions of dollars. If they have 10 million visits every month and they improve their traffic by 20%, that's 2 million extra visits per month, more than 20 million extra visits per year. And if the value of one click for them is, let's say, $1, that's just phenomenal ROI that they're getting almost instantly from the implementation. So I think that's kind of one of the things we're kind of calling out. But there are a lot of other equally important workflows that our solution provides.
Maybe you need to charge more if they're getting that type of uplift.
We're definitely thinking about this. I think we've already done one price increase since launch. So that was one of the points of the feedback that we got from people. They were saying that it provides so much value. It's kind of too cheap because we're hearing questions from procurement. If it's so good, why it's so cheap? So we did increase prices, actually, last year.
You have really good gross margins. I'm looking forward to that increasing maybe a little. Sorry, the analyst in me just keeps extrapolating. Let's talk more holistically on the product side. Gen AI is a theme that I know we're all paying attention to for a variety of different reasons. When I think about the Semrush platform historically, you both talked about SEO, right? SEO a lot. If we think about SEO over the last 20 years, it means Google and Google searches, right? In the last couple of years, we've seen these types of internet searches move off of the Google platform onto some of these Gen AI models, right? Maybe it's for a new pair of shoes. What's the best for me, et cetera? Why is that not bad for Semrush? And why is it actually a positive to how customers want to use your product?
So the way we think about this is that the more complexity in the system and the more channels are out there for consumers to use, the more complicated it becomes for marketers to navigate this landscape, and the more complicated it gets for marketers, the more value our software provides. Because we are kind of like a GPS that helps people to understand how they get from A to B. It's just in what we do, A to B is more traffic rather than getting to a particular location, but the idea is the same, and if ChatGPT or other tools get more popularity to the point where marketers care about them and care about how to optimize for those platforms, then we can start adding more functionality and charging more money for that.
And we're already starting to hear a lot of very kind of positive inquiries from some of our clients who say, "OK, can you help me to understand how I rank in ChatGPT's search? And what do they say about me? Is it positive? Is it negative? How can I influence that?" And we're building a lot of prototypes for them and kind of working on new functionality. And for us, it would be really great if more and more marketers started to care about how their businesses are presented in those places. And of course, that would mean they're willing to buy additional products that help them to optimize for that.
From a technical point of view, we have released several studies that both show what customers are searching in those new generations of AI search engines and how those new search engines rank content differently or similar to Google. So we have all these insights and all this data. And we're now just thinking about how to package this into the product and what would be the right price point, what would be the right buyer persona, what kind of level of demand we should expect short-term, long-term. But we're very, very excited about how this whole space is developing. We also believe that ultimately, it will be very adjacent to search engine optimization. We think discipline will be called AI optimization, so AIO. And there will be a lot of interest, especially from larger companies.
Like you said, Nike really cares how they are positioned in those places and what ChatGPT recommends for best running shoes. Now, small repair shop local might be not that concerned. But yeah, larger brands definitely we started, especially I think when New Year started, we started to hear more and more inquiries.
Moving on to the product functionality side, how do you all infuse some of these technologies then? Obviously, that data is interesting from a search perspective, which I like the monetization opportunity there. But how do you actually infuse it to the functionality that customers might use every day?
So in terms of you mean AI search engines, right?
Yeah, no, just within your own product, like ContentShake and stuff. How are customers using the functionality in those modules?
So we've been seeing sort of this trend very early where ultimately, all the channels are interconnected. And your performance in one channel helps you to perform better in other channels. And that's why we've always been aggregating all the data that we could get and thinking, "OK, what insights I can drive from this information?" And we came up with this idea that when you work on something, there is a certain type of content that you can create for social networks. And there is a certain type of content that you would create for search engines. But it works all much better if you kind of follow the same theme and if your insights sit in the same place where all the different tools can use the same data assets.
And one example is, like you said, ContentShake, where we focus on interesting trends and what your customers might be looking for. And at the same time, we look at what is the search volume for a particular topic and suggest the most interesting ideas to people. And then as they get the idea that they want to cover, we help them to create content that will rank really well in search engines, also will resonate on social media. And we use a lot of AI, LLM, APIs to do that. But if you use just LLM APIs, you're not going to get great results. You just get something very generic that doesn't really rank that well. And what we do, we analyze many, many, many different pieces of content that we know works really well, both in search engines and in social.
And we analyze what are the common things among all the content that works well. And then we can structure prompts in a certain way where LLM will give you very precise output that follows those instructions. And this output is very specific and ranks well. And inputs could include use this keyword, use this tone of voice, refer back to this search, use this data point as information. And if you look at the entire input that you need to give to a large language model to get the exact result you want, the volume of input is actually greater than the volume of output.
So that means that without tools like Semrush, who automate all this analysis and prompt generation for you, you're not going to spend that much time analyzing and creating this prompt because it would take more time to create a prompt than to just create the final output. So we save people a lot of time. And we enable them to do things they were not able to do before. And in doing so, we also help them to create content that performs really well. So that's, I think, the benefit of having this holistic map of the internet and all the data assets in one place. You can use it really well to help large language models and generative AI APIs to create things that work. And ContentShake is, of course, an example for text. But we have the same examples for ads.
We have the same examples for social media posts and ultimately any area of digital marketing.
I know you've taken an approach of having kind of three main levers for pricing around these modules. Can you help us understand why you chose the three different options to monetize the AI functionality?
I think it's an early area where there is no, some kind of book written on how to price AI features. So we essentially want to have multiple different pillars of AI monetization and kind of follow the same playbook that worked for us in the past with respect to other features that we introduced. And the logic is there are certain features where we expect such a high adoption that we just want it to be integrated across every SKU. And the way we monetize those kind of features, for example, our Copilot, is that eventually, as adoption goes up, you can increase prices. And of course, it improves conversion and retention. But the way you easily capture is through pricing power.
And then the second type of AI monetization is also something we've seen with other features that we introduced in the past, is that if you have a certain feature that is very, very valuable, but only for a specific audience, then you want to put it in a package for specifically this audience. And a good example would be our local product, where we use certain AI features as an upgrade trigger to move from entry-level local product to premium level of local product, where they have access to AI-powered reply-to-review feature that will essentially allow you to put on autopilot review management and let AI handle that. And then, of course, some products and features, they're so deep and so valuable that you could just sell them separately as an add-on and charge money for the subscription. And an example would be ContentShake, for example.
So we try all three types of AI monetization. And so far, all of them work really well.
OK. As we start thinking about go-to-market, you had an analyst day recently, obviously well received there. But you're making all sorts of changes. You're obviously investing more into go-to-market as you're moving up market. Those larger companies with larger deals want a little bit of a higher-touch process than just PLG. But with all the changes that you're making, whether it's changing incentives from gross ARR to net ARR and some of the metrics and some of that alignment, as we kind of stir all those things together, what does Semrush look like as kind of a business this year? I mean, how do we think about those changes? Have they impacted any of the business yet today, negatively or positively? And what do we think it does to the business here over the short term?
Yeah, I think a little bit of history would be helpful, and then I'll answer that. Semrush has been in business for 16 years. We started out with a single tool because at the time, search engine optimization was the only digital marketing channel that marketers could take advantage of, and at the time, we sold to the users. We were focused on SEO specialists, highly trained marketers that needed sophisticated tools to do their job. As the digital marketing landscape has continued to evolve, we've now diversified into six channels, so not just search, but paid social, brands, content, and local, and the evolution of our buyer has changed as well, so we started out with specific, highly trained marketers, saw that there was an opportunity with small business owners who also had a need to promote their websites and their brands and their presence online.
Over time, as this platform started to get established as a more sophisticated, enterprise-feature-rich platform, chief marketing officers, digital marketing VPs, and executives started getting interested because it wasn't just about optimizing the work that their individual marketers were doing. It was also about measuring and balancing their hard-earned dollars to be able to optimize it across all the different channels. As that started to materialize, we had to evolve our go-to-market motion to make sure that we were attracting and positioning our platform to leaders and more sophisticated marketing teams and evolve our go-to-market engine to make sure that we can evolve from more of a transactional play focused on the user to more of a trusted advisor status with leadership.
We've been, over the last few years, investing significantly in the caliber of resources who understand that transition from transactional to trusted advisor, the supporting organizations like deal desks, commercial attorneys, and sales operations to be able to facilitate those deals, in addition to those more experienced account executives. Overall, positioning Semrush to be able to basically close deals that are 10x-15 x higher than what our average ARR per customer is. That's been taking hold over the last couple of years and working really well for us. We're going to continue to promote that and evolve that and invest it in 2025. We believe that's going to be a significant driver of growth in the upcoming fiscal year. Having said that, we like all our kids. SMBs are still really important to us.
We like a lot of volume and a lot of companies engaging our platform because we get a secondary benefit where they're sharing data with us. The users are still incredibly important. But now, leadership and those interested in optimizing their budgets across these digital marketing channels is now an additional channel for us. So we're going to be balancing and investing in all three and leveraging all to drive growth. But we do think in 2025 that that enterprise opportunity with the go-to-market and the product combined creates an even greater opportunity for us in the near term.
I have about two or three minutes before I'm going to turn it over to the audience in case they want to ask any questions. I'm only through about half of my questions. So we go really fast or I just ask two. Let's start with average revenue per customer. I think that's a great question. It grew 13.4% in the last quarter. By my calculation, that was the fastest growth rate in that metric that we've seen over the last, I think, eight quarters now. Obviously, the enterprise solution is benefiting that. You just talked about the 10x-15x kind of uplift in a customer using that. But if we put that aside, is there anything else driving some of the expansion in that average revenue per account metric today?
For sure, yeah. We have multiple growth vectors beyond just expanding the number of accounts. But once we have an account and our ability to expand, there's multiple drivers. One is what we've been talking about. We have a comprehensive digital marketing platform across six core channels with data and intelligence at its core. We tend to land with one of those. So we have a tip of spear where each channel is independently positioned out in the market to be able to attract marketers. But then they understand that they're all interconnected, and you can't just focus on one. So they tend to expand and adopt more of our features and our products across those different channels. We continue to focus on pricing. We did talk about our enterprise portfolio and how we've increased the price already, even though we just launched it in May.
A third key factor is we're starting to launch much more sophisticated and complete solutions to, as you said, Scott, have a 10x-15x uplift on price compared to our average core portfolio. And it's working. As a proof point to that, we just launched this enterprise SEO product in May. Prior to adopting that, we have about 90-100 customers at the end of November this year. We'll update everyone after we report our fourth quarter earnings. But roughly about 90-100. Prior to adopting this enterprise solution, they were paying us around $10,000 in ARR per year. And on average, they're now paying $60,000. And we've actually seen one that's paying as high as $350,000 a year.
So there's good scalability and growth because of the completeness of our platform and the level of sophistication that's of particular interest to the enterprises that are out there.
You might have a new customer cohort to report against if you're going to sell more 350,000 deals.
Totally.
I'm looking forward to that day.
We have a 10,000 + 50,000. Maybe we'll do 250 ,000+.
Awesome. Last question for me is, what are you all most excited about in the next five years?
You want to start?
Yeah, I think I'm most excited about the fundamental kind of problem that we are helping to solve. And the way I usually frame this problem is that digital population number of people online is not really going to grow that much. And the amount of time that people spend daily scrolling the internet is not going to increase that much. We already spend too much time, probably. And if you look at marketing dollars and just in general, the go-to-market spend across the board, it only goes up and up and up and up and up. So the same number of people, more dollars chasing the same number of eyeballs is going to be harder and harder.
I think tools like Semrush are really the only way for people to understand where they should go, where they have a chance to compete, where it's going to be too tough. I'm really excited about our role in this growing ecosystem of kind of holding customers' hands and helping them to get safely and navigate this very challenging new reality.
And for me, I mean, this has been a long time coming. But we've invested heavily into a unique combination of data and technology assets that are game-changing for companies and for marketers. And at the same time that we've been doing those investments, the digital marketing landscape has continued to evolve. It's continued to become, as Eugene said, incredibly complex. And AI is just now another component that marketers and business owners need to be focused on that's going to be disruptive. And for marketers and businesses that just sort of maintain the status quo, they'll be left behind. And we have tools and resources, a complete digital marketing platform, highly sophisticated enterprise features.
I think Semrush is uniquely positioned where we're the only ones in our market who have made those investments, who have positioned ourselves with a loyal customer base that's now 117,000 strong, and are going to be well positioned to be able to capitalize on what we see as a $40 billion opportunity over the next five to 10 years.
Excellent. Happy to take any questions from the audience if there are any.
Curious in your enterprise solutions, what functions do your customers use? Is it purely marketing or is it other?
So with the current portfolio, it's mostly marketing. We do have use cases for R&D teams where they work with marketing on implementation. So for example, we call it technical optimization product. And what it does, it analyzes the website, and it finds technical issues like this page loads too slow or this is just bad customer experience because the page loads half of the page quickly, but the other half slowly. And it's frustrating. And we can highlight all of those user experience issues. And the solution is going to be usually on the engineering side, and marketing teams will have to collaborate. And for people for whom this is critical and they don't want to release anything that breaks the flow or breaks their rankings, we see those tools being implemented in their deployment process.
So before they release the new version of the website or updated version of the website to everyone, they would run this analysis on our dev side. And if it doesn't break anything, then they release. And if there are too many issues, they will not release. So that's one of the use cases where both marketing and engineering teams use the product. And I think our core focus, of course, is marketing. But marketing is a function that relies on a lot of other people. And the more things we provide for marketers, the more it will overlap with other functions where there are dependencies.
Any other questions?