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

So please welcome to the stage, Geoff Koegler.

Good morning, everyone. Welcome to HubSpot's Analyst Day here at UNBOUND. A big thanks to all of you that are joining us live here in Boston. It's really great to see you all. And thank you to those of you that are joining us on the webcast as well. We really appreciate you taking the time to join us today. I'm Geoff Koegler, Vice President of Investor Relations here at HubSpot. And this is my first UNBOUND. I'm incredibly excited to be here with the team and all of you. Speaking of UNBOUND, I hope you had the chance yesterday to catch Yamini, Dunkin' and Jon's keynotes. They did a great job walking through the latest trends across AI, marketing, sales and service and highlighting the amazing innovation that we're driving across our platform, including some exciting new product announcements that we made this week. So if you missed those, make sure you catch the replays. Also make sure you catch the replay of Darmesh's keynote on how to win with AI by becoming a builder. It was a really great session. No Analyst Day would be complete without a safe harbor statement. You can find the full details on our Investor Relations website at ir.hubspot.com. We'll also be posting the full presentation to the website shortly after the event concludes. We've got a great agenda for you all this morning. Yamini, Duncan, Jon and Kate will join us up on stage. They're going to unpack our vision to win in the outcomes era. They'll walk you through where we're headed, our strategy across product, pricing and go-to-market and how all of this will drive long-term revenue growth and profitability for HubSpot. After that, we'll have Dharmesh join the team up on stage, and we'll have some time for Q&A. So get those questions ready. And then for those of you that are live with us here in Boston, we'll have a short window to connect and grab some lunch afterwards as well. And then finally, at 12:30, we'll all head back downstairs to the main stage session with Sunita Williams. All right. With that, let's get started by hearing directly from some of our customers. [Presentation]

Please welcome to the stage, Yamini Rangan.

Good morning, and welcome to Analyst Day. It is so great to be in Boston. And for those joining us here in person, it's great to see so many of you here. I look forward to the conversations that follows. And if you are joining us online. Thank you so much for the support, and thank you so much for joining us here. Welcome. All right. Today, let's get started by talking about HubSpot's vision and where we are headed. Now AI is creating a new era in software, not just defined by the features that we deliver, but by the measurable customer outcomes that we can drive. That's a big change. We are transforming every part of HubSpot in order to lead in this new era from our product to our go to market to our pricing, to how we operate as a company. And we're doing this from a position of strength. Our connected customer data as well as trusted platform and the broad expansive distribution that we have and the ecosystem moat. They're all give us durable advantages, and we're staying very disciplined, not choosing between growth and profitability. We are committed to driving both durable growth as well as expanding margin. Today, you're going to hear from the leaders in product, go-to-market as well as Kate, and we want to walk you through our strategy to accomplish all of this. And let's get started. Yesterday, I talked about the outcomes era. And the outcomes era is here. The last few years, we've seen one of the largest investment cycles in technology, trillions of dollars invested in building the foundation for AI. But the next wave of value is actually going to come from applications that take that intelligence and convert it into measurable customer outcomes and growth. That is the outcomes era. And HubSpot is built to win it. Now there's just been a lot of debate on what AI means for SaaS, who gets to capture the value and SaaSpocalypse. I'm just going to address that directly. We don't believe that AI eliminates the value of platforms and applications. We believe it does raise the bar for how you convert that intelligence into real customer outcomes. And to do that, you need context. So here is our thesis. Our thesis is that intelligence is necessary, but it is not sufficient. Context is what turns that intelligence into measurable customer outcomes within businesses. Now AI models are trained on vast amounts of publicly available information. and they're going to continue to get better. Absolutely no doubt about that. But no matter how intelligent model gets, it does not know what's happening within the business, what's changing in the pipeline today, which deal is stalled, which customer is at risk. And all of that information is not just a static dumb database. It is a dynamic smart substrate with workflow logic that maintain state rules, permissions, governance and are it trails that make all of that information trustworthy for businesses. That is the context that HubSpot delivers. Now this is not just our belief. We actually quantified the value of context. We went across our customers that are using AI, and we measured how context impacts the outcomes. And this is what we found. We actually found that good contacts delivers great outcomes. We saw this with every one of the key metrics in marketing, sales and service, improve with good context across our customers. You can see that, MQLs generated up over 200%. Deals closed in one up, the number of customer meetings booked up. It takes a lot to actually move these metrics, but good context actually drives these types of outcomes. But here's the part that actually really surprised us. We've found that AI with bad context is worse than having no AI at all. This really surprised us. And you can again see every one of the metrics here, deteriorate with bad data and bad context. Now this is because I think as humans, we know how to control for this bad context. But AI does not, right? It might become in a product that is no longer in your product catalog, may actually use an inaccurate definition of ARR, my route and approval to a manager who just changed roles two days back. And that is why we believe that even as models improve pretty significantly the context within a business becomes even more valuable. And this is HubSpot's structural advantage. And this is why we are democratizing AI across go-to-market for scaling companies. Now for just over 20 years, HubSpot has done one thing consistently. We have excelled a democratizing complex technology to bring it to scaling companies to help them grow. And in that process, we've become domain experts in both go-to-market as well as SMBs. In Phase I, we democratized INBOUND. We helped tens of thousands of customers by giving them a playbook and a set of products to help them succeed in marketing. We created an all-new category called INBOUND marketing. In Phase II, we democratize CRM. We help hundreds of thousands of customers by giving them a unified platform with data across marketing, sales and service, and we made it easy to use. We went from a single hub company to a platform company with multiple hubs, and we crafted every hub organically. Now we have the opportunity to democratize AI. Our goal is to help millions of customers grow by providing them an agentic customer platform and making it easy to get the outcomes that they want with AI. This is our biggest opportunity yet as a company, and we're bringing all of our domain expertise in go-to-market and within SMB segment to make this possible. Now we're doing this from a position of strength. We have consistently grown our revenue. We've generated strong margins as well as free cash flow over the past many years, and we'll continue to build on that solid foundation. We also have a platform advantage that has allowed us to really bring the pace of innovation to our customers, and it has enabled us to grow our customer base to over 300,000 globally and drive multiproduct adoption across that customer base. And finally, we have exceptional distribution. We talk to millions of growth leaders every single month across our channels, across social, across academy, across the events that we do, and we have a powerful ecosystem mode. We have nearly 4,000 solution partners that extend the reach of our sales as well as customer success teams. Our execution really helped fuel the growth and now is the foundation for the next phase of our growth. So the question becomes what's happening with customers? How are they adopting AI? The landscape with customers is changing pretty dramatically. We've now, over the last year, talked to hundreds of customers looked at how they're adopting AI and three big shifts actually stand out. The first is customers want to build. AI is democratizing building, and that's the fundamental shift. For many decades, businesses just had to adapt the way that they work in order to fit the software that they had. Now AI, thankfully, is reversing that. And that means if you're a marketer, salesperson, service person, you can actually build. You can build artifacts, you can automate workflows. You can create new experiences all using natural language, no need to wait for weeks behind the developer sprint. But there is a clear way to build well and building well actually requires connected customer data. It requires growth context, permissions, governance, security, and that is the platform that HubSpot is building to deliver. We want to be the platform where go-to-market teams can build. Next, what we're finding is that customers care deeply about outcomes. Now again, the conversation is changing really rapidly. A few months ago, it was all about token maxing. Now our customers don't talk to us about token maxing. They want and care deeply about the outcomes that they want to drive within their go-to-market teams, and we love that because the outcomes within go-to-market are just incredibly clear. We help our customers build demand, win deals and delight their end customers. That is the focus for us. And third, I'd say that customers want proof of value as well as predictability in cost as they adopt AI, a huge shift again that we're seeing. In SaaS, the typical process was you evaluate features. You then implement the software, you drive change management and user training, and once the users begin to drive, they get the growth and the value out of software. But AI is changing that dramatically. You can just turn up a trial and look at what the value you get and the outcomes that you get within your environment. And that is -- means that there is need for proof of value before the commitment to software. And these three shifts combined is really changing what customers expect out of HubSpot right now. And this is exactly what our product strategy as well as our go-to-market strategy are designed to deliver. Now as we look ahead, we have a very clear strategy for AI transformation. We're not incrementally improving our business. We are transforming it across product, go-to-market, pricing and how we operate as a company. And this is what it means. It means that from a product perspective, we are moving from delivering features to delivering measurable customer outcomes. It means in go-to-market, we're not just selling software, we are proving value, and we are earning the right to expand and compound adoption from there. In terms of pricing, it's not about charging for access, it is clearly aligning to how customers get value and outcomes from HubSpot. And then as a company, we are scaling with people and AI. We are driving this full transformation from how we build to how we sell, to how we monetize to how we operate as a company. Very exciting transformation. Now Duncan and Jon are going to do a deep dive in terms of our product strategy as well as go-to-market strategy. So I want to take a few minutes to provide an overview across all of these four areas of transformation. Starting with our product. Our product strategy is to lead with outcomes. And that means every product decision now starts with a super simple question, can we deliver trusted, measurable outcomes. And can we do that fast? We have three clear focus areas that you're going to hear about today that will accomplish this strategy. First is that we are building an intelligence layer with trusted context. We're bringing together all of the growth contacts for customers in what you will see as a custom contact home. And this context home is what will provide our customers the visibility, the control and the trust that is needed in all of the information that powers every one of the AI and agentic use cases, really powerful. We believe this context is what will make every agent, every recommendation we make and every outcome that we deliver compound over time. Then we want to drive measurable outcomes within marketing sales and service using HubSpot agents and custom agents, two ways. Now HubSpot agents are domain-specific agents that we build. And at UNBOUND, we launched four new agents, campaign agent, content agent, nurture agent and revenue agent. You're going to see this in action today. And over time, we'll continue to expand the number of use cases that we support as well as the depth of outcomes that we deliver through HubSpot agents. At the same time, we are making it super easy for our customers to build on top of HubSpot, build workloads as well as their own custom agents. And the combination of HubSpot agents and custom agents is what will drive clear measurable outcomes within go-to-market. And then finally, really excited about the all-new breeze assistant. This is the interaction layer. And our vision is that customers shouldn't have to navigate the screens, click on menus and learn software. They can have powerful conversation with all of the value within HubSpot. We've completely rebuilt Breeze Assistant. It's no longer just a chat spot. You can ask questions, you can build workflows, you can build custom agents, you can build artifacts. This is why we are super excited about Breeze Assistant adoption. Now that adoption has grown pretty significantly this year. The beginning of the year was 37%. And that at the end of August, it was over 55%. Over 55% of our Pro Plus customers are using Breeze Assistant. That is great. And at the same time, customer satisfaction of Breeze Assistant has actually improved from 54% at the beginning of the year to nearly 68%. We are very excited about the progress that we're making in this powerful interaction layer. Now if you step back and you sum it up, we are the trusted context layer as well as agentic layer. And this is how we are transforming our product from delivering features to really delivering clear outcomes for our customers in this agentic era. Now product transformation is great. It is absolutely the start. But in order to capture the full opportunity with AI, we are transforming how we go to market, and how we meet customers where exactly they are. Customers are buying AI very differently than they bought SaaS. There's a fundamental shift that's happening here as well. They don't start by comparing featureless. They start by asking a simpler question, will this work for my business. And they want to see it work in their environment with their data. And when they see that, they want to scale with predictable costs that are transparent to them. And that changes how we think about going to market. We will land with focused use cases. We will drive outcomes very quickly and then continue to earn the right to expand from there. And to do that, we are actually strategically leveraging our distribution advantage. We have millions of users who use our product every single day. We have hundreds of thousands of customers, and we have scaled go-to-market motion. We're using that reach to seed the right AI use cases with the best trial experiences the best product experiences and very targeted pricing offers. The first use case, the agentic use case is strategically important to us because what we find is that when customers get measurable outcomes in the first use case, the second, third, fourth use case becomes just easy for them to adopt. Now if you take a step back, SMBs are very early innings of adopting AI. And that is our biggest opportunity. So our strategy is to seed the market is to prove value quickly and then compound adoption from there. And to do that, we're also leveraging the strength of our partner ecosystem. Our partner ecosystem is our forward deployed moat. Now delivering outcomes within AI is not turnkey. It can't get results just by simply turning an agent on. You got to look at the data foundation and the context foundation that you're building, redesign the workflows, reimagine how a business operates. And that is exactly where our partner ecosystem comes in. And this is one of HubSpot's biggest structural advantages. We have more than 4,000 solution partners, and they work with our customers every day to implement HubSpot, but also transform the go-to-market. They actually drive 25% of our sourced ARR. They help us close nearly 50% of our ARR and actively manage about 40% of our entire customer base. So we are investing in this partner ecosystem by providing the platform and the tools so that they can custom build on top of HubSpot by providing the right incentives to drive that motion and also making sure that they're taking us into every industry vertical that they have the expertise in. We believe that companies that will win in AI will have a very strong ecosystem, and that's why we are continuing to build on the moat that we have here. Okay. With that, let's actually talk about pricing. Now our pricing has also evolved alongside our platform. In Phase I, we built inbound marketing, and customers bought one hub, Marketing Hub. We had subscription pricing back then with an [indiscernible] contact tiers. Life was just simple. People knew exactly what to buy. Now in Phase II, as we expanded into a broader CRM platform with multiple hubs, some had seat space pricing and the others had subscription-based pricing. But now we are entering Phase III as we democratize AI and deliver an genic customer platform. Customers are not just simply buying software anymore. They're actually buying outcomes from the software which means our pricing needs to continue to evolve to meet where our customers are in AI. So our vision for this next stage of growth is simple. Hybrid model of seats and credits. Seats provides access to platform, data and AI capabilities and credits, they scale with the value and the outcomes that we deliver. And we want to make HubSpot easy to buy and easy to grow with. And we've been delivering against this vision. Many of you know that we started by simplifying our seats model back in 2024. And specifically, we did three things. We lowered the seat price. We removed the seat minimums that we had for Pro as well as enterprise, and we introduced a new seat called the core seat to capture, the value of the platform. And we saw clear results. We were able to add more customers. Those customers are healthier because they buy exactly what they want and continue to upgrade from there. And we were able to capture the value of the core seat through the platform. Now more than 25% of our Pro Plus customers have continued to add core seats. That's up 10% year-over-year. And at the same time, Pro Plus customers have continued to buy more Sales Hub and Service Hub seats. In fact, the seat count has grown by 16% year-over-year and seat ARR has grown 23%. That's what we are seeing in terms of the seat adoption. And now we are -- as we look ahead, we want to be even more clear about our focus. We want to remove barriers to AI adoption. That is the focus in terms of how we are approaching pricing. And in order to do that, we got to drive predictability in terms of the cost as well as flexibility in terms of driving the right adoption. This is a big change for application buyers, right? And so in terms of predictability, we are tying our pricing as close as possible to the value that we deliver as well as the outcomes that we deliver. And at the same time, we're providing visibility through dashboards and the ability to control spend our customers can set up thresholds right now. And in terms of the flexibility, we're making it super easy to swap between seats and credits. Super easy to buy in chance for credits. And we are continuing to test and iterate and learn within this model. For example, we have pricing test and experiment that was running in Nordics that makes it super easy to buy seats and credits, and we will continue to expand that pilot within just the EMEA market in Q4. And as we test and learn, we'll continue to share the progress with you. Over time, we expect seats and credits to become the primary source of our growth drivers. Speaking of which, let's actually talk about our growth model. Our core growth levers that have powered HubSpot over the last many years remain solid and consistent. We're continuing to drive up market expansion and the number of large deals continues to grow. We will deepen our platform consolidation as well as multi-hub adoption, and we will land and drive volume in down market. We plan to do that by focusing on single use cases that are powerful drive value quickly and then continue to expand to multiple use cases from there. As we look ahead, we have very clear emerging levers for AI monetization. Seats monetizes platform and AI value when we assist the work that people in go-to-market teams that are doing. And credits will monetize the AI value when we deliver work. And as I talked about, we think about it as HubSpot agents consuming credits, custom agents consuming credits as well as the APIs that we deliver consuming credits. So if we put this all together, customers grow with HubSpot. And as the value that we deliver with AI continues to grow, we expect our monetization to continue to grow. This is our durable model for growth. Okay. Let's finish by talking about how we are transforming as a company. AI is not just creating a new growth opportunity. It's really changing how we operate as an organization. And we've embedded AI into how we grow how we build and how we operate and transform ourselves into being almost AI native with the speed of that. Our product organization is just completely all in on AI. And they are one of the most advanced in terms of using our own agentic harness, something you'll hear Duncan talk about in a few minutes. Now nearly all of our developers are using agent coding. And as they do that, the lines per developer has increased more than 123%, and you'll see the interactions within our internal harness also continue to improve. In terms of go-to-market, we are an AI-first go-to-market organization. From a marketing perspective, AI handles almost all of our INBOUND chats. And from a sales perspective, prospecting agents help us set up the meetings as well as continue to help our teams drive more win rate. And in support, AI has enabled us to scale customer support without scaling the headcount. And as an organization, we have completely transformed. Everything happens in 6 weeks prints with continuous experimentation. All of this means that we are committed to driving durable growth as well as profitability, not by choosing between one or the other, but by really transforming how we operate as a company. Okay. To bring this all together, we know that the outcomes era will really be on by companies that can take that intelligence from AI and convert it into clear measurable outcomes for our customers. And we have clear structural advantages to win, a trusted platform with the right context that can drive customer outcomes, distribution that can accelerate adoption, an ecosystem moat that helps our customers transform with AI and an AI-first operating model that helps us balance growth as well as profitability. And the combination of these structural advantages as well as the deliberate transformation that we are doing is what will drive durable growth as well as long-term shareholder value. Thank you so much for your support, and I look forward to your questions during Q&A. [Presentation]

Please welcome to the stage, Duncan Lennox.

Hi, everyone. It's great to be here with you this morning. Yamini just made one thing very clear. AI needs context to deliver outcomes, and that's where I want to pick it up. My job today is to show you what context-driven AI looks like inside our product and how we're building HubSpot to democratize AI for scaling businesses. And it starts with our vision for the outcomes era. Build the best agentic go-to-market platform that delivers an easy, fast and trusted experience that customers love. And to do this, over the last year, we've rebuilt much of our platform from the ground up. We've reinvented the CRM for the outcomes era. It's still the beating heart for go-to-market teams, but it needs to drive outcomes through people and agents, not just be a system of record for tools and workflows. So we added a brain to make it all work seamlessly together. And let me show you what our new platform looks like, starting with how we've rebuilt the CRM. The CRM is really three things: It's data, trust and intelligence. As always, it has to start with unified data at its core bringing structured and unstructured data together. And this, of course, is what customers have always relied on HubSpot 4. Next, we built in security, governance, observability and quality, which is foundational to trust, which, of course, is more important now than ever, especially as agents and AI are taking action on someone's behalf. And finally, we've given the Smart CRM, a new brain with two key components: our agent harness that we call Aviator and the growth context graph. And this allows us to give customers what they've always wanted a CRM that updates itself. On top of this foundation is the action layer. This is where go-to-market outcomes across marketing sales and service are delivered. Whether that's through the dedicated apps that HubSpot has always been known for like Marketing Hub and Sales Hub or the agents, both those that we build ourselves at HubSpot and custom agents that customers and partners are building on our platform. They all run on the aviation harness in the smart CRM. And then there's Breeze Assistant as the interaction layer, but this is not the Breeze that you may know. We've completely reimagined it as a go-to-market expert that does the work for you. It figures out the best way to achieve your outcome, choosing the right tools and agents to get the job done. And this fundamentally changes how you work in HubSpot, bringing all of the power of agentic AI to users to drive the highest quality, best value, fastest outcomes all without them needing to know anything about AI. And when we talk about democratizing AI, this is exactly what we mean. And finally, we want customers to be able to get the best go-to-market outcomes from HubSpot regardless of how they choose to engage with our platform. So we'll get into all of the details about how we optimize for agents running on HubSpot and agents running HubSpot as part of our headless strategy. Okay. Let's start with the look under the hood of the smart CRM. Now I want to make an important distinction here. The CRM is not simply a system of record. It's a system of context. The intelligence layer for your go-to-market. That's the bet that we've made on CRM for the AI era, and we rebuilt it with two defining elements. First, our new agent harness called Aviator, and secondly, rich growth context built in, and I want to dive into both of these. Let's talk about Aviator, our agentic harness. Now a lot of people are throwing around this term at the moment, but few people are talking about what it actually takes for it to truly work. We've thought deeply about this which is why we rebuilt the CRM with the agent harness instead of bolting it on to an old system that wasn't built for AI outcomes. So I want to talk about what it is, and how ours is different. At its simplest, an agent harness is designed to let you run an execution loop calling an LLM to do work repeatedly until you get the desired result. But in practice, that's a pretty inefficient mechanism and deliver slower and more expensive results that often miss the mark. Our harness is deeply integrated into the heart of the CRM with quality, trust and safety and governance. And how we build Aviator is key to our ability to deliver the best go-to-market outcomes. First, we made it model agnostic. We think of this as taking the Switzerland of AI approach. Not every model is good at everything. And it changes on an almost weekly basis. So we're highly opinionated on which should be used for which task. Next, we use our domain expertise in go-to-market to build evals that run continuously to pick the right model. That means we know what great looks like. But our customers don't need to know or care about any of this. They can just be assured, HubSpot will deliver the best outcomes for them even as the state of models and AI continues to evolve rapidly. The second part of our unique strategy comes down to context. In the outcomes era, good data is necessary, essential in fact, but it's not sufficient. Building on our unified data model of structured and unstructured data, we developed the growth context graph. It computes insights and intelligence from across a customer's unified data. This intelligence then allows us to use precise high-quality context in real time for a given task or outcome. And growth context has three dimensions that work together. The first is business context what you do, everything that makes your company yours, your brand and voice, your product and positioning. Next is team context, who you are how you and your team work and win together, your roles and goals, your workflows and processes, your methodologies and habits. And finally, of course, customer context, who you serve the full history and nuance of every customer relationship, your ideal customer profile and personas, your communications contract and revenue details. A point solution might have one part of this, but that's why they hit a ceiling. But in HubSpot, all three live together, and now that's in the smart CRM. The Aviator harness uses growth context to provide the right context for a specific task at the exact moment that it's needed. And this is so important because the wrong context adds noise and cost which is how businesses find themselves wasting money on tokens with minimal ROI. Our unique combination allows us to deliver the highest quality at the lowest cost. And if at any time, the best outcome favors another model, we route accordingly. Okay. So we picked the right model. We inject the right context or without customers having to do anything. And it's all built into the trust layer of the Smart CRM, which has security, including things like role-based access, governance for things like permissions and approval workflows, observability for full visibility into what every agent did and why and quality, including safety and evolves. This is what makes us the go-to-market harness for scaling businesses. Now I want to show you how customers experience all of this in HubSpot. Yesterday on stage, I unveiled our all-new AI-native CRM. It's the CRM that you know and love now rebuilt for the outcomes era. This is because our agent harness now orchestrates and manages action, which changes everything. This new CRM works on our customers' behalf and update itself. This will save customers hours of manual updating, the headache for getting to add details. But most importantly, it will give them accurate insights at the moment they need them. And those insights are stronger and more proactive than ever because we build growth context directly into the Smart CRM. But we didn't stop there. businesses need to be able to see and control their growth context to understand what it's powering, and how it's delivering outcomes. So we build context home as their one-stop shop. This is a new capability that lets customers see everything HubSpot knows about their business and manage it directly. It also gives customers a score on how complete their context is along with recommendations for where to add additional context that will let us deliver even better outcomes for them. All they need to do is upload a file or point us out of source and HubSpot do the rest automatically. And all of this context is available to every agent and every AI feature across HubSpot, the moment it's relevant. From our own HubSpot agents to custom agents, customers build on our platform with agent builder. And I want to show you how this comes to life for customers and something that they use every day, e-mail. This is what a generic e-mail written by AI that has no context might sound like. We've all seen e-mails like this. It's AI slap. Now this is what happens when AI has the complete precise context it needs. And let me point out the pieces of growth context that went into this e-mail. First, from business context, we knew the product that was trialed, how it's positioned against the competitor and even the Gartner comparison to show that. From team context, we knew how the team typically follows up after a trial, the reps unique voice and tone and even that UNBOUND was the right moment to push for a meeting. And from customer context, we knew this customer-specific usability and price concerns. The team member that they should loop in for next steps and the per se pricing was particularly relevant to their company size. Every sentence in that e-mail earned its place and helps our customers drive better outcomes. And the user didn't have to foot around with figuring out what docs or other contexts to upload or what prompt to write or which agent to use. So let me recap what we've covered. We've rebuilt the Smart CRM as the foundation that makes everything else work. Our agent harness delivers high performance at a lower cost growth context delivers accuracy and control. This is the intelligence layer, the new brain of HubSpot. Now let's see how it powers action. When it comes to driving outcomes, agents are the most effective way to scale how businesses build demand, win deals and delight customers. In the past six months alone, we built several new agents and tools in HubSpot aimed at driving these exact outcomes, and I'll get to those in a minute. But first, I want to talk about another foundational update we've made to HubSpot how our customers actually use those agents. Yesterday, I introduced the all-new Breeze Assistant. Many of you know this tool. It's been on HubSpot since we first started building with AI. But what we unveiled is a complete overhaul. It might look similar, but just like the all-new CRM, it's been rebuilt for the outcomes era. You now get work done just by talking to Breeze. Instead of navigating menus or learning workflows or even having to understand the capabilities of our agents, customers just describe what they're trying to accomplish, and Breeze takes action across the platform on their behalf pulling in the right tools for the job. Customers tell us that they don't have time to become AI experts and Breeze Assistant means they don't have to. Now I know many of you watched the Spotlight from yesterday, and I definitely encourage you to do that if you haven't because there's some really exciting stuff in there. But I want to highlight some of the new tools and agents we launched, and how our customers use Breeze Assistant to drive outcomes. For marketers, the outcome is building demand, but this has never been harder. Organic reach is declining, buyers are asking AI instead of filling out forms, all while the volume of content needed to cut through the noise has gone up, and ROI has gone down. Marketers need a dramatically new way to build demand, so that's what we created. Our all-new marketing studio is their single home where they can just have a conversation with Breeze and then go from insight to a live campaign in minutes. To start, they just asked Breeze for a full campaign plan. It pulls in our new campaign agent to builder, everything from messaging angle to an asset plan and audience segmentation. Next, we've made it easy for marketers to create their campaign content just by asking Breeze. It calls on the new content agent, which takes the campaign plan and draft the assets in the correct brand voice and targeted to the ICP. And when marketers then have their campaign up and running and need to nurture the leads, they just tell breeze. It pulls in our new nurture agent to write automated e-mails for every contact using CRM data and engagement history. With the new home and marketing studio and Breeze Assistant as their guide, we're helping marketers build demand all from a single conversation. Now let's see a sales journey where the outcome is, of course, winning deals. We know reps are spending more time than ever on admin and less time actually selling. Deal stole, follow-ups get missed and context doesn't carry between conversations. So we're giving sales teams valuable time back to win deals, and they start just like our marketers by talking to Breeze Assistant. Let's say they want to break into a big prospect. They just tell Breeze and it calls on prospecting agent. Immediately, it pulls the buying committee and draft personalized e-mails for each person. We already saw the difference growth context makes in those e-mails. Next, when the reps meet with prospects, they want to capture the details from the meeting. They can use our note taker to dictate updates, so nothing gets lost. And then deal progression feeds those updates back to the CRM automatically. So deals keep moving and reps keep selling. No other platform makes it this easy to prospect, research, coordinate, update the CRM and land the deal. So just to summarize what all of this means for customers is a platform that works for them, a go-to-market harness that selects the right model and context, a CRM that updates itself and delivers insights and a single tool in Breeze Assistant that launches agents just by chatting. Together, these democratize AI for scaling businesses because they take the burden away from teams so they can focus on their customers. Now in a world where customers are increasingly using agents, we're very aware of a different customer challenge and that's agents brawl. Leaders are losing sight of what's running, what it's doing and whether it's working. This leads to difficulty managing the agents, higher TCO and a poor customer experience, all for agents that don't actually work together. We believe there's a better way and customers will want a single view into it all. We built Agent Hub to help solve that. It's one place to see, manage and build agents across the entire go-to-market and they all operate from that shared context, which a collection of point agents can't do. agent Hub is where customers gain confidence that their HubSpot agents and custom agents are actually delivering the outcomes their business needs. We're also seeing a new kind of go-to-market team member emerge, the builder. These are people who want to quickly build agents that extend what they can do on HubSpot whether it's as simple as qualifying every lead before a rep touches it or as sophisticated as monitoring your entire pipeline and flagging at-risk deals before anyone has to ask. That's exactly what we're making possible with our new agent builder. This lets customers and partners build agents that run on Aviator and use growth context just like our own HubSpot agents. And you can build all of it just by chatting with Breeze Assistant. Our last layer is all about extending the way customers and partners interact with our platform. And I want to spend some time here because I think it's one of the most important things HubSpot has built. There's two parts to our ecosystem strategy. Agents can run HubSpot using our data and intelligence as a building block. Think of custom agents businesses can build on our platform. And agents can run HubSpot operating the platform end to end through APIs without a human ever logging in. This is what most people refer to as a headless CRM. We're unique because we think about our API in layers. First is the data APIs. This is raw access to everything in the CRM, over 1,000 public endpoints covering contacts, deals, sequences, content. And we are committed to giving API parity with everything you can do in the UI. We want developers building on us, not hitting walls. The second are our context APIs. Instead of raw data, you get insight that we've already computed. Think of it as asking HubSpot a question and getting an answer instead of a spreadsheet. This is actually the same layer our own HubSpot agents run on. That's how good it is. And the third and most important, our work APIs. This calls either a HubSpot agent or a custom agent to drive an outcome end-to-end. Let me give you an example of how this works. Imagine you're a rep, and you want to know which deals in your pipeline are at risk. With the data API, you'd pull all your deal data, bring it into an external third-party tool and figure it out for yourself. That's a lot of data and a lot of work. You're also going to burn a lot of tokens. With the context API, you say, give me the 10 most at-risk deals in my pipeline, and you get that insight back directly. And with the work API, you can say, figure out which deals are at risk and give me an action plan. You'll get the 10 deals and a unique action plan, next best action for each one. We believe deeply in our agentic ecosystem strategy, and we've been executing on it for the last 18 months. Over a year ago, HubSpot became the first CRM to support MCP. So agents can interact natively with our software, and we did this first with Anthropic. Then in June 2025, we were the first CRM to launch a deep research connector with ChatGPT. In July of that same year, we became the first CRM connector with Claude. And in May 2026, we shipped the agent CLI, a completely new layer that lets agents automate bulk scheduled work on HubSpot with no human in the loop. By mid-2026, we were the first CRM to have connectors with all four leading frontier models, adding Gemini and CoPilot. And today, we're the #1 CRM connector on both Claude and ChatGPT with weekly active users on ChatGPT up 250% since June. And yesterday on stage, I announced that HubSpot is the first CRM to integrate with ChatGPT ads. And we launched the AI growth bundle that gives new customers 65% of HubSpot Starter and HubSpot Credits, a free ChatGPT business seat when they buy one and up to $750 spend match on ChatGPT ads. By connecting to the deep context in HubSpot, ads become more powerful and customers get attribution that's only possible by integrating their CRM. This is what real momentum looks like. These aren't just announcements. These are real products with real users and real growth. Now I want to wrap it up. We've heard a lot today but I want to actually talk about what actually makes HubSpot different. AI that runs on a foundation rebuilt with the right harness, the right context and the right governance. And it has a compounding effect. Every action makes the next one better for a customer. That's what Aviator and growth context give us, and no one else has that foundation. Second, we made it easy for customers. Most agentic platforms give you power and then ask you to go figure it out for yourself. Breeze Assistant means customers don't have to. They just describe what they want. Agent Hub means they're not losing track of what's running. And the CRM updates itself, so go-to-market information is always relevant. We're not just making AI available. We're making it usable and that's what it takes to democratize AI at scale. Third, we don't believe in walled gardens. We believe in customer choice and meeting them where they are, and we've used these principles to build a 3-tier API strategy. And finally, trust, which is at the core of everything we do. Every company building AI right now is learning that customers are nervous. Businesses want to know what's running, what it has access to, what it did. Trust is about transparency and control, and that's what we've built into every layer of the agentic customer platform. from Aviator and growth context in the CRM to Breeze Assistant and our own HubSpot agents. So customers can trust us as their go-to-market harness. And here's where all of these things we've talked about add up, the outcomes. Customers using HubSpot with rich growth context across the platform are seeing real results from AI today. They're seeing demand built with nearly 4x more MQLs. They're winning deals at over 3x the rate, and they're delighting customers with twice as many tickets closed. We set out to build the best-in-class agented customer platform that drives go-to-market outcomes that customers love. We believe that what we've shown you today and introduced to our customers this week not only meets the moment, but gives businesses the platform they need to grow well into the future. Thank you so much. [Presentation]

Please welcome to stage, Jon Dick.

Thank you, Duncan. So great to meet you all. How are you? Nice to see you. Thanks for coming to UNBOUND. I'm Jon. I'm the Chief Customer Officer at HubSpot. I lead sales and customer success. But I'm a long-time HubSpotter. I've been at HubSpot for about 10 years now and have worked across pretty much every part of our go-to-market. Most recently, I led our customer success team. And leading that team, I had the opportunity to get to know our customers deeply, their goals, their challenges and their hopes when they choose HubSpot. And I think that our customers are going to absolutely love the HubSpot that we're building for the outcomes era. Not only do I think that the product that Duncan just talked to you about is going to absolutely blow them away but the transformation of the business model will align HubSpot's growth with our customers' growth in a way that was just never possible before. It's good for customers, and it is good for us. So my mission as a Chief Customer Officer is to build a go-to-market that meets the moment for the outcomes era. It means democratizing AI, delivering exceptional results for customers and ensuring durable growth in the process. And that's what I'm going to talk to you about today, how our customers are changing, how our go-to-market is transforming, and how we're going to drive growth, both down market and upmarket. So let me just start with the customer. In the SaaS era, the customer journey was very well established. You've seen it before, it looked something like this. And it was a journey that was built on establishing upfront promise. What can the solution possibly deliver for my business. This has all changed with AI. In the outcomes era, the customer journey is built on upfront proof. What will this solution deliver for my business. And I think that new journey looks more like this. Let me just talk you through it at a very high level. So the customer journey used to begin when buyers started doing research on their own. They found a whole bunch of solutions for their problem. In the new journey, buyers are outsourcing this. They're outsourcing it to LLMs and to influencers. And so the whole thing looks very different. They're being given a short list of solutions to consider for their problem, and they're just showing up much more ready to buy. But buying also looks very different now. Buyers are no longer interested in spending a tremendous amount of time upfront evaluating if a solution works for their business. Instead, they want to jump right in. They want to start building for their business. They don't want to demo from a vendor. They want the vendor to help them actually build their first use case on the product in their environment. And that first use case then becomes the business case. Instead of developing a theoretical ROI and committing based on promised value, buyers are able to look at actual proof and have confidence in the outcomes that the solution will provide. And with that confidence, they grow, and their spend grows too, not only by spending more in that first use cases at scales, but also by adding other use cases because in this new customer journey, one use case leads to the next. So this new customer journey has really inspired us at HubSpot, and it's caused us to really obsess about two big things in our go-to-market. The first is how we drive distribution so that more companies consider HubSpot on day 1 and continue to consider us on day 100. Second is how we engage customers, to help them actually build that first use case and see value immediately. Let me start with distribution. So distribution has been an advantage for HubSpot since our founding, and it continues to be a real cornerstone of our go-to-market strategy. The big shift, of course, is that buyers increasingly rely on more trusted sources to make purchase decisions, especially when it comes to AI products. They bypass the traditional channels. And instead, they looked at the LLM, the influencers communities to learn what they should be buying. And we've been ahead of this shift at HubSpot, and we've been diversifying into these channels for several years now. Our YouTube and Newsletter reach, for instance, beat major business media publications. And you can see that with the live stream of our Spotlight from yesterday, we're already at 1.65 million views on YouTube, well ahead of other keynotes that have been posted. Meanwhile, our global aided awareness for HubSpot upmarket across core geos and segments is at all-time highs, and we continue to be #1 in AI share of voice for the CRM category across all of our major markets. So this is our playbook, and we are continuing to invest in this. But distribution, of course, isn't only about new customers because the other place that buyers look are to the companies that they already work with, and they already trust. With over 300,000 customers, of course, we have a substantial distribution advantage and a trust advantage within the SMB space. And as you know, we've been successfully leveraging this advantage when it comes to upgrade and multi-hub expansion for some time, and that continues to go well. But we're also now focused on leveraging our existing customer base for the emerging AI adoption opportunity. We are actively working to get as many of our customers building their first AI use cases on HubSpot, and we're making really good progress with that. 19% of Pro Plus customers are using HubSpot agents as of August, which is up 2x since the beginning of the year. So we're working with them actively. And so this brings me to how we are doing that. So we are transforming customer engagement at HubSpot. And we're putting our focus on helping build that initial use case, helping customers achieve value and then growing from there. What this looks like in practice is fewer handoffs and more help. To do that, the lines between sales and customer success really start to blend in a whole new way. That's why we combine the sales and the customer success team under my leadership earlier this year, why we've aligned everybody in sales and customer success on customer outcomes, and why we are rebuilding our motions with customers. But we, of course, are not the only ones who help our customers at HubSpot. Our engagement model also includes our partners who are really a extension of our sales and customer success teams. They act as HubSpot's forward deployed team. Our partners are a major advantage for HubSpot, and we are leveraging them to help our customers adopt AI. Partners have the industry experience, the technical expertise and the change management resources that our customers need in this moment. So accordingly, we are doing a lot more with our partners. We are delivering more co-selling, we are doing more co-delivery. We are rewarding them for their efforts by providing incentives for customer AI adoption, and it's paying off. When a partner is involved in an upmarket deal, we see nearly an 8-point higher win rate, 75% higher ASP and a 14-point lift in net revenue retention. So as you can see, when partners are involved in selling and expanding, it really works for driving value for our customers. So we are going to continue to do that. So now let's actually talk more about our customers. And I want to start with the group that's been the foundation of the HubSpot story for our entire 20 years, which is small businesses. So what we're seeing right now down market is that customers have more options than ever, but the same challenges. They're more price-sensitive, they're time constrained. They just need easy ways to grow. And that is what we have always delivered at HubSpot. And I think by what you saw from Duncan, what we will continue to deliver because we've built them an even easier way to grow in the age of AI. I mean no more data entry, no more managing omnichannel complexity, no more learning the software. I've been using HubSpot's mobile notetaker on the HubSpot app all week in meeting with customers, and it's been absolutely great. They can build and scale on HubSpot without friction now. And so this gives us an opportunity to reintroduce HubSpot to the downmarket customers in a whole new way. And so the first thing we're doing is we're positioning ourselves differently. We are obsessing much more about the outcomes that we drive for customers than the technology that drives it. That's what our customers care about. They care about building domain, they care about winning deals, they care about automating all the stuff they hate. Second, we are ramping up our promotion in Q4, and we are pursuing new partnerships and helping more customers see this whole new and improved AI-native HubSpot. And third, we're being more flexible with our customers. We are letting them shift their spend between seats and credits just so it's easier to build, prove and grow. So better product, better positioned with the pricing flexibility that customers expect. Now let me tell you two stories. I want to tell you two stories of downmarket customers who built and quickly saw value at HubSpot. The first is Fin. There is security awareness in fishing simulation platform. They're an existing HubSpot customer who had a common problem. Their reps were spending more time doing prospect research than actually prospecting. So the reps were spending their time coming through thousands of e-mails and deals, searching for insights on who to reengage, what messaging work, how to personalize their outreach. And they knew that they needed a better solution. And so they did what any modern buyer would do in the new customer journey, and they started building, but they didn't just build on one solution. They build with a Frontier Lab. They built with a point solution, and they built with HubSpot, three different experiments to see which was going to work for them. HubSpot won, and we won because of our trusted context, prospecting agent went out at research, all the old deals by pulling notes, the contact details, the call history, pulled it all from HubSpot-drafted relevant reengagement messaging for the reps. Now you saw from Duncan the difference between an e-mail written by generic AI by -- versus one written by HubSpot with growth context. So it should come as no surprise that Fin's e-mail response rates more than doubled on HubSpot, and they won multiple deals that had previously been written off. So these results were great for them. It led them to 8x their credit commitment with us, and they are now on track to continue to increase their spend this year. Just love the story because we want on proof, not on promise. Now another example of a downmarket customer that I wanted to share is Vision Wave Holdings. They're a microcap defense technology company. And their COO had a problem. He had a major defense trade show coming up, and he had no way to collect leads at scale. He needed outcomes in like literally days, not months. And so we didn't have time for the old traditional customer journey and to go through that whole process. So we just came to HubSpot. He came to HubSpot ready to get started. He got on to our free product and he began building. In one week, using only Breeze Assistant, he built a custom lead registration system for the trade show. He can figure the sales processes and he ingested the company's existing leads and deals. The COO described the experience as the fastest and best CRM implementation in his 30-year career, and he upgraded to our enterprise product touchlessly because of it. As for the trade show, according to him, Vision Wave got an overwhelming pipeline of deals from governments and Tier 1 contractors around the globe. I love this story because it shows the power of Breeze Assistant in action for our customers and how powerful it can be, especially down market. So let me tie it all together. With our new AI native CRM, our value proposition down market is easier to use software, faster time to value and better outcomes. Okay. Let me switch gears now and talk about the fastest-growing part of our customer base, our upmarket customers. So what we are seeing upmarket is more deliberation than ever before. And honestly, I get it, it makes sense. It's a once-in-a-generation platform shift and companies want to make sure that they get it right. What this is showing up as for us is more scrutiny from buyers, larger buying committees and a requirement to just derisk the implementation. Now the good news is that despite this, our product and our go-to-market have never been stronger for a upmarket buyer. As Duncan showed, we have a legitimate platform advantage. Our trusted context drives better outcomes at a lower cost and our headless strategy future-proofs HubSpot for our customers. Customers, when they come to HubSpot can be confident that we will scale with them despite a changing AI landscape. Meanwhile, as you saw the core marketing sales, service and ops use cases, they have all gotten more valuable for our customers. On the go-to-market side, we have more credibility upmarket, more logos, more approved, more peer companies willing to be references for our new customers. And so all of this together means that we just have a real opportunity upmarket. So here are the new ways that we are meeting the moment. The first place we're focused is the initial sale. We know these upmarket buyers are more hesitant and have higher expectations. So we have updated our sales playbook to match. We're doing more account-based marketing. We are involving C-suite decision-makers earlier, and we have built out a dedicated specialized sales team for service and CPQ use cases to help those buyers. To derisk the sale, we are investing in more resources to ensure that our customers see tangible value from their first use case as quickly as possible. What this looks like in practice is funding more partner services, building outcome-oriented implementation plans and offering HubSpot go-to-market engineers to support their AI setup. And to help drive expansion into AI, we are partnering with our customers much more deeply than we ever have before. We are doing discovery focused on how their core go-to-market processes work. We're mapping their problems to agentic solutions, and we're granting them credits to actually trial our agents to solve those problems. We are embedding our HubSpot partners into the delivery and ongoing success of all of that as well. So a better playbook, more protection upfront, more partnership long term, all guided by greater personal engagement market. And we're already seeing the results. We are competing in and winning much bigger deals, deals over $120,000 of ARR, up 38% year-over-year, and our partner attachment to upmarket customers has grown as well. Now let me walk you through two stories of upmarket customers that I really love as well. These are stories of customers who are driving incredible outcomes with HubSpot. So the first is Nation's Roof. Nation's Roof is one of the largest commercial roofing contractors in the U.S. They've got over 40 locations across the country. So they originally came to HubSpot for marketing. They needed more leads, and they needed to automate more of their marketing activities. And while the marketing team got going on HubSpot and experienced what it was like to engage customers with HubSpot and experience the ease of use of HubSpot, their sales team was on a different planet. They were on an incumbent platform that they referred to as a Franken system. They told us it was over-customized, it was over expensive, and it was so painful to use that rep simply avoided it. I hear the story like multiple times a week, by the way, from different companies. Within a year, the CRO decided that it was time to make a change. So he went to the market looking for something different. And I remember meeting him, it was very distinctly the way that he described what he was looking for and his hopes and dreams in a new CRM platform. He said, "I just want a CRM where the reps don't have to do data entry. That's what I want. That's what I need." And he got excited as I started talking to him about what we were building with our note taker and our self-updating CRM and our mobile notetaker. He also got excited about the idea that the sales and marketing team would actually work out of the same system and could work together. And you got the most excited about the lower total cost of ownership that he would have to be consolidated all of it on to HubSpot. So that's what they did. What started as a marketing use case, quickly turned into four enterprise-level hubs and 50,000 additional credits, bringing their annual commitment with us from $10,200 up to over $300,000. The next example I want to talk about is a company called Aventus. So Aventus provides software that helps financial institutions detect market manipulation and meet regulatory requirements. The firm sells into one of the most compliance intensive segments of fintech. The cycles are long, the buyers are technical. The addressable market is small and very high value. This means that prospecting for them requires extensive research, depth and personalization. But that doesn't mean it needs to be done by a human. They had a vision for building a prospecting system that would augment their sellers and not replace those sellers, one that could do the research, track the signals and work with their reps. Now to their credit, they jumped right in. They pulled some engineers and were like, let's just build this prospecting system. They went at it. But when they learn that HubSpot had built a prospecting agent in the same system that their sales reps already worked in, they decided to give it a try. They made a small credit commitment, and they ran a pilot on HubSpot. They love the early results. What they loved was that each rep got an agent that was trained on their voice and their selling style, and that agent did all the work. It pulled the context on where lead came from, it pulled the engagement history, it pulled the campaign goals to draft and outreach. All the reps had to do is review it and approve. Based on this, they made the decision that it wasn't worth the cost for them in dollars or engineering time to replicate what HubSpot had already built. So they went all in on their first use case, and they started to scale. Since then, their teams have seen a 27% increase in influenced pipeline, they've got back 87 selling days across the team. And because of it, they've also began to expand into new agent use cases. They are now enriching, tagging and routing good fit accounts with data agent, and they're building a big ABM program for their account management team for a top 50 ABM approach. So in the last four months, they've increased their credit commitment 3.3x with HubSpot, and they are still expanding. In both cases, our upmarket value proposition of an easy, fast, trusted platform with lower TCO delivered exceptional value to our customers. As I hope you can see, we have an immense opportunity upmarket and down market, and we are executing against both of these segments very thoughtfully. So I want to close where I started. It is a noisy time in the paradigm shift and there are lots of options, but HubSpot is positioned to win for the outcomes era. Against the incumbents, our AI native capabilities and our trusted platform deliver on buyers' expectations with a CRM that is easier, faster and provides better outcomes at a lower total cost of ownership. Against custom-coded tools, and everybody wants to vibecode their own CRM, our message to customers is that starting on this path is easy, but scaling it is exceptionally hard, and HubSpot does the hard stuff really well. We provide the governance, the compliance, the maintenance, the integrations, the model neutrality, the ongoing innovation, everything you need to confidently grow. And we do it without the complexity and cost of doing it yourself, and we make it available in whatever way you want to access it. Against AI point solutions, we think history will repeat itself here. In software era, nobody wanted a franking system. And in the AI era, we don't think anybody will want agents brawl. Our unified platform has both breadth and depth. It delivers the trusted context that every agent needs, and it scales exceptionally well. So this era is going to reward the platform that delivers outcomes the fastest. And that platform is HubSpot. And that's why we're leveraging our position to win more customers and drive more growth. So Kate Bueker is now going to take you through our business performance. But first, I'd like to share another example of a customer getting outsized value from HubSpot. Thank you all very much. [Presentation]

Please welcome to the stage, Kate Bueker.

Thank you, Jon. Hello, everyone. It is great to see you all back here in Boston. You've heard from Yamini, Duncan and Jon about where HubSpot is headed as we transform with AI. I am excited to share how this is going to drive durable revenue growth and margin expansion in the years ahead. All right. In terms of our agenda, I'm going to start by sharing how the solid foundation that we built as we democratized INBOUND and CRM puts us in a position of strength to democratize AI. Next, I'll walk through how that strategy drives new growth levers as we align our business to delivering outcomes for customers. And finally, all share how will drive margin expansion and earnings growth as we transform how we build, grow and operate. Over the last 20 years, HubSpot has built the leading go-to-market platform for scaling companies and a solid foundation for growth in the age of AI. This foundation includes our financial scale, platform advantage and exceptional distribution. It is what sets us up to win as we democratize AI. Let's start with our financial scale. HubSpot has a consistent track record of delivering revenue growth, margin expansion and strong free cash flow growth. Since 2021, our revenue has grown at a compound annual growth rate of 23%. At the same time, we've delivered consistent margin expansion, expanding our non-GAAP operating margin by 12 points. And our cash position is strong at $1.4 billion as of the end of Q2. In 2026, we expect to deliver $750 million in free cash flow, representing a 30% compound annual growth rate since 2021. Next is our platform advantage. We've grown our customer base at a compound annual growth rate of 20% over the last 4.5 years to over 300,000 today. We've also driven significant multi-hub adoption, with 70% of our installed base ARR leveraging three or more hubs. Across our 300,000 customers, the scale of our CRM in context is massive, with 5.5 billion company records and 18 billion contact records. And we have nearly 2,000 technology partners that have integrated into our ecosystem. Finally, the platform is unlocking new ways for customers to engage with HubSpot's platform. I know Duncan shared, we were the first CRM to health connectors with all four of the leading frontier models and we are the #1 connector on both Claude and ChatGPT. Last, our exceptional distribution. We are one of the 10 largest business media networks worldwide with over 350 million organic views over the last 12 months. We have a sales and customer support team that is 4,500 strong, and we have 4,000 solution partners that help extend our reach by helping customers build their go-to-markets on HubSpot. And our distribution extends across the globe with 50% of our ARR coming from markets outside the United States. Our strong foundation sets us up to win in the AI era. Next, let's talk about how we will drive growth. Our growth strategy remains strong. Our core growth drivers, our upmarket expansion, multi-hub adoption and downmarket volume. Our strategy also enables a set of emerging growth drivers across seats and credits that align our business model to delivering outcomes for our customers. Let's dive in starting with our core growth drivers, which should all look familiar to you. Our upmarket performance remains strong, reflecting continued demand from larger customers for a unified AI-powered platform. We saw this again in Q2 with large deals growing 38% year-over-year. As a result, deals with over 120,000 in ARR represent 18% of our installed base, up more than 3x since the beginning of 2021. Our multi-hub momentum continues to grow as new and existing customers consolidate their go-to-market technology stacks on HubSpot. The share of our new Pro Plus ARR that is multi-hub has grown steadily. With more than 80% coming from customers with two or more hubs and more than 55% coming from customers using three or more hubs. As a result, only 9% of our installed base ARR is on a single hub today. and more than 70% use three or more. Down market volume remains a key element of our strategy. You heard this from Jon. As of Q2, we had 150,000 starter customers on HubSpot, up from 50,000 in 2021 and representing about half of our total customers. We are transforming our downmarket product, to deliver an AI-native CRM that helps small customers achieve the outcomes they care about at a very low cost. We aim to capture customers early in their journey and grow with them, and we remain committed to this segment. So our current growth drivers are working. And as we execute our strategy to democratize AI, we are seeing traction with a set of emerging growth drivers. These AI growth drivers include seats and credits. Seats are about making human users of HubSpot even more effective. And credits are how we monetize the outcomes we are delivering for our customers. [indiscernible] has been a meaningful driver for HubSpot with paid seats up 80% since our 2024 pricing simplification. We had more than 2.5 million total paid seats at the end of Q2, including more than 1 million paid core seats. We're committed to making every go-to-market employee as productive as possible. As a result, we've continued to add value to our seats and notably, our core seat with features like Breeze Assistant, note taker and now a self-updating CRM. Breeze Assistant adoption continues to scale, growing from 37% of Pro Plus customers at the beginning of the year to 57% in August, and Breeze Assistant genic actions are up 4.4x. We will continue to monetize seats through the -- monetize the core value of our AI and platform capabilities, and we're going to continue to increase the value of seats and drive deeper penetration within our customers. Now let's talk about agents and credits. Duncan outlined our agentic strategy, which includes three opportunities: HubSpot agents, custom agents and APIs. Each of these will consume credits and drive monetization over time. Across all of these opportunities, we will measure our progress with a consistent framework, reach, depth, quality and growth. First, reach. how many customers are adopting our agents. Second is depth, how often our customers taking agentic action. Third is quality. What is the quality of the business outcomes we are delivering, and finally, growth, how is all of this translating into credit and seed consumption. Let's dive in. The first opportunity we see is with HubSpot agents. Our strategy here is to deliver best-in-class agents across marketing, sales and service use cases. In addition to customer prospecting and data agent, we launched campaign, nurture, content and revenue agents this week at unbound. Let's unpack the early momentum we are seeing, starting with reach. 19% of Pro Plus customers using HubSpot used HubSpot agents as of August, and this is up 2x from the beginning of the year. Over the same time, we have seen the number of monthly agentic actions increased by 3.5x, reflecting the quality of the outcomes we're delivering for customers. And we've seen monthly credit consumption more than double, even with the change that we made in April to lower agent prices, and shift to outcome-based pricing, which created a near-term headwind. And equally encouraging, we are seeing credit usage increasingly diversified across our agents. Customers are not just adopting a single agent. They are beginning to use agents across the entire customer journey. Let me give you an example. Sesame HR is a 400-person software company with more than 18,000 customers worldwide. As the company grew, its support team struggled to keep up. At best, they could cover 70% of their tickets, leaving the rest unanswered. Instead of hiring more reps to handle the volume, Sesame ran a one-month pilot with customer agent, and they saw results. Within a few months, they could -- they were able to cover 100% of the tickets with 60% of the tickets resolved through customer agent without any human escalation. Customers got a better experience and the support team had more time to focus on higher-value work. The value is translating directly into credit consumption. Sesame quickly used its initial credits, and they purchased additional credits that represented about 100x the initial amount. This resulted in $80,000 in genic ARR. The company has been very encouraged by the success with customer agent, and they are testing additional HubSpot agents. Sesame is piloting by our intent to help its sales team drive conversion rates and better outcomes, just like its support team. Okay. The second opportunity we see is with custom agents. As Duncan shared, the strategy here is to make it incredibly easy to build agents and go-to-market automation directly on HubSpot. Breeze Assistant and our solution partners will help drive adoption. Well, early, we have seen strong initial interest and engagement. Over 3,000 custom agents have already been deployed since we launched in July. And the most common adoption in deals is -- are things like internal workflows around deal management and revenue operations, places where customers can quickly save time and drive consistency in their process. We see a significant monetization opportunity with custom agents as our partners and customers build more agents and lever and consume credits. Let me share an example. Cyber Advisors is a nationwide cybersecurity leader with over 400 employees. The company saw an opportunity to make it go to market more efficient, while driving incremental growth. And they turned to HubSpot agents to help. To start, they launch pilots with customer agent, content agent and prospecting agent, and they saw results. Prospecting agent now sends a personalized e-mail to every new lead before the team member picks up the phone. And so far, more than $700,000 in closed revenue has been attributed to prospecting agent. But cyber advisors aren't just using HubSpot agents. They are building custom agents on HubSpot that suit the unique needs of their business. For example, the customer built a scoping agent, that parses call transcripts in real time and creates and sends a ticket to the solutions architect to prepare a quote. This is a process that used to take days and deals would often go cold as a result. Now it happens in minutes. Cyber Advisors also built an upsell agent that automatically analyzes customer counts based on things like industry and company size, and then it recommends services that they are most likely to need to the sales reps, allowing the sales reps to have much more targeted conversations. Together, HubSpot agents and custom agents are transforming cyber advisers go to market, and they are mapping out plans to build additional agents on HubSpot's platform. The third opportunity we see is with APIs. This is all about extending the ways that customers and partners work with HubSpot's platform. As Duncan shared, this includes accessing our platform through data, context and work APIs. We're confident in our ability to monetize the value we provide to customers no matter what work surface partners and customers choose to operate. All right. I'll end where I began. As we deliver AI value with HubSpot agents, custom agents and APIs, we expect credit consumption to scale as we monetize the work and the outcomes we're delivering for our customers. And as we see traction with these emerging growth drivers, we want to make it as easy as possible to buy and scale with HubSpot. This means lowering the barrier to entry to get started and giving customers more choice and control over their AI spend. We took a big step in this direction in 2024. We simplified pricing for sales and service by lowering seat prices, removing seat minimums and introducing the core seat to give all go-to-market employees access to the platform. These decisions eliminated large upfront commitments and made it easier for customers to adopt and grow with HubSpot. What we expected was more customers lower initial ASPs and better net revenue retention, and that is what we saw. The change drove approximately 2 points of benefit in net revenue retention in 2025. And and we are seeing a similar benefit in 2026. The decision to simplify seats was the right one, and it gives us confidence that further simplifying our pricing to drive AI adoption is the right thing to do for our customers, and also for our business. So let's talk about where we're going. We want to make it easier to consume functionality across all of our hubs, by giving customers predictability in their spend and flexibility in how they purchase HubSpot. We've made AI pricing more predictable by aligning agent pricing to outcomes by making credit usage visible across our platform and by letting customers set limits so they are not surprised by runaway AI costs. And customers have more flexibility than ever in how they choose to purchase HubSpot. Customers can pay for credits by buying capacity packs or simply by paying as they go. And as Yamini shared -- we have been piloting a simpler seats and credit space model with new customers in the Nordics and Benelux. Our early learnings give us confidence that we're on the right track and we are expanding our pilot to new customers in EMEA starting next month. We will continue to test, learn and evolve pricing in a thoughtful way. And over time, we expect that seats and credits will be the primary drivers of our revenue. Let's recap. Our core growth drivers of upmarket expansion, multi-hub adoption and downmarket volume remain large opportunities. Our strategy drives new growth levers of credits and seats as we align our business model to delivering outcomes for customers, and we will continue to remove barriers to AI adoption and simplify pricing. Let's shift gears and talk about the meaningful margin expansion opportunity ahead. Yamini set the table well for me. As we transform how we build, grow and operate, we expect to deliver durable margin expansion and drive outsized earnings growth. Let's dive more deeply into how we think about delivering leverage, starting with COGS and gross margin. We plan to invest aggressively to drive agent adoption and usage across our platform. As a result, we expect that COGS growth will outpace revenue growth in 2027. That said, Aviator gives us the flexibility to leverage models from different vendors and route tasks based on which model will deliver the best quality at the best price. This gives us a structural advantage. As model capabilities grow and prices change, we will continue to optimize. Turning to OpEx. We're transforming how we operate internally, and we see meaningful opportunities to deliver leverage across the organization. In R&D, we'll continue to invest to drive innovation, but our best-in-class internal agentic platform lets us do this more efficiently and deliver leverage. In sales and marketing, we use AI today to handle the majority of our support tickets to create and localized marketing content and to drive growth in AEO visibility. We're going to build on this momentum and continue transforming our teams with AI. We're focused on standardizing usage and driving consistent adoption of AI-enabled processes. And in G&A, we'll see modest leverage by continuing to lean in to automation, systems and data. We are confident in our ability to balance gross margin pressure with OpEx leverage and drive continued margin expansion in the years ahead. Looking ahead, our 2026 guidance for non-GAAP operating margin is 21%, representing a 2-point increase year-over-year. In 2027, we expect non-GAAP operating margin of 23% to 24%, representing 2 to 3 points of incremental margin expansion, and we are raising our long-term target for operating profitability. We expect to deliver non-GAAP operating margins of 30%, GAAP operating margins of 20% and stock-based compensation as a percentage of revenue of 10% by 2030. Before wrapping up, I want to reinforce our long-standing principles of balancing growth and profitability and share more about how we intend to manage the business moving forward. First, we will solve for long-term revenue growth. We have a deep conviction in our strategy, and we're making the strategic decisions to set ourselves up for durable, long-term revenue growth. Second, deliver operating margin. We expect to continue expanding GAAP and non-GAAP operating margins. Third, we will deploy capital strategically. Using free cash flow for organic and inorganic opportunities that accelerate our AI strategy and for opportunistic share repurchases. As of Q2 2026, we repurchased more than $1 billion in shares over the prior 12 months. Fourth, manage dilution. We'll manage dilution through a disciplined stock-based compensation strategy. If we take all of these together, we believe that the business can compound earnings growth at a very healthy rate into the future. Let me wrap with three key takeaways from today. First, the solid foundation we built as we democratized INBOUND, CRM and built a durable platform company puts us in a position of strength to democratize AI. Second, our strategy creates emerging growth drivers as we align our business model to delivering outcomes for customers. And finally, we are unlocking durable margin expansion and earnings growth as we transform how we build, grow and operate. Thank you for your time today and for your continued support of HubSpot. With that, we'll move into our Q&A session.

All right. Great. So we're going to move into the Q&A session now. I'm sure there's going to be lots of questions. So if you have one, please raise your hand. We've got a couple of mic runners that are going to be running around and so you can just stay seated. We would just ask that you please introduce yourself before you start your questions. We know who you are and try to stick to one question, so we can get to as many folks as we possibly can get to. So I'll give it just a moment here for hands to go up, and we'll get going here. Yes. Let's start with Jackson.

Jackson Ader at KeyBanc Capital Markets. Thank you so much for doing this. I guess we'll start with Yamini or Dharmesh really. I mean the big question right now that I think I have and investors have are around the head list strategy. Can you just explain it simply is like why is headless good for an application software company. I feel like I've spent my entire life thinking that the higher up the stack you are, the closer to the user, the more value accrues to that company rather than being down the stack. Why is that different in 2026?

Okay. I'll start and then Duncan as well as Dharmesh can add, if they have. So take a huge step back. I think software was always meant for users, and there's a way in which users got value, and that continues. But what is different with AI is that agents can operate on top of software and drive value as well. So there's a big paradigm shift that is happening. And the way we think about head list, and I think Duncan did a great job of like laying out what our headless strategy is. There are three ways in which now people can interact with HubSpot. The first is by having Breeze Assistant as the key way in which you get the value out of HubSpot. Think about that as users not having to learn how to navigate, not having to learn how to move screens and really having a conversation to get the most value from an application like HubSpot. That is a new interaction mechanism, and that's what is growing, and we are seeing that adoption within HomeSpot. The second way is through connectors. Everybody is talking about connectors. We were 1 of the first ones to actually build connectors with all four of the Frontier model companies. And we've been having that within our ecosystem for more than 15 months. And that is a new interaction way. Just like way back when the Internet came those browser and applications became available on the browser. Now there is a new interface in which customers can actually go and ask questions about the software applications that they're using, and that's what connectors enable our connectors, we've seen nearly 350,000 users on connectors. When they leverage it. What they're doing is they're going to an LLM of their choice that they're working on and getting insights getting reports, maybe even creating artifacts to really get more value, and we've seen engagement go up and the cohort of customers that are using those connectors, we've seen 12-point improvement in retention. So more users are using HubSpot. They're engaged more deeply with HubSpot. And because of that, the retention is going higher. So that's the second mode of interacting with software that is completely new with the age of AI. The third is actually an agent talking to software and getting more value. The first two are users interacting with software to get value. The third is really an agent that runs. And we are now making it easy to get to the data layer, but more importantly, from our perspective, to the context and the work layer, which is what Duncan laid out in terms of the strategy. And why is that valuable? Because the number of ways in which you can build with AI, you can automate more things with AI than you ever did before with just the software applications before, it is expanded, which means agents get to build, and that means our data, our context and the work that we deliver also continues to expand, and that's how we think about the strategy. So look, lots changing, and we feel really confident in terms of where we are going with our strategy. I don't know if you want to add.

The only thing I'd add on top of it is if you think about that 3-layered API strategy, when you're interacting with HubSpot at the context and work layer, you're getting all the benefit of us doing our fundamental job, which is our job is to go out there and deliver the highest quality, best value, fastest go-to-market outcome. So as part of all of that, you get everything running on the aviator agent harness, leveraging growth context. So at the end of the day, that's always our #1 job. The second thing is we want to be able to meet customers where they are. So if they're interacting with us via an LLM like Claude and ChatGPT, we want them to be able to do that really easily.

Okay. Great. Let's maybe go down here to Elizabeth.

Elizabeth Porter from Morgan Stanley. You highlighted the flexibility of being able to choose between seats and credit consumption. And just as that credit consumption starts to expand, how much of the opportunity should we think about that it truly being incremental to the wallet versus just a shift from what would normally be priced into a seat or a hub. Just talk to some of the confidence that you're seeing that credit expansion really drives average revenue per customer higher than simply shifting the composition of spend.

Yes. Thanks a lot, Elizabeth. I'll start and feel free to chime in as well. Look, the easiest way in terms of the framework that we think about, is that if AI and our platform adds value to someone within a go-to-market team, it belongs to seats monetization. So you'll see us add more and more value through that, which is breeze assistant where people charge for pilots. We just include that within our core seat as well as no taker as well as a lot of the functionality. So if we help go-to-market team members get much more effective, that belongs to the seat. And if we deliver outcomes or value that scale beyond people that are using our software, then that belongs to credit. That's the simple way in which we frame these two. Now over a period of time, we do think that the agentic use cases are going to continue to expand. Today, we laid out our framework for how we think about it. HubSpot agents or domain-specific agents will continue to expand both the number of these agents as well as the depth of outcomes that they deliver. And as we do that, we'll see it in our framework of reach, depth as well as the quality. Custom agents pretty early, but we're seeing really good published rates in terms of how many customers are deploying custom agents as well as custom workflows, they consume credits, and then we laid out the structure for API. So the question you're getting to is how can on you be sure that these are all additive? Well, we're very early in the stages, but the examples that both Jon as well as Kate gave today show the cases where it is additive. It is additive because the customer agent side, we're adding to how our customers are resolving tickets, it's much bigger than the opportunity that we had before. prospecting agent, we're continuing to see how we can drive even more outcomes by delivering qualified leads, that is additive. So at least so far, we've continued to see. I think in the fullness of time, even if the seats are capped based on the headcount, we'll continue to deliver more genetic use cases. And as those agent use cases expand, we think there is a huge opportunity for us. It is early stages, but lots and lots of opportunities.

Yes. I guess just to maybe restate and reinforce what Yamini had to say, like with small- and medium-sized businesses, there are many, many cases where they just aren't getting to all the stuff that they can and should be doing to grow their business. You again heard that from me and from Jon, and that the new value that we can deliver through agents or even just our AI-enabled customer platform allows them to do more work and we're delivering more value than we ever have before.

All right. Let's go over to Matt maybe here.

Matt VanVliet from Cantor. I guess, as we see more companies approach this sort of headless architecture, and it seems great to unlock a lot of the value that sort of trapped in the data model today. But how is more of the work maybe gets done on an interface that isn't native to HubSpot? Do you ensure that the data gravity remains around the Smart CRM, and you ensure that data continues to be written back into instead of over time, they start pulling in data sources from other areas, whether it's cloud data warehouses or even other systems. And it sort of eventually diminishes the value of HubSpot. And so how does that multi-hub approach and getting that into more customers' hands sort of fortify that?

8 SP1 Sure. I mean, look, the way we think about it is this is why Aviator in the agentic harness is so important to be built directly into the smart CRM. The -- what we're doing is focusing on that goal, I was talking about delivering those highest quality, best value fastest go-to-market outcomes. And we're doing that by being able to always have that shared context that's available for all of our agents, our own agents or custom agents to be able to use on top of that. And then exposing that and making the ability to drive those outcomes available via things like the context and work APIs, even if that's headless and inside of another product like a Claude or ChatGPT. So that goal of us being able to always have that mix of structured and unstructured data, computing -- precomputing insights out of that in the context layer and then using that to inject into our own work layer mean that you'll always be able to get those best outcomes directly through HubSpot, regardless of how you're interacting with it. And we think that's going to continue to be a huge opportunity for customers, even in headless.

One thing I'll add just on the headless since this is a point of discussion is that headless exposes the data work API through that mechanism. But the tote thing we're going to start to see, and this is evolving already. is we will see interfaces coming across the wire as well. So we'll have a hybrid conversational interface. I don't think the future is everyone is just going to be chatting with their business software, and that's the only way to interact with it because the affordances aren't there. It's not that user-friendly. So we'll start to see is like, okay, yes, we're going to pass some data across the wire where get some working insights across the wire, but we're also going to have UI elements that are exposed from HubSpot because we've been doing this for a while, that says whether within Breeze Assistant or within third-party AI application, it's going to be a combination of data and UI that's embedded in that conversation. So it's not just data traveling across the wire. And kind of final thing would be within Breeze Assistant, which we're seeing increasing usage of. Yes, it has access to all of our context, but also has access to the user context as as we know what part of HubSpot you're in, would we know what you're doing, and we can actually bring the power of both the Frontier model and HubSpot's domain expertise and side breeze assistant. So I think of the AI apps more as it's like a browser. This is like a container for functionality. It's the new way to embed software. And resistant is our embedded way. So it has access same frontier models. It can do model routing. You can do and now UI embedding. So over the fullness of time, my sense is that, yes, for ad hoc queries, you're going to go ahead listen, maybe you'll sit within Claude or ChatGPT, there's increasing gravity just to Breeze Assistance simply by where it sits in the stack, that has access to everything that's going to be more economical and efficient to.

Brian Peterson from Raymond James. So as you guys hit Phase III of this growth trajectory, I'd love to understand how you're thinking about the net new customer opportunity versus expansion with existing customers. And Kate, I know you're not guiding to that metric, and you historically haven't, but if we look at what you outlined, it would seem to be that the NRR should be increasing. Is that the right way to think about it over the long term? And in durably, what should that NRR look like?

Yes, I think it's a really good question. I do think that there are a number of opportunities that you should have heard throughout today to really drive NRR over the next couple of years. Like certainly, seats and credits are two big opportunities that will fuel net revenue retention over time. I think in the near term, we've been very clear about where net revenue retention goes, which is largely flat based on some of the dynamics we shared in Q2 earnings. But over time, if we can execute this strategy well, there are a lot of tailwinds to net revenue retention.

Tyler.

Tyler Radke from Citi. So lot of great vision on AI, the things that are going to drive long-term revenue growth. At the same time, you've talked about the evolution of pricing, more promo discounting down market changes to what or how you're monetizing and charging for the AI credits. Can you just talk about the puts and takes on revenue should -- does any of this have an impact in how we should think about the shape of revenue growth or ARR growth near term? And then I think you referenced some early good traction in the Nordics, for instance, that you're rolling out to Europe. Maybe just talk a little bit more specifically about what you saw there? And maybe is that going to extend globally to?

Do you want to take first part?