Good afternoon, and welcome to the HubSpot's Q1 2025 Earnings Call. My name is Gigi, and I'll be your operator for today. At this time, all participant lines are in listen-only mode, and there will be an opportunity for questions and answers after management's prepared remarks. If you would like to enter the queue for questions, you may do so by dialing star followed by one one on your telephone keypad. I would now like to hand the conference over to Senior Director of Investor Relations, Ryan Burkart. Please go ahead. Thanks, Operator. Good afternoon, and welcome to HubSpot's First Quarter 2025 Earnings Conference Call. Today, we'll be discussing the results announced in the press release that was issued after the market closed. With me on the call this afternoon is Yamini Rangan, our Chief Executive Officer, Dharmesh Shah, our Co-Founder and CTO, and Kate Bueker, our Chief Financial Officer. Before we start, I'd like to draw your attention to the safe harbor statement included in today's press release. During this call, we'll make statements related to our business that may be considered forward-looking within the meaning of Section 27A of the Securities Exchange Act of 1933 as amended, and Section 21E of the Securities Exchange Act of 1934 as amended. All statements other than statements of historical fact are forward-looking statements, including those regarding management's expectations of future financial and operational performance and operational expenditures, expected growth, FX movement, and business outlook, including our financial guidance for the second fiscal quarter and full year 2025. Forward-looking statements reflect our views only as of today, and except as required by law, we undertake no obligation to update or revise these forward-looking statements. Please refer to the cautionary language in today's press release and our Form 10-Q, which will be filed with the SEC this afternoon for a discussion of the risks and uncertainties that could cause actual results to differ materially from expectations. During the course of today's call, we'll refer to certain non-GAAP financial measures as defined by Regulation G. The GAAP financial measure most directly comparable to each non-GAAP financial measure used or discussed, and a reconciliation of the differences between such measures, can be found within our first quarter 2025 earnings press release in the Investor Relations section of our website. Now, it's my pleasure to turn the call over to HubSpot's Chief Executive Officer, Yamini Rangan. Yamini? Thank you, Ryan, and welcome, everyone. Today, I'll focus on our Q1 2025 results and the key trends driving our performance. Then, I'll highlight the clear momentum we are seeing with AI adoption, how our spring spotlight innovations are accelerating it, and why we believe HubSpot is uniquely positioned to lead and win with AI. Let's dive in. Q1 was a solid start for HubSpot, with revenue growing 18% year-over-year in constant currency. Total customers grew by 19% to over 258,000 globally, fueled by over 10,000 net customer additions in the quarter. I'm also pleased to announce that our Board of Directors has authorized our first-ever share repurchase program with up to $500 million under authorization, a clear signal of the confidence we have in our business and the growth opportunity ahead. This strength is reflected in our Q1 results, driven by three consistent pillars: platform consolidation, up-market expansion, and strength in customer acquisition down-market. Let's start with platform consolidation. HubSpot's customer platform is becoming the clear choice for businesses looking to drive AI innovation while reducing total cost of ownership with a unified, simpler tech stack. Over 37% of our Pro+ customers by ARR are using four or more hubs, up seven points year-over-year. In Q1, we saw more new Pro+ customers start with our full platform than ever before. This is a clear signal that our platform for strategy is working. In up-market, we continue to see real momentum driven by product innovation and go-to-market execution. We delivered key enterprise-grade functionality that larger customers need, including journey orchestration in Marketing Hub Enterprise to run sophisticated workflows, enhanced sandbox functionality to support complex implementations, and the launch of three new data centers in Sydney, Oregon, and Montreal. As a result, large deal growth was strong, up 23% year-over-year, with clear momentum among our installed base. Our partners continue to be a key driver up-market, with co-selling up 41% in Q1. In down-market, we see strength in customer acquisition driven by pricing changes and improvements in the free-to-starter conversion. We made our product easy to buy and easy to upgrade, and those changes are a clear tailwind. Free-to-starter conversion is up year-over-year, driven by better onboarding, product improvements, and flexible monthly payment options. Our Smart CRM and core seat offerings are gaining strong traction, expanding our reach beyond traditional personas to include ops, admin, and finance users who need powerful editing capabilities. 24% of our Pro+ customers on the seats model have purchased additional core seats. It's rewarding to see customers scale with HubSpot, and the consistency of our results shows our platform is delivering real value. Okay, before we move into AI innovation, I want to acknowledge the macro environment. Uncertainty remains the constant. We entered the year with more growth conversations, and while that has not shifted dramatically, we are seeing a sharper focus on value. HubSpot is built for moments like these, and we have a proven playbook to communicate value. As a crafted customer platform, we help businesses consolidate tools, cut costs, and unlock the power of a unified customer view. Our ease of use and fast time to value allow customers to onboard quickly and see results. With AI embedded across all hubs, we deliver immediate efficiency, and we believe that AI's value is amplified further when companies prioritize driving productivity. In uncertain times, HubSpot is the predictable choice for scaling businesses. Okay, with that, let's shift gears and talk about AI innovation and momentum. Our AI strategy has been clear. We will embed AI into every hub and every part of the platform. One product, one AI-first customer platform, no bolt-ons. We've executed against that strategy by introducing hundreds of embedded features across hubs, launching Copilot to help go-to-market teams get work done faster, and rolling out agents that can do the work for them. We have reinforced the strategy through strategic acquisitions, including Frame AI and Dashworks, to accelerate our roadmap. And that strategy is working. Over the past year, Content Hub attach rates have tripled, and Service Hub adoption has improved because of embedded AI. Copilot adoption has accelerated, with the number of users who have engaged with Copilot more than doubling from 270,000 users in Q4 to over 660,000 in Q1. Our agents are gaining traction. Customer Agent adoption has doubled in three months, now serving over 2,500 customers with average resolution rates of over 50%, and some seeing 80% AI resolution. Service Hub customers with Customer Agent close tickets 39% faster than customers not using it. Prospecting Agent, now in public beta, automatically researches target accounts, personalizes outreach, and engages prospects to help sales teams grow pipeline. We now have over 2,300 customers activating on Prospecting Agent, signaling clear demand. These wins are important, not just because they show clear adoption and value, but because they validate a core belief. Our customers want AI embedded within HubSpot, not bolted on. Many have organizational mandates to drive AI initiatives and appreciate our ease of use and fast time to value. We are doubling down on Customer Agent because it's driving real results. We're seeing clear product-market fit, and therefore, in June, we are expanding it beyond Service Hub to all Pro and Enterprise customers across every hub. What's been fascinating is how customers are using Customer Agent across the entire journey, not just for post-sale support. Whether answering marketing questions like, "Where's your next webinar?" or sales questions like, "How does your pricing work?" Customer Agent is already delivering value at every stage of the customer journey. And that's why we believe Customer Agent will become the single most important AI agent every business will adopt, and the easiest way for companies to scale with AI. Starting June 2nd, Breeze Customer Agent will be available through our existing credits-based system, the same simple model we introduced at inbound with Breeze Intelligence. Here's how it's going to work. To make it easy for customers to get started, Pro and Enterprise customers will receive a monthly allotment of credits included in their subscription. This means they can begin using Customer Agent immediately at no extra cost. As they see success and want to scale, customers can purchase additional credits, either in flexible amounts or in easy-to-choose packs, giving them full control and predictability over their spend. This is just a step in our broader strategy to monetize AI using both seats and credits. Over time, we'll extend credits-based pricing to other AI agents and features, staying true to our philosophy of delivering value before monetizing. While we believe this approach will be a strong long-term growth driver, because we're launching this mid-year and adoption will take time to ramp, we do not expect a material impact to our 2025 revenue from this change. Okay, let's talk about how we are sustaining our AI momentum. We have continued to innovate rapidly and released over 200 new features at our first-ever in-person Spring Spotlight event. Our innovations focused on three clear themes: deeply embedding AI across the customer journey, advancing multi-agent orchestration, and driving up-market scalability. We have significantly enhanced Customer Agent, Prospecting Agent, and Content Agent to help customers market, sell, and support better. For example, Content Agent now proactively suggests blog topics, learns from marketer-provided context, and automates internal blog linking, creating high-quality targeted content faster. Prospecting Agent now combines HubSpot and external data to personalize outreach, delivering insights directly into Sales Workspace. Now, as we lead the way in agent-to-human collaboration, we are taking a major leap forward with agent-to-agent orchestration, a shift we believe will define how work gets done in the future. At Spring Spotlight, we introduced Knowledge Base Agent, a breakthrough in multi-agent orchestration. As we watched Customer Agent adoption closely, we saw a clear pattern. Customers with strong, well-documented knowledge bases achieved the highest AI resolution rates, and to help more customers succeed, we built Knowledge Base Agent to work alongside Customer Agent, automatically identifying knowledge gaps and drafting new articles from unstructured data like support tickets, calls, and emails. This is agent-to-agent orchestration in action. Two agents working together using unified HubSpot data to deliver faster results. Alongside these AI advances, we have expanded our enterprise capabilities with powerful new features like lookalike lists, journey automation, and multi-account management in Marketing Hub Enterprise, plus enhanced workspaces for sales, service, and customer success teams. Together, these innovations give customers the confidence that by choosing HubSpot, they are AI-first and are future-proofing their business for what's next. Our approach to AI is differentiated, and that is what will set HubSpot apart. We unify structured, unstructured, and external data at scale, giving our AI a richer, more complete foundation. We layer deep context across the entire customer journey: brand, tone, voice, customer interactions, something only a true platform, not point solutions or agents can deliver. With nearly two decades of experience serving SMBs, we understand exactly what growing businesses need from AI, and we're building features and agents that deliver real outcomes, not complexity. And through Agent.ai, we're incubating a vibrant AI agent ecosystem, already engaging over 1.7 million users and 17,000 agent builders. We uniquely bring together data, context, customer focus, and ecosystem innovation, and that's why we're positioned to lead the next era of AI-powered growth. With that, I'll hand it over to our CFO, Kate Bueker, to walk you through our financial and operating results. Kate. Thanks, Yamini. Let's turn to our first quarter 2025 financial results. Q1 revenue grew 18% year-over-year in constant currency and 16% on an as-reported basis. Subscription revenue grew 16% year-over-year, while services and other revenue increased 13% on an as-reported basis. Q1 domestic revenue grew 16% year-over-year. International revenue growth was 19% in constant currency and 15% as reported, representing 47% of total revenue. We added over 10,000 net new customers in Q1, modestly above our expectations, and ended the quarter with a total of 258,000 customers, growing 19% year-over-year. Average subscription revenue per customer was $11,000 in Q1, down 2% year-over-year in constant currency and 4% on an as-reported basis. Over the next few quarters, we expect net additions to moderate to roughly 9,000 and ASRPC growth to be approximately flat. We continued to see healthy customer dollar retention in Q1, holding in the high 80s. Net revenue retention was 102% in Q1, down two points sequentially as expected. For the full year of 2025, we continue to expect net revenue retention to be up a couple of points year-over-year, primarily driven by the impact of our seats-based pricing model change. Calculated billings were $767 million in Q1, growing 18% year-over-year in constant currency and 20% on an as-reported basis. The remainder of my comments will refer to non-GAAP measures. Q1 operating margin was 14%, down one point compared to the year-ago period. As a reminder, there was a one-point headwind to operating margins due to an increased company match rate for our 401(k) contributions that shifts more costs earlier in the year. Net income was $96 million in Q1, or $1.78 per fully diluted share. Free cash flow was $122 million, or 17% of revenue in Q1. Finally, our cash and marketable securities totaled $2.2 billion at the end of March. As Yamini shared, our board has authorized HubSpot's first-ever share repurchase program of up to $500 million over the next 12 months. This reflects our strong conviction in our long-term opportunity and our commitment to delivering shareholder value. While we will continue to prioritize investments in organic innovation and remain selective with strategic M&A, our healthy balance sheet and strong free cash flow give us the flexibility to settle our convertible bond in June and execute this repurchase program, which will help offset dilution. Okay, before I move into guidance, I want to touch quickly on the macro environment. I think Yamini said it perfectly. Uncertainty remains the only constant. While we haven't seen a significant change in our business trends, we are seeing a heightened focus on value and a higher level of uncertainty regarding the economic environment, which we expect to persist through the remainder of the year. With that, let's dive into guidance for the second quarter and full year of 2025. For the second quarter, total as-reported revenue is expected to be in the range of $738 million-$740 million, up 16% year-over-year in both constant currency and on an as-reported basis. Non-GAAP operating profit is expected to be between $124 million and $125 million, representing a 17% operating profit margin. Non-GAAP diluted net income per share is expected to be between $2.10 and $2.12. This assumes 53.2 million fully diluted shares outstanding. For the full year of 2025, total as-reported revenue is now expected to be in the range of $3.036 billion-$3.044 billion, up 16% year-over-year in both constant currency and on an as-reported basis. Non-GAAP operating profit is now expected to be between $558 million and $562 million, representing an 18% operating profit margin. Non-GAAP diluted net income per share is now expected to be between $9.29 and $9.37. This assumes 53.3 million fully diluted shares outstanding. As you adjust your models, please keep in mind the following. Our EPS guidance includes the impact from both the planned convertible bond settlement and our share repurchase program. We expect the repurchase program to be slightly diluted to EPS in the second half of 2025 due to the lower interest income from reduced cash. That said, the reduction in share count will build over the next 12 months to offset the lower interest income. We continue to expect CapEx as a percentage of revenue to be roughly 5%, and now expect free cash flow to be about $570 million for the full year of 2025, with seasonally stronger free cash flow in Q4. With that, I will turn the call back over to Yamini. Thank you so much, Kate. As we close, I want to express my heartfelt gratitude to Brian Halligan, our co-founder, who has decided to transition from his role as executive chair to a regular member of our board. After three and a half years as executive chair, Brian is stepping back to devote more time to his work with Propeller Ventures and his teaching role at MIT. I am very grateful for Brian's support, guidance, and investment in me during his time as executive chair. Brian will continue to serve as a board member, a founder, and a large shareholder supporting HubSpot from the boardroom. Lorrie Norrington, our longstanding lead independent director, will assume the board chair role. I want to thank our customers, our partners, and investors for the continued support. And a huge, huge thank you to all HubSpotters around the world for staying focused on solving for our customers every single day. With that, operator, let's please open up the call for questions. Thank you. If you would like to ask a question, please dial star followed by one one on your telephone keypad now. If you change your mind, please dial star followed by one one again to exit the queue. When preparing to ask your question, please ensure your phone is unmuted and limit yourself to one question per person. One moment for our first question. First question today is from Mark Murphy from JPMorgan. Thank you so much, and congrats on a very steady performance. So Dharmesh, I was hoping you could give us an update on Agent.ai. And just as part of that, there was some talk about multi-agent orchestration. How are you doing that across the hubs? So for instance, an agent in Marketing Hub, how will it talk to an agent in Sales Hub? And I'm trying to understand if you're going to use something like Model Context Protocol or some other communication mechanism, and then have a quick follow-up. Okay, thanks for the questions, so on the Agent.ai front, so as Yamini mentioned, Agent.ai is one of our innovation bets, and the thesis behind the bet is that the future is going to consist of tens of thousands, hundreds of thousands of agents all collaborating and working with each other to accomplish higher-order goals, and so our kind of goal behind Agent.ai is to kind of empower citizen builders to build this next generation of agents, and we're really pleased with the progress that that innovation bet has made. Now, one of the ways I think these agent-to-agent collaborations will happen is with what's called Model Context Protocol, which you brought up. And what Model Context Protocol is, is an open standard that defines how AI applications and agents can communicate with a variety of tools and backends and services. And so this benefits HubSpot in a couple of ways. One is, earlier this week, I'm super proud the team launched into public beta our MCP server support, which allows any AI application, current and future, to be able to communicate with HubSpot's backend over a natural language interface with an LLM. And so where this unlocks potential is that now HubSpot can be used by all sorts of applications. They can get to that data easily. So think of it as a next-generation API. So past APIs were built for classic applications. Model Context Protocol allows agents to be able to consume HubSpot services. The flip side of it is HubSpot will also be an MCP client. This will allow us to then interact with all the third-party systems that are out there and be able to bring that data into our normalized hybrid structured and unstructured form. So it kind of expands broadly the footprint of data that we will have access to. So we're super excited both by the kind of agent network as represented by Agent.ai and the opportunity that MCP, the Model Context Protocol, unlocks. Thanks for the question. Thank you. One moment for our next question. Our next question comes from the line of Samad Samana from Jefferies. Hi, good evening, and thank you for taking my question. Great to see the strong execution in a tough backdrop. Maybe, Kate, for you, just as I think about the revised guidance, it's obviously impressive, but it now implies a different shape of growth as the year progresses than where we started. And so I was wondering if you could maybe just help us contextualize how we should think about the quarterly progression. And then maybe related just on the margin side, sticking with guidance, how much you're absorbing in terms of OpEx from M&A? Because I'm guessing you would have revised that maybe higher otherwise. Just maybe help us understand that guidance just a little bit better. Thank you so much. Oh, sure thing. I would just maybe start, Samad, by saying, well, there are maybe a little bit more complex pieces to the puzzle. The way that we've approached guidance for this quarter is very much consistent with the way we always approach guidance. So I just maybe I'll take you through the pieces, and it will help a little bit here. I think you're well aware, foreign exchange has been quite volatile. Since we reported our earnings in February, the U.S. dollar has weakened pretty meaningfully, which is favorable to our as-reported revenue results. And we flowed this through to our full-year guidance. So we flowed through $50 million of FX upside, which is all of the movement that we saw in rates between February and today. That said, we did not flow through our Q1 outperformance into the full-year outlook. And the decision not to do that is really a reflection of the increased macro uncertainty that you heard both Yamini and I talk about in the prepared remarks. A couple of things maybe just to highlight as you think about our full-year guide. Yamini said it pretty clearly. We have not seen any material headwinds to the business through April, but the environment is obviously highly uncertain, and we're actively monitoring trends in the business. Secondly, as always, our guidance is going to reflect a range of scenarios that we run in preparation for this call. We have not assumed a full-scale downturn in the economic situation this year, but we do believe that our revised outlook is appropriately prudent, and it really does reflect sort of a confident view of the business that we see today. In terms of the operating profit side and the impact of M&A, it's honestly very small. Thank you. One moment for our next question. Our next question comes from the line of Arjun Bhatia from William Blair. Thank you so much. Two questions. One, Dharmesh, if I can follow up on Mark Murphy's earlier question on Agent.ai. Is it right to assume that the experience of an agent that's coming into HubSpot through Agent.ai is going to be the same or not materially different than the agents that HubSpot is building, that you're building yourself? Or are there differences that we should be thinking about just in terms of the customer experience and how that gets developed? And then one for just on the macro, I know you pointed out a little bit of customers are putting higher focus on value. I'm curious if you're seeing certain verticals, certain end markets, or certain types of customers be more hesitant than others. Is there any way to kind of just get a little bit more granular into where you're seeing that higher focus on value thus far? Thank you. Let me start with this, and then I'll have Dharmesh step in on the agent. I'll start with the overall macro question that you have. As we've said a couple of times here, we've not seen significant changes in demand patterns, and that holds true by industry and segment and geography. Across all of those, we're seeing very similar patterns across the board. Now, look, I think there's just a broader narrative that SMBs may be disproportionately impacted by economic downturn. And while we'll never say never, we simply have not seen that play out in the business over the last few years. And we've kind of been through a lot, a pandemic, a downturn, and macro uncertainty. And the reason for that is that we are a platform that is a must-have. We're not just a nice-to-have, and we're kind of the nerve center of our customers' revenue-generating activities. And the platform that they rely on to connect to their end customers and grow. And our customer base is broad and diverse. And so we're not reliant on one industry, one geography, or a specific segment. And because of that broad customer base, the results tend to be very resilient. Now, you asked a question in terms of agents. Maybe I'll start with explaining our strategy. Our AI strategy is to embed AI deeply into every hub and across the whole platform to deliver a co-pilot, which is a digital assistant for all front office employees, and to deliver a set of agents that are first-party agents that we're building that can help our customers accomplish work. So we're building an agent platform, and that takes all of the unified structured, unstructured data and the context we have across our 258,000 customers and brings that to bear. That is kind of our strategy from a core agentic delivery of value from our customer's perspective. Now, as Dharmesh explained, we also have an innovation bet where we want to have an ecosystem of vibrant developers of agents. And we're certainly looking at that, but the core part of our strategy is to build customer agent, prospecting agent, and deliver value for our customers. Thank you. One moment for our next question. Our next question comes from the line of Gabriela Borges from Goldman Sachs. Hi, good afternoon. Thanks for taking my question. Yamini, I wanted to revisit some of the shared data you mentioned that you're seeing in the mid-market, so that 200-1,000 employee cohort. Now that you have the additional HIPAA qualifications and compliance, how do you think about which customers or which cohorts of customers in the mid-market make the most sense for HubSpot? When you have your salespeople giving you feedback, are there specific targets or specific types of customers, either by vertical or by mindset, that work better as potential HubSpot customers versus perhaps staying on their existing providers? Thank you. Hey, Gabriela, thanks a lot for the question. I think just stepping back, we have been focusing on 2-2,000, and you're absolutely right. The mid-market portion of our segmentation is kind of between 25-200. But I will say that your broader question on what are we seeing, which verticals, where do we see even further opening. I think the reason we have been able to consistently deliver results is because we're seeing momentum in all of our segments. We've had a strategy where we want to drive acquisition at the lower end, and we want to continue to add value. Over the last few years, you've seen us do things like HIPAA support, but also deliver more complex and sophisticated needs of the customers in the upmarket segments, and all of that has played through, and again, as I mentioned slightly before, it's not in one segment or in one vertical, and this is really driving a better product-market fit for mid-market and upmarket customers, and we're doing it without compromising ease of use. When I talk to our customers, when I talk to our sales reps on where there's momentum, they will come back to the same reason. We have a clear value proposition for mid-market and upmarket customers, which is easy to use, fast time to value, and unified data, and because of that, there is just a ton of momentum in both mid-market as well as upmarket. Thank you. One moment for our next question. Our next question comes from the line of Alex Zukin from Wolfe Research. Hey, guys. Thanks for taking the question. I guess maybe getting kind of the AI and macro question differently. As you look at the delta between upmarket traction and maybe kind of downmarket traction at the same time, any different patterns of either Copilot or agentic adoption that you're noticing between the two various cohorts? And then anything even through the month of April in the pipeline around sales cycles that you're seeing maybe different between the enterprise versus low-end pipeline? Yeah, Alex, thanks a lot for the question. So if I look at the sales pipeline, it remains healthy across all of the segments, and we've not seen significant changes in pipeline quality or velocity on any one of those segments, whether it is downmarket, mid-market, or upmarket. And as you can imagine, we're monitoring all of the key leading indicators, both top of funnel as well as pipeline conversion, very closely. Now, if I were to give you color in terms of what's happening within the pipeline in Q1 as well as the beginning of Q2, I would remark at a couple of different trends. The first one is we see a lot of multi-hub deals. And the consolidation of tech stack, the pattern that I talked about in the prepared remarks, that is what is continuing to drive that trend. Our prospects are frustrated with disconnected point solutions and hidden costs and overly complex implementations and setup. And our message, but also proven ability to simplify their tech stack and lower the total cost of ownership while delivering value, is resonating, and that's what we see in the pipeline. Now, in terms of AI, I have a ton of conversations. They really like the fact that AI is built into the platform. Many of our customers, and this is kind of across segments, they have internal mandates of getting started with AI, and they come to us, and we are likely one of the first AI solutions that they're working with. And they appreciate the fact that it's embedded, it is easy to use, it's easy to get started with, and they are excited about how fast we're innovating. And I don't necessarily think that there is a difference between segment. The real factor is, is there someone within the customer who's leaning into AI? That tends to make the difference. But overall, we're winning because we are delivering clear value. We are helping our customers adopt AI quickly. Thank you. One moment for our next question. Our next question comes from the line of Rishi Jaluria from RBC. Oh, wonderful. Thanks so much for taking my questions. And nice to see continued strong execution in a tough macro environment. I wanted to drill a little bit more into pricing for all these AI solutions. Look, I totally understand the idea that you want to get it in front of customers, go through use case discovery before really trying to collect money. And I totally get that. I think that's the right thing to do. If we think about this idea of HubSpot credits for Breeze Customer Agents and expanding that, I guess two questions there. Number one, how should we be thinking about specifically on what vector are customers paying? Is it for the amount of times they're using it? Is it outcome-based pricing? Is it token-based, resource-based? Any kind of puts and takes there would be helpful. And maybe alongside that, given how early we are in the technology, how do you balance wanting to obviously turn this into revenue over time, but also customers are notoriously terrible at forecasting their own usage of even something like AWS, let alone cutting-edge technology that pricing can be very opaque and confusing. How do you kind of help remove that friction and make it more predictable for your customers? Thank you. Hey, Rishi. Thank you so much for the question. I think it's like a two-part question that I'm going to try and answer. The first one is, how are you thinking about this credit-based system? And then how are you going to make it easy for your customers to kind of adopt this? Both of those are super important and obviously something that we have been thinking about quite a bit. Maybe I'll start with grounding us on the philosophy of monetization. Our philosophy is to deliver value. And once we see consistent repeat value and we have high confidence, then we begin to add monetization levers. That has served HubSpot really well throughout our history. And that is exactly what we're doing with introducing customer agent as part of the credit-based system. And we launched credit-based pricing at Inbound last year for Breeze Intelligence. And now, because of the clear value and the adoption that we have seen with customer agent, we're adding customer agent to that same credit-based system. And really, the way we think about this is we want to make it super easy for our customers to get started, to start using, and to scale with agents. And therefore, the way it'll work is Pro and Enterprise customers will have an allotment of credits that is included within their subscription. That means when they begin to start using Customer Agent, they can do that without adding additional costs. And as they increase usage and they want to scale, they can buy credits, and they can either buy pay as you go or they can buy packs of credits. And this gets to the second part of your question, which is by having super easy packs of credits, it gives our customers, especially scaling businesses, predictability and control over their spend. And that's what we have seen even in the earlier adoption of Breeze Intelligence credit packs. They want the ability to be able to predict it, and that's exactly what we are doing. What you can expect is that we'll just have a universal credit system. And because we are a platform, not a point solution, that credit consumption will be based on the value that these agents will deliver and provide. So in the case of customer agent, it'll be the number of conversations that we enable with AI. In the case of prospecting agents, it might be the number of meetings that we book. So it'll consume credits based on the value that we are delivering. I feel really good that we are expanding AI monetization to include both seats space and credit space component. And I feel the platform approach that we have taken to bring these agents to bear will really help us scale our monetization alongside the value that we are delivering to our customers. Thank you. One moment for our next question. Our next question comes from the line of Parker Lane from Stifel. Yeah, hi, team. Good afternoon. Thanks for taking the question. Kate, I think we anniversaried the new seats-based pricing during this quarter. I was wondering, when you look at the stability of gross retention and what we see in NRR, what are you seeing from those cohorts of customers as we get to that anniversary mark? Is there perhaps better potential gross retention trends by landing people at the more appropriate levels or more suitable levels to their needs? Yeah, Parker, thank you so much for the question. We are continuing to see really nice momentum as it relates to the seats-based pricing model change that we made about a year ago. What we are seeing is really strong and consistent seat upgrade trends. And that happens both because customers get started with the seats that they need, and they grow from there. And it also happens because customers are adopting more core seats, which is a reflection of the value that we're delivering in the core CRM, which we always knew. Migrations continue. They're going well. We continue to expect that most of the customer base will migrate by the end of 2025. And as we shared at Analyst Day, we expect that 50%-60% of the NRR of our existing customer base will go through their first renewal by the end of 2025. I also shared in my prepared remarks that net revenue retention is likely to be up a couple of points this year. That is in large part a result of the tailwinds we're seeing from the seats-based pricing model change. Thank you. One moment for our next question. Our next question comes from the line of Elizabeth Porter from Morgan Stanley. Great. Thank you so much. I wanted to follow up on the announcement about customer agent expanding beyond Service Hub. So just given customer agent just went GA, I think, earlier this year. Can you just talk to some of the success that you're seeing that gives you confidence to roll out across the broader platform so soon? And also, given just the benefits of agents plus data, over time, is there an opportunity, you think, to use agents to push into more adjacent categories that may not necessarily be core hubs today? Thank you. Elizabeth, I love that question. I'll take both the parts there. And we're very excited about the momentum that we're seeing with customer agent. I talked a little bit about the adoption that we're seeing. We have over 2,500 customers, average of 50+ resolution rates. In a number of cases, it just takes 15-30 minutes to set it up, especially if the customers have well-documented knowledge base and history of tickets. This is a place where we feel very comfortable for customers to get started with and get value fast. What is even more interesting, and you mentioned this, is when we were observing the early adoption patterns carefully, what we found is that customers are already using it beyond post-sale support requests. We're seeing customers already use it within marketing. When someone asks them, "Hey, where's the next webinar?" or "Where should I go for this particular event?" Customer Agent is answering that. Same thing with sales. Sales teams are leveraging it to handle inquiries on pricing and product features and driving better prospect engagement. So the big aha for us is that, one, customers really want the context across the whole customer journey. And two, we are uniquely positioned as a platform to be able to deliver that type of an agent. And so we really believe that customer agent is one where every customer can benefit from. And so excited about that. I think the second part of your question, agents plus data, does that have a bigger impact? Look, we think that we're just getting started with AI. And if I were to frame it, if everything that we did with our apps was to help people do their jobs, I think agents allow us to do the work to help people. And that's a huge transformation. And that means that the opportunity becomes much bigger. It's not really just the software spend that we're thinking about. It's the labor replacement and the labor augmentation that we can now be part of. And so is there adjacent categories? Absolutely. I think there's multiple ways in which we can add value to our customers. And that is what is exciting about not just where we are getting started, but where we can go over the next few years. Super exciting. Thank you. One moment for our next question. Our next question comes from the line of Brad Sills from Bank of America. Oh, wonderful. Thank you so much. I wanted to ask a question about agents as well here, if I could. Could you give us a sense for how plug-and-play agents are in HubSpot? Is it as plug-and-play as what HubSpot's known for? The simplicity of getting out single-hub, multi-hub deals? Is there a certain amount of data prep that's involved here? We're kind of used to thinking about this as an effort for customers in a deployment cycle, the data layer, and then the orchestration. Or are these just features that run within HubSpot and customers don't think of it any differently than that? So just if you could provide a little sense for the effort involved, if at all, to spin up an agent. Yeah. I would say that you'd have to expect that we'll take any technology and we'll make it easy for our customers to adopt. I think that's why we have won, and that's why we will continue to win in the agentic world as well, Brad. And maybe to answer your question a little bit of depth, agents can augment work for every go-to-market employee. And in order for agents to do it effectively, you need a few things. First of all, agents thrive on data, structured data, unstructured data, external data. And we have always been fantastic at having that structured data. Now we're expanding and getting all of that unstructured data so we can give agents deep, actionable insights from that data right out of the box. The second thing is agents need context. Imagine an employee getting onboarded within a company. Well, that employee needs context. What's your product? Who's your customer? What's your brand voice? How do you communicate to your ideal customer profile? That's exactly what an agent needs. And we have that ability to bring that context across the customer journey. And so the agents that we're developing naturally have the deep context across the whole customer journey. And then agents need intelligence. They need to be able to reason. They need to be able to act. And that's why our platform-centric approach works. So all of this, what does it translate into for our customers? Well, our customers will need data. And if not, we'll help them get the data in a single unified manner. And once they have that data, then adopting agents is not such a big deal. Once they have the data, especially for customer agents, we're seeing it takes like 15-30 minutes if you have well-documented knowledge bases, if you have a history of support tickets, if you have enough to feed that agent. It takes no more than that. So our job is to take complex technology and be world-class at applying that technology to SMBs. And that's what we're excited about, and that's how we're driving the innovation cycle. Thank you. One moment for our next question. Our next question comes from the line of Joshua Reilly from Needham. All right. Thanks for taking my question. How are you thinking about the normalization of upgrades in 2025 after a few years where you could argue there was some pent-up demand building? And how are you thinking about this contribution to net new ARR growth for the year versus the last few years? Thank you. Yeah, Josh, thank you very much for the question. I think that I would answer the question maybe by taking a step back and talking about what we are seeing with net revenue retention, which is really where you see the impact of upgrades show up in our metrics. As you know, net revenue retention was kind of seasonally down a couple of points, but the conversation we had last quarter was that we are going to see net revenue retention up a couple of points in 2025 relative to 2024. This is a reflection of a couple of things. The first is really a continued strong customer dollar retention. We consistently see customer dollar retention hanging out in the high 80s, which is a very healthy number. And then what we are seeing in terms of upgrades and sort of the year-over-year growth is very much focused around seats expansion. And we made the pricing model change to the new seats-based pricing model about a year ago. We are seeing seats upgrade rates be up year-over-year. And again, that's a combination of new customers who start on the new seats-based pricing model, buying what they need, adopting more seats over time as their businesses grow. And it's also a reflection of higher paid seat ratios. We are monetizing the core CRM through our core seats. That was something new in the new pricing model. And we are seeing customers really see value in that core CRM functionality. And that's the main driver of the increased expansion year-over-year. Thank you. One moment for our next question. Our next question comes from the line of Keith Bachman from BMO. Hi. Thank you very much. I wanted to try to understand how you're thinking broadly about consumption revenue adding to the model. So to break it down, you did introduce consumption-based pricing last year. Yamini, you clearly said that the customer agents wouldn't contribute this year. But I'm trying to understand if I think philosophically about the next couple of years, how will consumption-based pricing contribute to revenue growth? And either the answer could be something to the extent of we see it adding X basis points or more likely, is there any way to characterize if you thought about your growth algorithm that 90% comes from traditional seats-based and 10% comes from consumption? But any dynamics you could give us on how, at least philosophically, you're thinking about consumption adding to growth over the longer term? Thank you. Yeah. Thanks, Keith. And And I'll just start by appreciating the reiteration you had that the new credit-based model will not have a material impact on 2025. So thank you. That said, over the medium to long term, we are really excited about the opportunity to drive top-line growth through our consumption model, through credits on AI and other forms of consumption. That said, and you heard this from Yamini, we do believe that AI is a mix of seeds and consumption through the credit system. And it's honestly early for us to really understand how that balance will play out over time. I would say as we learn, we will continue to share more information with you. Thank you. One moment for our next question. Our next question comes from the line of Ken Wong from Oppenheimer & Co. Thank you for taking my question. Kate, you mentioned moderating customer net adds down to 9,000 from 9 to 10. Is that just kind of the seasonal dynamics of the back half, or are you deliberately trimming that funnel a little bit due to the macro uncertainty? Yeah. Thanks, Ken, for the question. I think you know over the last few quarters, we've talked about 9,000-10,000 as the range to be focused on for net adds. This quarter, we came in a little bit above, which was nice. Maybe I would say the following. We continue to see healthy gross additions through all of our customer sizes. We continue to see really nice professional and enterprise net customer adds. At the starter tier, we are lapping these large cohorts of starter additions that were driven by pricing and packaging changes. And it's that expectation that that's going to continue here that leads us to point toward a net adds number that's closer to 9,000. Now, you didn't ask the question, but I'm going to answer it anyway, which is what's going to happen to ASRPC as a result because that's frankly just an output metric. In Q1, you saw ASRPC down low single digits, which was in line with what we told you last quarter. We're going to expect over the next couple of quarters that ASRPC growth is basically going to be flat. Thank you. One moment for our next question. Our next question comes from the line of Jackson Ader from KeyBanc Capital Markets. Great. Thanks for taking our questions, guys. Kate, I think part of the expectation for net retention rate to improve also has to do with the pricing adjustment, right? The pricing increase of no more than 5% upon renewal on the seats-based model. I'm just curious, given now that we're introducing more credit-based pricing on the higher tiers and uncertainty persists, any kind of update on what you're thinking in terms of your willingness or flexibility for customers coming up for renewal on the pricing increase in the second half? Yeah. Thanks very much for the question. You are right that the net revenue retention increase year-over-year is a combination of healthier seat upgrade rates and the pricing impact. Customers migrate from the old legacy model to the new seats-based pricing model at a neutral ARR. And then when they come up for their first renewal, we said we would take up to 5%. That has been going the sort of total migration of customers over to the new seats-based model has been going well, going very similar to plan. I think the expectation is that that will continue through the year. Thank you. We have time for one more after this. Our next question comes from the line of Kirk Materne from Evercore ISI. Yeah. Thanks very much for taking the question. I don't know how much to take this, Dharmesh or Kate or Yamini, but I was just kind of curious. Can you give us an update on sort of the use of AI internally at HubSpot and maybe how that's trending and if there's any sort of impact on you all from a, I guess, an OpEx growth perspective? I know there's a lot, Dharmesh, on your side in terms of coding agents and the benefits of those. We've heard a lot from some of your peers. So just wondering if you could give us a quick update on that. Thanks. Yeah, Kirk. I'm happy to take this question and answer it, Yamini. Look, we were very early in our pivot to using AI internally. Just as quickly as we pivoted from a product roadmap perspective, we've been experimenting and scaling internally. 2023, 2024 was a lot of experimentation and learning. And this year and next year is going to be a lot of scaling from the lessons that we have learned. And probably I'll highlight a few areas that we are using internally and getting a ton of value. The first one is support. Now, we've talked about support being a slam dunk use case in terms of leveraging AI. Last year, 35% of our support tickets were resolved with AI. Already in Q1, we're seeing up to 41%. And we've been able to hold headcount flat within our support organization for the last 18 months. And we've been able to drive more proactive support based on how we're doing this. So that's an area. I'd say the prospecting is another one. You can imagine AI being super easy to look at accounts and send emails to accounts and set up meetings. And in Q1, we booked nearly 11,000 meetings using AI for our sales team. So it's driving productivity there. And of course, marketing. We've been doing a lot of work internally with marketing. And we've been at the forefront of leveraging it. And we've seen clear improvements in email conversion rates, up to 80% improvement in email conversion rates through leveraging AI for personalization. And we've been able to move most of our web chats that are now being handled by AI and automation. So within our go-to-market engine, there's just a ton of experimenting and scaling with AI. And we feel really good about what that will lead to in terms of midterm and long-term efficiencies. At the same time, we're using it within our product organization. Our developers are using AI to increase their productivity. And one of the reasons we've been able to accelerate our pace of innovation over the last couple of years, and certainly this last year, has been leveraging AI. What is interesting is that the first gen of developer productivity was via Copilot types of things like GitHub. And now we're on to using Cursor and Claude Code and leveraging agents for developer productivity. And that is super exciting. So overall, we have a huge set of initiatives that we're scaling. And this is all going to have midterm to long-term impact in terms of efficiency. But we also have this added benefit of learning fast and sharing those best practices with our customers. So feel very good about it. Thank you. This concludes the HubSpot Q1 2025 Earnings Call. Thank you to everyone who was able to join us today. You may now disconnect your line.