HubSpot, Inc. (HUBS) Earnings Call Transcript & Summary

All right. Fantastic. We will go ahead and kick it off. It's a real pleasure to welcome Yamini Rangan back on stage with me, CEO of HubSpot, particularly off of the massive success of INBOUND last week. Thank you so much for joining us.

Thank you, Gabriela. It's always a pleasure to come back and thanks for having us.

So I know how much energy you and your leadership team put into the messaging at INBOUND and the product and the functionality that you're announcing. Now that you've had a whole weekend to digestion process, would love to hear what stood out to you? What was the one takeaway from INBOUND that perhaps investors should be aware of?

Yes. For those of you who don't know, we have a conference called INBOUND, it's a customer conference. We had about 13,000 people. And for the first time, we did it here in San Francisco. For the last 15 years, we have done it in Boston. So this time, it was in Moscone. And I'd say that across the Board, the pace of innovation at HubSpot shown. We had 200-plus updates and features and product releases across the Board, across the entire platform. But if I were to pick a couple of themes, I would say the first one is data. And we all know this is the foundation for anything to do with AI is data. And so we actually transformed what we had originally called an operations hub into a data hub -- and this now brings the data together. It uses AI to make it very smart, you can enrich any data by running LLM queries -- and it allows for customers to take that data, bring it into reports and workflows. So I think almost every single customer I spoke to since INBOUND have mentioned data and the value of having the data in a simple streamlined manner in order to enable AI. And then the second big thing, which we will talk a lot about is marketing. There's just such a big change that's happening in marketing now, and our customers want us to guide them through the process with playbooks, with products as well as the entire ecosystem. So that was a big theme. And then third, all the stuff that innovation that we have done in terms of AI, specifically the AI agents and the marketplace and the custom agent builder that we launched last week. I'm sure we'll talk about each of those, but those were the top 3 themes.

Yes, absolutely. Let's start on the data piece of it. One of the things that I've appreciated about HubSpot is you're the CRM and the front office suite for growing and scaling companies. So talk to us a little bit, when you look at your customer base, how far along do you think they are in getting that data strategy organized such that they're then ready to take the next steps and adopt AI agents?

Yes. And I think that's a broader question. One of the reasons why HubSpot consistently wins and more folks are consolidating on HubSpot is the data story. And so that has been part of the reason why we now have over 270,000 customers, and that's because of one simple thing. Our core platform brings together data for marketing sales, service ops teams. And if we ask our customers that's a single reason that they buy, it will be one of the top three reasons. And so I think that's number one. Specifically, what's happening with AI is that you can now do much more in terms of looking at intense signals even before someone shows up on your website. That's a huge change in terms of what you can do with data and having the ability to capture intent signals on multiple channels, and then being able to take actions and then being able to drive agents is such a critical part of the strategy for scaling companies. And you put these two together, you need to have a clear data story. To your question, where are our customers in the journey? Like any of those, they are in multiple parts of the journey. There are some that have been using HubSpot for years. They already have their customer data together. Now they want to get better intent signals and do much more sophisticated things with AI. They are more advanced. And then there are some that are just starting their AI journey and realize that they need clean, not siloed data to begin the journey and they are starting. So I would say kind of across the spectrum.

Yes, absolutely. And the other thing you mentioned there is marketing. And I thought the way that you introduced the group really takes INBOUND and brings it into 2025. Tell us a little bit more about how you purchase this concept of the loop and how it takes into account the way that the interactions are changing between the brands and their customers.

Yes. So for those of you that were not there at our conference, we launched a new marketing playbook called the loop -- and the context is this. We all know that search is getting disrupted. And the easiest way to think about why search is getting disrupted is AI overviews are providing answers at the top of every search. And the data point that you should all remember is that 60% of Google searches today end up with 0 clicks, which means less traffic to website and less content leads. So that is the disruption that is happening. The untold part of the story, the other part of the story is that what AI is taking away in terms of content leads it is doing something very different. It is allowing for you to capture intense signals, match your customers' intent with the information that they need to see and driving much better conversion in the marketing funnel. So in marketing language, if you want to say the top of the funnel is getting disrupted because of AI overviews. The bottom of the funnel is actually transforming with better conversions from AI. So there are two parts to the story. And if you put these two together, marketing playbook needs to change. It needs to fundamentally look very different. And therefore, we launched a new playbook. We've been iterating on this playbook for the past couple of years. It's powered HubSpot's growth and now we launched it for our customers. And the way to think about it is threefold. First of all, you should start with human voice and authenticity. Otherwise, AI is just going to give out [ SAM ]. Then you got to like use the data, all the intent signals that you can now gather about your customers to drive much better personalization of messages at scale, that's kind of the second step within the loop. And then you've got to be in every channel. The diversification of channels in marketing is such an important part of the story. You cannot wait for customers to come to your website. You have to be in social channels. You have to be in podcast, you have to be newsletters and there is a new channel now called LLM and therefore, there's a whole approach in marketing called AEO, which is AI Engine Optimization. So to put this in a nutshell, the marketing playbook is changing. What we launched is how you use human authenticity with AI's efficiency to diversify your content and channels and grow with AI. This is a big deal for our customers. This is a big massive opportunity for HubSpot, because we're not only providing the playbook that we launched, but we are also supporting it with our products across Content Hub, Data Hub and Marketing Hub. And we are helping them reimagining how they can drive growth with AI and was a pretty exciting moment for us to launch it. And all week long and continuing into this week, we've been talking to customers about how to help them adopt it.

So you made a really interesting comment there on, this has been incubating at HubSpot for a couple of years as you've transformed your own internal marketing facility. So tell us a little more about that. What has been the iteration process internally? How do you feel about your own internal strategy to get new customers on HubSpot.

Yes. I think that's a great question, Gabriela. So a lot of the articles this year has been about search disruption. However, for us, the journey began in 2022, because in 2022, second half of 2022, we noticed that our customers were spending less time on blogs and much more time on social, on podcast and newsletters. And this was before AI overviews even came into the picture. So we did a few things. We actually made a huge set of bets back then to diversify our marketing channels. Some of you may remember that back in 2022, we acquired a media company called the Hustle. And at the time, it was like what are we doing? But the reason we did that is we wanted a business podcast network. And that network strategy has diversified, and we now get 90% growth in terms of leads from that podcast network channel. Same thing back in 2022, we also acquired a couple of newsletters, because we saw that e-mail newsletters is also a new way to reach customers, and that was pretty big. So the changes that we made in terms of our marketing channels and strategy started in 2022, and it was launching YouTube, launching Instagram, launching podcast network, launching newsletters and all of that has paid off. And if you put it all together, we've seen a pretty dramatic shift in terms of HubSpot demand mix shifting over the past three years, leads obviously from blogs and web apps have gone down over the past three years. But at the same time, we've seen leads from social, from e-mail, newsletters and podcast more than double in that same three-year period. So diversification is really important. One of the other things that we've started experimenting and this is early, it's new is AEO that I just mentioned. It's a pretty early approach to kind of being part of an answer rather than being part of the search result. SCO and the playbook of the prior generation was all about being in the top 5 blue links that you get. Now there's a new approach and a methodology of showing up in the results, the one single answer that an LLM provides. And it's super nascent, but we've been experimenting with it. And what we are finding in terms of AEO is that the strategy, the content is the same. But the strategies that you need to use in order to come up in the answer is quite different than what you did to show up in the links and we've been experimenting with it. Our internal leads from AI and LLM have grown tremendously over the past couple of years, small base. But the more interesting thing is that leads from LLMs are converting 3x better. And that is because when someone asks the question of an LLM, it is -- they are much deeper in research. They have much higher intent and they are ready to act. And this, again, by the way, is a big breakthrough for marketeers. In marketing, you don't get 3x conversion improvement in any channel. You mostly get 5% to 10% improvement. So this is actually a pretty big breakthrough. And so now we're helping our customers begin to think about AEO strategy. And this is going to be multiple years in the making in terms of how to show up in an answer and how to figure out how LLMs work in terms of marketing messages. But the bottom line is this. We diversified our channels. We've been ahead and looking at all of the changing dynamics within the marketing landscape and the things that HubSpot does really well is this. We iterate -- and then once we know the playbook, we educate our customers, and then we activate our entire ecosystem of solution partners to go out and help our customers win. That is what we did with INBOUND. It was playbook. It was a set of products, and it was an entire ecosystem behind it, and that's exactly the approach and strategy that we are taking with loop, and we think the opportunity is pretty massive for HubSpot.

This idea of AEO is a really interesting one. I know it's early days. Any initial observations as to what that strategy could look like? How do you figure out? Is it primarily through iterations. How do you figure out what works well and how to get that lead conversion.

Yes. I think -- to answer the question, there's -- you have to understand the difference between how to show up in a blue link and how to show up in an answer. The average LLM question is 23 words. And the average search was 5 words. And that begins to tell the difference between AEO and SCO. That means customers and prospects are asking very specific questions of LLM, -- it's not, "hey, what CRM should I use"? It is I am -- would manufacturer in the Midwest serving population that looks like this ideal customer profile, what are the ways in which I can grow my lead. Now in order for you to show up in that answer, you got to have very specific staff, you have to have quotes. You have to repeat your content across multiple sources. And so there's a whole strategy and marketing customers are super excited to do it. And the best way I would describe it is a ton of iteration and experimentation with your content. The way our teams are figuring it out is not just have one case study but have like very specific case studies for every industry with very high repetition of the content in multiple channels. And when you begin to do that, you're also going to need the visibility because what's happening within AEO is that you got to know, are you first of all showing up in the kinds of questions that you want to show up. You need marketing tools to be able to do that. Then what is your share of voice within LLMs compared to all of your competitors? You need marketing tools to be able to do that? And then how do you drive better conversion of your content through LLMs and AI referrals, you need marketing tools to do that. And so I think that's where we get excited about the opportunity in front of us.

Yes, absolutely. And you're describing a strategy that has required a lot of heavy lifting from an R&D, from a product innovation standpoint. The question I want to ask you is, -- the beauty of HubSpot's tech stack with the primary color is you're able to do this kind of evolution or transformation. I wouldn't use the word seamlessly, but you're able to execute it in a way that's incredibly fast based. So my question is, are you seeing HubSpot separate out more from the existing SaaS competition because of the technical decisions that Dharmesh and Brian made very early on and because of how you've organized your R&D team?

Yes. I love this point about the primary colors that you're making. And we've said this for many years. A couple of choices that we made strategically really early on. One is that we're going to build our platform and organically build it, and that drives value for our customers, because when customers have to take a bunch of cobble together acquisitions and put it, they take the owners on like integrating it and unifying the data. We want to take that paying for customers and that really has a huge impact in terms of a smaller company trying to grow. That was choice #1, which is like we'll build it. And then as soon as we made the decision, we also made a decision to be platform first -- and what that means is that we have a set of primitives within the platform. Think of it as automation, reporting, data and those primitives are across multiple applications and hubs, which means if we improve a platform level work in automation, that's going to show up in Marketing Hub. That's going to show up in Sales Hub and Service Hub. And to your point, we've taken an exactly similar approach for AI. We have a set of AI primitives, AI skills being one within the platform, which allows us to build agents faster and build products and features within the hubs much better. And so -- what I think you're getting to is that is speed a moat. And I do think being platform first and having a set of core capabilities within the platform allows us to move with speed. And one of the things that this translates into for our customers is just the pace of innovation. I talk to customers all the time, and I tell them when you buy HubSpot, you're future-proofing, your technology investment because we are driving a pace of innovation, and that deeply resonates with our customers.

Yes, absolutely. Well, let me flip the question around from a competition standpoint. You're iterating on an existing tech stack that was built 10-plus years ago. You set it up to succeed. But there are companies that have been born in the last 1, 2, 3 years that will say, we're starting with a clean sheet of paper. The AI native tech stack is fundamentally different from the SaaS tech stack. Therefore, we can build something clean and more beautiful, more dynamic, more disruptive. How would you respond to that claim that the AI native companies can be more successful relative to companies that are starting with an existing tech stack?

Is it fundamentally different in terms of the tech stack.

I wouldn't love to hear you. Yes.

I think that there are changes that AI is enabling. And I would characterize the changes that AI are enabling into 3 things. The first one is that context that AI can process is much broader. So if you think about an application like HubSpot, we always had structured data. That's what we are known for. CRM is structured data of companies, contacts, deals, tickets. That's what we are known for. With AI, we can now process unstructured data and external intent signals at a much better pace than before. And so our contact layer has now transformed into bringing together structured data, unstructured data and external data that we got from the Clearbit acquisition. And I would say that is huge differentiator for HubSpot and our architecture, because if you just start with unstructured data, you still have to go back and build. So if you're an AI-native startup starting with just e-mail or Gmail or calendar, for instance, you have to go back and build the structured data. So that's like number one. The second change that is happening is where you take actions. It used to be that we can take actions or we can -- our customers would take actions on hubs, go to marketing hub, create a campaign, go to a sales hub and close your deals. That was the way in which our customers were using it. Now with AI, we're also doing work for them. And the agents that we have launched over the last 1.5 years, we now have 15 plus agents that we have built or the ecosystem is building, those agents do work, and that is the transformation that we've already enabled within our platform, and that gives the flexibility of our customers for go-to-market employees working in hubs to get work done or using agents to do more work for them. So that's the second layer in terms of the transformation. And then we bring together all of this so that hybrid teams of humans and agents can work together. And that is kind of the orchestration layer. So what I would maybe suggest is that AI has enabled us to add more value by bringing in more data that we can help our customers with, more places to drive action and ability to orchestrate across hybrid teams. And I don't think this is one where there is fresh sheet of paper actually has an advantage. I would say the context that you have from, in our case, 19 years, 270,000 customers worth of context is much deeper and that provides the right kind of foundation for AI to drive effort at work.

Yes. There's a really interesting concept here around concept -- around context. And what I want to better understand is you obviously partner with an incredible wealth of companies that can essentially enrich the customer experience because they partner with you. And so how do you govern what works or what your partners have access to via APIs, understanding that the customer owns their data such that you maintain the advantage of having the context, but you also enable partners to the extent you need to keep the customers happy.

And you are particularly talking about our application ecosystem or are you talking about our LLM providers that we have built connectors and APIs.

Actually, we should do both, because it leads into the LLM competition question.

Yes. I think -- so there's a lot that is changing in the world of LLM. It's a new AI OS that is getting formed, just like you had mobile platforms that enable a set of applications and you had AWS, there have always been like these foundational platforms. We think that LLMs provide a new foundation for insights to show up. And my perspective is this, I think, AI and SaaS are complementary. And there are very key things that both bring together that add more value for our customers like HubSpot. The first is LLM can absolutely deliver insights. But without the context of the customer, without the context of your sales pipeline, without the context of the campaigns that you have run across multiple regions, it cannot deliver growth insights. HubSpot brings the context for growth. Every campaign that we have launched, every e-mail that has been sent, everything within the sales pipeline, every deal that is closed. Every CPQ transaction -- that is the context within HubSpot. And we're now enabling our customers to access that context and the insights from any location, including an LLM. So that's like number one. The second thing is that while LLM can take action on behalf of a single user, a sales rep wanting to write an e-mail, what HubSpot does is we maintain state across thousands of users whose roles are changing, whose permissions need to be monitored in a granular manner and whose task require the context of that particular team. That doesn't just auto generate by itself. And so HubSpot maintains that state across thousands of users. And are you a salesperson, what pipeline can you have access, what action can you take, that context of the user is within HubSpot, and it's very complementary to what an LLM can do for a single user. So that's the second thing. And then the other one I would say is that LLM are fantastic in terms of interactions and insights. We have the logic, HubSpot has the logic to take the action. So the way we see the complementary value of LLMs is that our customers can go and do a deep research. For example, on an ideal customer profile within a particular company. Once they've done the deep research if they need to launch multi-language, multi-country campaign that then reaches across 5 million contacts those actions. Those workflows, that logic and the ability to then drive leads, that is within HubSpot. And so I know there's a raging debate going on, but I -- we have a very clear point that these two are very complementary, and we bring the context across the domain, the user as well as the rich understanding that we have in terms of SMBs to help them actually drive growth. And we're actually super excited, because what we see is the opportunity to drive even more value for customers by building AI deeply into our platform and our hubs.

Yes. I think you already articulated this really well. You had Dario from Anthropic on with you last week, and he was actually explicit in saying that he's not going to target marketing as a domain. And so in your opinion, is that a function of all the domain knowledge in the context that you were just talking about? Is there anything else that you would add as a barrier to entry for a large language model to get into marketing as a domain?

I mean, I don't know if it is a barrier to entry, but I think that our business is completely different. Like we spent all the time thinking about what our marketing team, our sales team and the service team in a 1,000-person company or a 2,000-person company needs. Now -- and we built a business around it. We understand it deeply. We get that context. And we deliver that context. The example that I would give is an LLM can absolutely write an e-mail and generate an e-mail for a sales rep. But without knowing who is the buyer? What were the last 20 conversations? Who is the competitor? How do you handle objections for that competitor that e-mail is going to be generic -- and we have that type of context that then drives growth. And so I again go back to the perspective that there is a lot of complementary value that we can create. And you have the conversations with Anthropic CEO, Dario, like he said the same thing. There's a lot that the platform can enable, but not really in the areas that we are focused on, which is why I do believe it's complementary.

Yes, absolutely. Okay. I want to talk a little bit about go-to-market and then the growth out. So one of the things that HubSpot as a leadership team has been very consistent in saying, you want the barbell approach. You want the growing and scaling companies in the 200 to 2,000 employee cohort, but you also still really care about the small companies joining HubSpot for the first time. I'm curious if you've noticed any change in how those 2 cohorts, so the new early HubSpot customers versus the more established, perhaps the ones that you've displaced competitors on, how they're approaching some of the 15 AI agents that you're adopting. Does the education, does the go-to-market increasingly change between small companies and large companies?

I don't know if the adoption of AI is based on the size of the customer. I do think it comes from the level of AI fluency that they're trying to drive within the companies. And a lot of times, there's someone in the C-suite who is like pushing for AI. That is the most common denominator in terms of AI curious to AI leading, and we have a spectrum of customers all the way from AI curious to leading with AI. But talking about the two parts of the strategy, we do think both are important. We care a lot about customers who are just getting started. In fact, we offer a set of free tools for them to get started free. And over the last couple of years, we've consistently removed the friction for a very small business that is just starting to scale. We've lowered the price. We've removed seat minimums. We've made it easy for them to try by and grow with HubSpot. And by consistently doing that, we've seen the volume of additions from that continue to maintain. And that's our ongoing strategy. We want to make sure that we get started with our customers as early in their journey, and we continue to grow or help them grow and therefore, grow ourselves. And then on the upmarket side, we've had a very successful playbook for the last few years, and it's been a consistent driver of our own growth, and that starts with having a great product that is easy to use, one of the #1 reasons why we get picked over competitors is because we are easy while being powerful and then you get the engagement. And then we drive a level of adoption that typically you don't see in complex software -- and so our product has gotten better, the level of adoption from customers is better, and we have a partner ecosystem that has continued to grow and continue to scale upmarket customers in their ability to adopt HubSpot. And so both these parts of the strategy are just exceptionally important, and we have consistently proven that the playbook is working there.

And I know we've been talking now since 2021 on -- we've been through a period where the net retention rate at HubSpot has been under pressure because of the COVID over buying cycle. And so my question for you is, if we were to bring all of these product enhancements, you've had the pricing evolution, you've got the success that you're having going upmarket. When do you think that translates to a structurally higher net retention rate? Or maybe it's not a net retention, maybe it's a structurally higher growth rate. To put it simply, do you think that the company can accelerate growth? And how do you get there?

Yes, we absolutely think we can accelerate growth. In fact, at the Analyst Day last week, we shared how we have internally already reaccelerated our net new ARR. That is the forward-looking indicator in terms of growth. And what we have seen is a few things. One is that we have consistently been the platform that customers have consolidated on as they look at their tech stack, as they look at the level of complexity, if they want to lower cost and if they want to grow faster, they've consolidated, and that has been a trend over the past many quarters in terms of platform consolidation. That plus the upmarket and downmarket momentum that we talked about, both -- all 3 of those have been current levers for growth, where we think the additional new emerging levers for growth are the things that we started talking about, which is marketing is going through a fairly massive transformation -- and we now have a clear playbook for our customers to follow that includes our data hub, our marketing hub and content hub. So inherently, it is a multi-hub adoption story, and that's an emerging lever for growth. Very similar to that is sales. We just launched CPQ. And the combination of sales data and CPQ is a multi-hub driver for us to consistently grow -- in addition to that, for the last couple of years, we have walked you through our pricing changes and that nets to a consistent way in which we can drive more seat upgrades, both in terms of the sales service sub-seat as well the core seats. So that's an emerging lever. And we feel that especially the core seat for us is a huge opportunity that is AI, data and platform adoption. So that's a new emerging lever for us. And we've talked about AI -- and as we drive AI adoption within our customer base, ARR potential increases. So just to net it out, we clearly think that we have internally reaccelerated that's going to show up in external metrics. The way we have done that is by driving platform consolidation, consistent execution upmarket and downmarket. And now we have some emerging drivers of growth, which will be marketing sales, CPQ as well as the pricing changes that we made.

Yes. I want to stay on this concept of pricing for a few more minutes. I know how much time we spend with customers and making sure that HubSpot is delivering value to. So part and parcel with the AI adoption, you now have a move towards more value-oriented pricing or consumption-based. So what are some of the guiding factors that help you step the right price point for the usage-based pricing algorithm. And how do you execute on that layering into the model without compromising or cannibalizing.

Yes. A couple of different things there. So I would first start with our pricing philosophy. And we've consistently focused on driving value before we monetize. I know we've repeated this over and over. But the way we think about pricing is that we first need to identify use cases that deliver repeatable value for our customers. And when we do that, then we have high confidence. So we're strategically patient. And that approach of value before monetization has worked with us for a very long time, and you've tracked us for a while. Now in terms of our pricing model itself, it is hybrid. It is seats plus credit that we just launched -- we launched credits last year at INBOUND, and we have consistently added to it. Last year, we started with a couple of agents. This year, it now has customer agent. It now is prospecting agents. So all of the agent actions as well as the newly renamed Data Hub and the data things, they consume credits. And we think about it as if we are delivering AI value for a specific role within a hub, then that belongs to seat monetization. And if it is an agentic work that we are delivering or data that supports that agent that belongs to the credits. And so we think that the hybrid model will work. So you asked the last part of the question, how can you make sure that if seats gets compressed and then credit stake, by having a hybrid model, we do think that if we deliver more value with AI -- and then we are delivering more value. So we're going to be able to monetize it with credits. And we have not seen seat compression. But even when we do the hybrid model is going to help us make sure that we balance the value. But in any case, we are delivering more value. We're delivering more value from a seats perspective, and we're delivering value from AI and credit consumption perspective. And so I think overall, people always talk about [ PxQ ], we get like pretty fixated on [ PxQ ], but there is also a value equation. If price times value, if we are actually delivering more value than you're going to be able to monetize it. And that's why we have a hybrid pricing strategy.

Fantastic. Please join me in thanking Yamini for your time. Yamini thanks for your time.

Thank you for having me.