Klaviyo, Inc. (KVYO) Q4 2025 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by. My name is Krista, and I will be your conference operator today. At this time, I would like to welcome you to the Klaviyo's Fourth Quarter and Full Year 2025 Earnings Conference Call. [Operator Instructions]. I would now like to turn the conference over to Ryan Flaim, Director of Investor Relations. Ryan, please go ahead.

Thank you, and welcome, everyone. We appreciate you joining us. Joining me today are Klaviyo Co-Founder and Co-CEO, Andrew Bialecki; CFO, Amanda Whalen; and joining us for the first time, Co-CEO, Chano Fernandez. Welcome, Chano. Andrew, Chano and Amanda will share their views on the quarter and fiscal year, and then we'll open up the line for your questions. Our commentary today will include non-GAAP measures. Reconciliations to the most directly comparable GAAP measures can be found in today's earnings press release or earnings release supplemental materials, which can be found on our Investor Relations website. Additionally, some of our comments today contain forward-looking statements that are subject to risks, uncertainties and assumptions, which could change. Should any of these risks materialize or should our assumptions prove to be incorrect, actual company results could differ materially from those forward-looking statements. A description of these risk factors, uncertainties and assumptions and other factors that could affect our financial results are included in our filings with the SEC. We do not undertake any responsibility to update these forward-looking statements, except as required by law. Andrew, that concludes my intro. We're ready to begin. Over to you.

Thanks, Ryan, and thank you all for joining us today. 2025 was a breakout year for Klaviyo. We grew revenue by 32% to $1.2 billion and delivered 14% non-GAAP operating margin. We're now serving more than 193,000 customers in over 100 countries, and we saw strong momentum across every part of the business, especially in our enterprise customer base and internationally. The future is autonomous customer experiences, where every interaction between a business or organization and its consumers will be defined and delivered by AI. Every marketing message, every customer service request, every web visitor, mobile session and even in-person experiences will be designed for the individual consumer and delivered autonomously. To allow any business to deliver autonomous customer experiences, we've built the autonomous B2C CRM. Our technology marries the customer database we founded Klaviyo on and our robust marketing messaging infrastructure with agents for marketing and customer service that will autonomously create, deliver and optimize customer experiences on behalf of a business. And this agent layer that is designing and delivering experiences to billions of consumers is trained from our deep expertise and the trillions of experiences we've delivered for businesses already. Because of these 2 things, proven infrastructure for real-time customer data and personalized experiences and now agents trained to execute marketing and customer service autonomously, we are powering the era of autonomous customer experiences and our results are showing it. In a year defined by rapid transformation, we drove 32% revenue growth for Klaviyo, with every quarter growing 30% or more year-over-year. Klaviyo customers are growing, too. And in Q4, our 10,000 largest customers grew their revenue, also known as GMV, by 11% year-on-year, which was roughly twice as much as the broader market's growth. Before I go deeper, 3 points to keep in mind. First, AI is making it easier to deliver high-quality experiences consumers want, not generic experiences, but ones that delight and convert, and we'll share a few examples of this today. Second, AI is improving the quality of experiences via personalization and training on past performance to meet the ever-increasing bar to wow consumers. We've hit a point where content and experiences generated by humans alone are not enough. Using AI is required to keep consumers engaged. Third, AI is making experiences more accessible. AI has introduced new modalities like the chat and soon-to-be-available voice capabilities of our customer agent that offer new services for consumers and businesses to meet. The upshot is that AI is an accelerant and opportunity for our customers and for us. Our customers and our incentives are well aligned. We're an engagement and revenue engine for them, and we sell Klaviyo based on engaged consumer relationships and outcomes, not on seats. AI continues to improve the speed, quality and places consumers and businesses meet. This means as our customers around the world look to use AI to meet that moment, we are scaling with them. Our ability to be the agent and the platform of choice finds its roots in how our platform was built. This is our durable advantage. At the core is the database and data infrastructure, specifically built to handle the scale of consumer data and indexing and enriching through machine learning and serving hundreds of thousands of requests per second with millisecond latencies. It allows Klaviyo to ingest, aggregate and govern first-party data in real time, so every consumer behavior, transaction, preference and consent is available to our users and now to our agents to deliver the best possible consumer experience. That foundation is coupled with our marketing platform, which not only provides high throughput, scalable systems to render messages, deliver them with excellent deliverability and enforce compliance, but also integrates with our customer data infrastructure to make last-mile personalization decisions on content, incentives, timing and channels at the moment we deliver a message or experience to an end consumer. Speed and scale matter. By keeping consumer profiles continuously updated and queryable, Klaviyo turns real-time consumer activities into tailored messages and experiences delivered instantly. This architecture enables a fundamentally different consumer experience. Here's how it works in practice. When say a consumer browses their favorite restaurant website, books a reservation, uses our chat agent to make a modification to that reservation, buys a gift card for a friend while they're there and then leaves a review, Klaviyo is compiling richer context with every interaction, who they are, what they browse, products and services they consume and how they prefer to be contacted. All of this is stored in a single profile. We also expose that progressively built profile at each step, and our system uses that context to deliver the optimal experience. That live context is immediately available to the system, allowing an agentic customer representative to answer the question, recommend the right product and follow up with a personalized message on any channel, all in real time and all without switching systems or losing context. This means the system is compounding its learning. The more our customers use our system, the smarter it gets for them and the smarter it gets for the entire ecosystem of customers. Last year, Klaviyo processed 0.5 trillion customer interactions across 8 billion consumer profiles, translating into 3.7 billion daily signals, all feeding decisions and actions that help our customers engage more precisely and drive stronger outcomes. That same foundation fuels the agent layer. We have launched agents, including our Marketing Agent and Customer Agents. We've also designed the platform to be open, meaning that customers can use the agents Klaviyo builds or bring their own. Whether it's external models like Claude interacting through an MCP server or ChatGPT through the Klaviyo app, any agent, any model with all the context from one platform. Our own agents are deeply grounded in domain expertise earned from hundreds of thousands of customers and trillions of data points that generate highly specific revenue-driving output. That combination ensures they don't generate generic output, they make impactful decisions that are trained from relevant compliant and effective customer experiences. Bringing this all together, Klaviyo's autonomous B2C CRM is like the central nervous system for AI-driven consumer engagement, connecting reasoning to real customer context and turning it into actions across channels. The result is clear customer impact. Consumer interactions running through Klaviyo generate measurable revenue for our customers, what we call Klaviyo attributed value for KAV. In Q4, KAV continued to outpace business revenue, underscoring the value of running consumer engagement on one unified platform. Customers around the world are already putting the autonomous capabilities built into Klaviyo to work. Marketing Agent is a clear example. More than half of the campaigns created by customers using Marketing Agent are now generated by AI, with many of our customers' agent-generated campaigns performing as well as and often better than campaigns created manually while taking significantly less time to launch. A great example is Adore Beauty. They're an online skin care and beauty store where growth depends on frequent, relevant communication. Using Marketing Agent, a small team can create and send fully on-brand campaigns in minutes instead of hours. The results have been remarkable. Open rates increased by 50% and revenue per campaign grew by 40%, while allowing the team to run more campaigns without adding additional headcount. We see the same positive pattern in service. Customer Agent is live and handling real customer conversations, answering questions, resolving issues, recommending products and increasingly supporting transactions, all using the same integrated data model as marketing. Since launch, resolution rates have increased by 20 points with monthly resolution volume increasing by more than 50% since Black Friday, Cyber Monday. LifeStraw is a great example. Using Customer Agent to resolve customer queries with real-time customer and product data, LifeStraw achieved a 111% increase in AI-generated sales from agent recommendations, and an over 100% increase in average order value and a 75% resolution rate over the last 90 days. Many of these interactions involve technical questions that can't be handled with scripted responses and require deep contextual understanding of their product catalog. Next, we're expanding what these agents can do and where they operate, adding more channels with e-mail and WhatsApp and beta and voice coming soon, along with more skills and more control while keeping everything grounded in the same data, delivery and measurement infrastructure. Marketing Agent and Customer Agent reinforce one another. Service interactions improve marketing decisions and marketing activity improves service outcomes. As usage grows, that feedback loop is strengthened. This is the foundation we're building on, a system that scales globally, improves with use and delivers measurable outcomes for our customers. When people think about AI at Klaviyo, they should think about it in 3 ways. First, AI is embedded directly into our infrastructure. It uses first-party customer data, like behaviors, purchases and engagement to predict how customers are likely to act and makes those insights immediately usable across marketing, service and other customer interactions. Second, AI sits on top of that infrastructure in the form of agents that do the work of marketing and customer service, creating campaigns, answering questions, resolving issues and driving transactions using the same underlying data and execution system. Third, Klaviyo is an open platform embedded in the broader AI ecosystem, providing the real-time customer context and infrastructure LLMs require. Taken together, this gives us a clear view of where customer engagement is headed. Businesses are moving towards more autonomous experiences. As they do, interaction volume and complexity increase, driving demand for infrastructure that can unify, execute decisions and measure outcomes in one place. That's the role that Klaviyo plays. We've built a platform that coordinates data, decisions and execution across the entire customer life cycle, making autonomous engagement practical, measurable and scalable for businesses of all sizes. The next era of customer experience will be autonomous, powered by unified data and real-time action. We are making that future accessible to the world's leading consumer brands. I want to thank our customers and our partners for their trust, and I want to thank Klaviyo's around the world for their dedication to our customers. We're excited to continue building, investing and leading as we bring our autonomous B2C CRM platform to businesses of every size. And with that, I'll turn it over to Chano.

Thanks, AB. I'm excited to join you, and thank you all for being with us today. One month into the role, it is clear to me that Klaviyo is built for where the market is going: real-time data, intelligence that learns from outcomes and actions across marketing and service that's increasingly autonomous. I've spent my career scaling enterprise technology businesses, and I joined Klaviyo because I see a B2C CRM category transition happening and a team that knows how to win it. Let me start with enterprise. Enterprises are rethinking how they run customer engagement because their own customers' expectations have changed. They want personalization, immediacy and experiences that feel seamless as they move across channels. When I talk to these companies, I hear a consistent need. They want one platform that can power the entire customer relationship. What they have instead are separate databases, separate customer-facing tools and separate decision engines that weren't designed to work together. The result is slower decisions, inconsistency and personalization that breaks down as their customers move across channels. The real cause of that fragmentation is yield. When data decisions and executions are disconnected, companies leave revenue on the table. This is a massive opportunity for Klaviyo, one we're already capturing and are laser-focused on expanding by proving that consolidation produces results. A great example is Cymbiotika, a natural supplement brand. They leverage segmentation tools during the holiday season to focus on their most engaged audiences. Using Klaviyo, they reduced their e-mail volume by 31% year-over-year, while increasing total revenue by 44%. This strategy nearly doubled the revenue per recipient. Another great example is Proper Hotels, which selected Klaviyo as it scaled its luxury hospitality portfolio to gain deeper visibility into its guest database, a stronger attribution across channels and more robust reporting than traditional hospitality CRMs could provide. Since migrating to Klaviyo in October, Proper has seen overall CRM revenue grown by more than 200% year-on-year. Stories like this illustrate why momentum is building and why large consumer businesses are relying on Klaviyo data, intelligence and action. Klaviyo's continued investment in AI-driven capabilities, our MCP server and integrations with LLM further reinforce our role as a long-term infrastructure partner, enabling more dynamic segmentation, faster experimentation and increasingly personalized life cycle messaging. We have our largest enterprise pipeline ever and it's across verticals and geographies. Importantly, we are seeing this pipeline convert to a more repeatable motion, clearer stage gate, tighter product alignment and faster time to value proofs. In fact, the number of customers generating at least $1 million of ARR doubled last year. Inside Klaviyo, our team is ready for this demand. Over the last several months, we have put more structure into how we work larger opportunities. We're aligning product, sales and engineering earlier in the process. This approach is allowing us to scale our motion in a repeatable way. We're also continuing to expand our partner network. This includes an exciting new partnership with Accenture, where we are combining Klaviyo Autonomous B2C CRM with their services to drive faster and more integrated customer outcomes for some of the largest brands in the world, like Stanley 1913. We're also strengthening leadership across our go-to-market team. We hired a new global Chief Revenue Officer, Eric Fearday, whom I know well and who brings significant experience in enterprise deals. Several key sales roles are also now held by sales leadership with deep enterprise experience. At the same time, we are preserving what makes Klaviyo unique, which is a product-led culture and an engineering mindset that moves fast and solves customer problems. We tell prospective customers, "You can dream it, you can build it own Klaviyo." International is another area where we're seeing increased momentum. Revenue outside the Americas now represents more than 1/3 of the business at the end of Q4. During the last quarter, Klaviyo partnered with KIKO Milano, Italy's #1 makeup brand, operating more than 1,300 retail stores across over 70 markets worldwide. Service is becoming a meaningful contributor in these markets as well. All of this ties to where customer engagement is going. The shift towards autonomous, conversational and agent-driven experiences is happening faster than we could have anticipated just a few years ago and the opportunity that lies ahead is large and exciting. Everything I've seen since joining reinforces my belief that the opportunity ahead is great and that we're still in the early stages. We are building to win. Now I'll turn over to Amanda to walk through our financial performance.

Thank you, Chano and AB. 2025 was a defining year for Klaviyo and a clear reinforcement of our position as the autonomous B2C CRM that powers more valuable customer experiences. We delivered strong growth, expanding profitability and record free cash flow. Our peak season performance demonstrates that businesses rely on Klaviyo to understand their customers in real time and act instantly across channels. For the full year, revenue reached $1.234 billion, up 32% year-over-year. International revenue growth accelerated to 42%. Our largest customers, those contributing more than $50,000 of ARR, grew over 37%. Our new service category is the fastest-growing product launch in our history. And this broad-based strength drove NRR to 110%, a year-over-year expansion of more than 200 basis points. Q4 was a standout quarter. We delivered revenue of $350 million, up 30% year-over-year, driving our annualized revenue run rate to $1.4 billion. Q4 was defined by the shift toward autonomy and the validation of our model. Customers increased spend with us because we are the engine of their growth. Customers are leveraging data and AI to move from broad-based volume to precision targeting, focusing on smaller, higher intent audiences. This shift drove superior outcomes for our customers because an automated flow drives on average over 10x more revenue per message than a static campaign. We saw particular momentum in mobile with over 29% of SMB+ customers now utilizing text and WhatsApp. Let's look at our growth engines. We are successfully delivering on our growth strategy by adding more products and channels, more countries and larger customers. AI acts as a multiplier, accelerating our momentum. In Q4, each of our growth engines delivered meaningful progress. This dynamic accelerated our NRR to 110%. NRR strength was driven by expanded usage of our e-mail and text products, cross-sell of text messaging and WhatsApp and the scaling of marketing analytics and service. There was also contribution from the profile enforcement change we made in February. Service is on the steepest adoption curve in our company's history. From $1 million businesses to $100 million enterprises, high ROI personalization and real-time action resonates at every scale. The data confirms the success of our multiproduct strategy. 60% of our ARR now comes from multiproduct customers. With consolidation accelerating, over 15% of our ARR now comes from customers adopting at least 3 products. This mix continues to shift upwards as brands recognize the strategic value of unifying on a single platform. Half Magic demonstrates this value. By consolidating marketing, service and analytics on Klaviyo, this fast-growing beauty brand drove a 5x increase in repeat purchasers and 110% growth in automation revenue over 12 months, plus 2x higher average order value from orders attributed to our customer health. We are aggressively expanding our addressable market. Q4 was a standout period for mid-market and enterprise momentum, highlighted by wins, including KIKO Milano and Bayer and expansions like TaylorMade and Nine West. Our customers with greater than $50,000 in ARR increased by 37% year-on-year to 3,912. We added 349 new -- net new 50,000 customers in the cohort, beating our previous record by over 25%. International is strong with revenue outside the Americas growing 41% year-over-year in Q4, led by strength in Italy, Spain and Portugal, alongside traction with global brands like Bayer and KIKO Milano. Our business model is fundamentally aligned with customer success. Because we price based on the number of active profiles and usage rather than seats, our revenue scales naturally as customers use more data, adopt more tools and drive more value. As customers integrate AI-driven workflows, interaction volume increases flowing through our platform and deepening the alignment between customer value creation and our long-term growth. Turning to our financial performance. Non-GAAP operating income for the quarter was $51 million, representing a 15% non-GAAP operating margin. This reflects a 900 basis point expansion year-over-year, or 400 basis points excluding the impact of last year's bonus implementation. We delivered a non-GAAP gross margin of 73% in the fourth quarter, consistent with the anticipated seasonal mix shift toward higher text messaging and WhatsApp volumes. Our underlying infrastructure efficiency is providing increasing leverage. This operational scale significantly mitigated the margin impact of text messaging and WhatsApp, resulting in a sequential improvement in the year-over-year gross margin trajectory throughout the fiscal year. Operating discipline remains a priority. Non-GAAP operating expenses were 58% of revenue, our lowest level since the IPO. AI is driving meaningful internal productivity, enabling faster development cycles without commensurate headcount growth. Free cash flow surged 61% year-over-year to $87 million. This was driven by strong flow-through from operating income and record collections, highlighting the high quality of our earnings. For the full year, the strength of our business model is clear. Non-GAAP operating income was $169 million, a 14% non-GAAP operating margin, which was up 170 basis points year-over-year. We delivered a free cash flow margin of 16%, continuing to outpace operating margins. And we surpassed $1 billion in cash on hand for the first time. Momentum heading into 2026 is strong, and we are raising our full year 2026 guidance. We now project revenue between $1.501 billion and $1.509 billion, representing 21.5% to 22.5% year-on-year growth. This is driven by high-quality expansion within our installed base and the continued scaling of new products. Importantly, our 2026 outlook is derisked. It assumes minimal revenue contribution from our newest AI and service products. We view these innovations as embedded upside that reinforce the durability of our growth runway. We expect continued operating margin expansion. For 2026, we project non-GAAP operating income of $218 million to $224 million, and non-GAAP operating margin of approximately 14.5% to 15%. Our business model enables us to reinvest in growth while simultaneously delivering profitability. We are a growth company who remains disciplined in striking the right balance, leveraging our operational rigor to fuel reinvestment, ensuring we capture the opportunity ahead and deliver growth over the long term. For Q1, we expect revenue of $346 million to $350 million, representing growth of 23.5% to 25% and non-GAAP operating income of $50 million to $53 million or a non-GAAP operating margin of 14.5% to 15%. On linearity, we expect revenue similar to 2025, weighted towards the second half due to seasonality and product ramps. Operating income will also follow a similar first half and second half cadence as 2025. As our business continues to grow and our scale increases, our visibility into the core has sharpened. Our outlook reflects high confidence in the baseline while maintaining a prudent approach regarding the timing of upside from new initiatives. In summary, Klaviyo is ushering in the era of autonomous consumer experiences with our market-leading B2C CRM. We are the engine helping businesses turn real-time data into intelligence and action. Customer experience is already becoming autonomous, unified and real time, and Klaviyo is the actionable infrastructure powering that shift at scale. Before I open it up to questions, I want to leave you with 3 key takeaways. First, AI increases the speed, quality and reach of consumer engagement, and customers are relying on Klaviyo to deliver that engagement. Second, our advantage is structural. Our infrastructure is what allows agents and AI tools to operate consistently as volume and complexity increase. And third, our model is built for this moment. Klaviyo is a revenue yield engine. As AI increases, the volume and sophistication of consumer engagement, value creation and our growth scale together. With that, we'll open up the call for questions.

[Operator Instructions]. And your first question comes from Ryan MacWilliams with Wells Fargo.

This is Cyrus on for Ryan. What was the impact from the portfolio enforcement change in the quarter? And I just have one quick follow-up on SMS.

Sure. Can you ask the question, the portfolio change?

Sure.

The portfolio enforcement change, yes.

Yes. In NRR, if we take a step back, there's 3 big impacts or 3 big drivers of NRR. The first and largest is the impact of growing usage of our platform across our e-mail and SMS products. The second is increasing cross-sell of our text messaging and newer products, including marketing analytics and service. And then we did have impact on -- or contribution from portfolio or the profile enforcement that happened in February. But that was the smallest of the 3. What we're really proudest of this quarter is the way that our customers are using Klaviyo to drive higher quality outcomes. They're seeing increases in their returns that they're seeing on each message because they're doubling down on automation and doubling down on high-value messages that drive with those automated flows up to 10x more revenue per message compared to a more static campaign. We're also seeing strong success in that cross-sell with over 29% now of our customers in the SMB+ category using SMS and our text messaging products. And service, as we've said, is off to a really strong start. So multiple levers contributing there to NRR over the coming year.

Perfect. And on SMS, how did SMS perform in the quarter versus expectations? I know the Black Friday, Cyber Monday statistics you guys put out were very strong. So how did that compare with your internal expectations? And how can we think about that going into 2026?

Yes, I can take that. So as Amanda said, people are getting more out of the profiles. They're building better relationships. So that's causing them to store more profiles with us. But SMS, very strong. Some of the things you cited. One thing that's happening with text messaging is we're undergoing a real transition here from SMS to RCS, which allows for much better experiences. So you can think about branded accounts, the ability to embed richer media, carousels, even take actions directly from within a text message. So we look at that as there's a lot of upside there. One thing I'm very proud of our team is we've gotten close to customers, and we've really designed some really compelling experiences where text messaging, I mean, not only can you now even say, browse an entire almost website, a whole catalog, we're getting close to where you even almost say buying decisions directly from RCS and text messaging. So I think there's a lot of upside to that channel. And in parallel, we launched WhatsApp generally in the last few months. And similarly, while WhatsApp is used more internationally, we're seeing a very similar pattern. The types of experiences that you can build, again, with richer media, allowing people to browse more of, say, products and services directly from a message or from a thread, that's working really well. And then the last thing, we talked about Customer Agent and this idea of having this helpful assistant that's always on that's powered by Klaviyo. Those now integrate directly into text messaging in WhatsApp. So we're also seeing a lot of consumers are asking follow-up questions and that's driving usage of both our Customer Agents as well as more adoption of those channels. And obviously, consumers are going to go where they can get the best experience. So we're kind of normalizing this idea that you can have a conversation with the business not just over a chat interface, but also in your text messages or inside of WhatsApp.

Your next question comes from the line of Elizabeth Porter with Morgan Stanley.

Awesome. This is Katie Keyser on for Elizabeth. AB, maybe on the service opportunity, can you give us a little bit more visibility into adoption and engagement trends by cohort, specifically when we're thinking SMB versus mid-market versus enterprise? Is adoption there skewing more towards the agent side relative to the help desk in any of those cohorts? And then maybe just translating that into the model. Amanda, definitely heard you that the guidance is derisked, but just any color you can provide on what level of service contribution is actually embedded within the '26 guide would be great.

Okay. So I'll take you a little bit of what I'm hearing from customers, and I'll Amanda comment on the embedding into our model. So the first with service, we talked about Customer Agent. Our belief is that every business is going to have this digital representative, this Customer Agent. And we think Klaviyo is very well positioned to offer that because of the data set that we have around who your customers are for the businesses that we work with and our ability to store that data and access in real time. And also just the broad set of customer experiences that we've already powered and using that to train and make our Customer Agent better for each of the businesses that we serve. So we mentioned some examples of businesses that are already adopting Customer Agent. We're seeing adoption from our entrepreneurs and SMB customers all the way up into our mid-market enterprise customers. Just as an example, I was talking with a customer the other day, the high-end fashion brand, and they were telling a story about how their Customer Agent closed the sale for $800 dress without anybody -- any interaction from anybody on their team. And they're starting to see that as the new normal. So we're very excited about this. I think adoption, when I talk to customers, there's really 2 rate limiters, and we're working on both of these. The first is they say, "Hey, they're not sure how to train up an agent." So when you think of a lot of companies that are understanding how to do that. We're actually building technology that, one, will teach you how to do that, but also will automatically do it. So out of the box, our agents are already really, really good. That's especially important if you think about for our entrepreneur and SMB customers and gives us a good platform for our mid-market enterprise customers where they can get a really high-quality agent in a very short amount of time. And the second one is that they're nervous about like, "Hey, the quality of the responses." So again, product that we're building in is the ability to, frankly, generate a bunch of simulated conversations and show those to our customers so they can get comfortable that the quality is really high. So as we continue to push on the product and our ability to deliver that, I think we're going to find that everybody is going to have a Customer Agent and obviously, we want it to run on top of Klaviyo. Amanda, do you want to comment on the guide?

Yes. In terms of the guide, there is minimal contribution from service built in. The way that we think about it is it's embedded upside in the model. We're seeing the strong traction and the strong customer response that AB just talked about. But we'll build that in over the course of the year as we see the results there continue to build and grow. We're just early in that product adoption cycle. So if you take a step back and think about our guide for next year, we're going into fiscal 2026 from a position of strength. Our guide is a strong baseline that reflects the momentum that we're seeing across our core business in international, in mid-market and enterprise and with new products that we're continuing to introduce. And that has the opportunity to only continue to accelerate as we gain momentum from not only our new service products, but also the new AI products that we're introducing as well.

Your next question comes from the line of Gabriela Borges with Goldman Sachs.

AB, I noticed throughout your prepared remarks, there's this theme about context and some of the proprietary context that Klaviyo has. My question for you is, what is the limiting factor to an LLM or an AI native company spinning up alongside of you and somehow replicating or otherwise accessing that context through API such that it diminishes your moat barrier over time. Maybe just a little bit on the limiting factor to someone else abstracting your context.

Yes. Thanks for the question, Gabriela. So I get this a lot of, hey, "How easy is it to rebuild Klaviyo or what gives you that advantage?" So let me say this, like we've looked at -- we designed Klaviyo to be ready for this agentic era or as we talk about this, everybody is building agents and this autonomy that's going to sit on top of the infrastructure that exists. There's 2 things that give us a lot of confidence here. The first is we already have this great data set. We talked about 0.5 trillion data points just last year, trillions over the course of Klavio's history and obviously, it's expanding every single day. We said almost 4 billion new signals that we get every single day. And the signals for us are really -- it's not just data points about what customers are doing, but it's also how they're interacting with the experiences that Klaviyo powers. So you can think about every message, whether it's an e-mail, a text message, a WhatsApp, a mobile notification, every experience is our Customer Agent. Not only do we have AI that's designing that and tuning it, but it's also learning from that data set about what works best and then using reinforcement learning and other techniques to improve itself. So first is the data. The second is, frankly, our infrastructure. When you think about our database, we designed it specifically for these real-time use cases. It's very hard not only to store data, but to index it and serve it such that every time a page loads, every time a message needs to be sent, you can query specific attributes about a person or specific things that they've done and make decisions in real time in milliseconds and integrate that into the experience. That -- combined that data infrastructure, combined within our messaging infrastructure, its ability to make sure you're compliant, pick the right channel, embed that personalization all in real time. I think those 2 bits, the data plus the infrastructure are really what set us apart.

Your next question comes from the line of Raimo Lenschow with Barclays.

Perfect. I'll stick to the one question. Chano, It's good to connect again. The one question I got from investors though was that we all know you as kind of co-CEO, Workday, et cetera, which is kind of more upmarket enterprise. Can you talk a little bit about how that fits in here at Klaviyo, which is kind of more slightly lower end, not lower end, but you know what I mean like more SMB space. Like how is the skill set that you bring kind of help here?

Thank you, Raimo. It's great talking to you again and connecting again, right? I guess there would be 2 dimensions, right? One is, obviously, my background is more scaling up enterprise businesses. The second one, I would say, is operational strength and how we're bringing some better operational cadence in terms of scaling up more predictable the business as a whole, right? So I would say my focus is on both, right? We commented that on the enterprise, we are really excited about the opportunity. We talk about our best pipeline ever. And also last year, we doubled the number of customers, over $1 million ARR, and we are putting the experience in terms of the go-to-market leadership. We're investing on the compliance and the governance to serve. And I believe the momentum is very clear, right? In terms more on the experience itself on the operational side, I would say, first, we're focusing on using AI internally as much as we use externally, and there is some amazing productivity gains that are already seen in some of our numbers. We are focusing on creating much better operational alignment across product engineering and go-to-market. So we keep decisions going fast and execution going at speed of light at these days, right? And I just mentioned as well, getting up to that enterprise readiness. So all that, hopefully, that contributes. Saying as my previous co-CEO experience, I'm clearly very excited to be part of the journey here and a company that I've been familiar with now from the Board for the last few years as well.

Your next question comes from the line of DJ Hynes with Canaccord.

Congrats on the really strong results, and I appreciate all the comments around kind of the infrastructure advantages that you have. AB, I was hoping maybe you could talk a little bit about the Accenture partnership and kind of what they see in Klaviyo. I mean when I hear of these GSI relationships, I think other services work. That seems a little bit at odds with kind of your core model. So just talk about kind of what they see in Klaviyo and kind of how you can leverage their reach to help build the business.

Yes. Let me make a couple of comments about Accenture.

DJ, could you just mute your line? I think there's some echo. Go ahead, AB.

Okay. I'll try. So with Accenture and how we're thinking about partners generally, look, this AI era and this building out customer experiences that are designed and delivered by AI, we think it's going to require a whole bunch of new skills and new services that tens of millions of businesses around the world. And certainly, the largest enterprises in the world are going to need help adopt it. So I'm very excited with the work that the conversations I've had with Accenture about what these new practices will be. And I think when it comes to our Marketing Agent and designing subagents within that, that will help look through your marketing, define gaps, help surface those then create the right creative all the way to our Customer Agent, helping that tune various experiences maybe based on what's worked in the past. We think there's a ton of new work for all of our agencies and SI partners. But anyway, Chano's been working closely with Accenture. Chano, anything you want to add?

Yes. DJ, great talking to you again. So I would say, clearly, Accenture is starting to provide services around marketing reinvention and service reinvention. Some of those services, they need a good powering on AI solutions, and they see us as a great partner to bring that to market. They see a great opportunity for companies that are looking to solve more for that fragmentation. I was talking on my script in terms of closing the loop for the entire customer relationship experience, and they see Klaviyo as a great solution to provide that one. And clearly, as I mentioned before, we're moving up more towards the enterprise, though to be clear, we still are fully convinced that it's going to be our core business and our more SMB business is a great cornerstone for us, and it's going to keep thriving. But obviously, enterprise opening us complete new opportunity. We're partnering with the likes of Accenture. It's going to be instrumental for both of us. So yes, tons of timing, stay tuned, I guess, we'll have much more to share over the coming months.

Your next question comes from the line of Samad Samana with Jefferies.

I guess, AB, I want to maybe revisit Gabriela's question a little bit because I think part of what we're all trying to figure out is if the marketer starts increasingly interfacing with Klaviyo via ChatGPT or another interfaces maybe making a tool call and just how the value will still ultimately be being provided by Klaviyo. But I guess the perceived value question of what the marketer thinks is creating the value for them, how do you contend with that and make sure that the user knows that it's actually Klaviyo and that you make sure that the economic benefit also accrues to you? And then I have one follow-up.

Yes, sure. So an analogy that we've been using internally and Chano's been using is when we think about our infrastructure, it's kind of like how good is the engine underneath that is both understanding who your customers are and actually provide those experiences, whether it's messaging or some of the conversational architecture that we have now with our Customer Agent and obviously marrying those 2 together. And I think we look at our infrastructure, I mentioned how tuned it is for the scale and really the real-time nature of consumer interactions. We just look at it as like we have the engine of like an F1 car, and you can't just go -- you can't just take any old engine and like plug it into something, make it look shiny and have it work. So with -- on top of that -- and the other thing I didn't share with Gabriela is like, "Look, we talk to customers and say, "Would you want to swap out Klaviyo's system for something that was maybe cheaper to run?" And the answer for all of them is no. And the reason is because the ROI for Klaviyo is so good people don't want to miss out on stunning customer experiences -- end consumer experiences, and they don't want to miss out on the increased engagement and increased revenue that, that comes with. That's been a very durable pattern, durable thing I've seen when I talk to our customers. So if you flip it back then, like I put this in the prepared remarks, but we actually look at AI as if it's sort of the driver of that car, well, gosh, I mean, how do we get our agents and potentially other agents using more of Klaviyo's underlying engine and infrastructure. And that's where we come back to what we found with AI is and our agents is it's making it cheaper to execute. We're finding a lot of customers who now because of AI, they have the time where they didn't have the time and the space. Now they can get to things because they can take an idea and turn it into marketing much faster. And it's also increasing the quality. I mentioned how our -- because our database, our underlying customer data infrastructure is married with the marketing messaging and we can see what's working. we're able to hint back to agents, both Klaviyo's agents and then in the future, potentially other agents, what's going to work best for each consumer. So we're able to inject that personalization, that context in at the last moment. And that's having -- we're seeing that have tremendous uplifts. I mean we talked about these results. The fact that we're seeing open rates increase not by a little bit, by 50%, revenue increasing by a similar amount. This is, I think, the future. I think the idea that you would run either marketing or customer service agents and you wouldn't run them on the best possible infrastructure just doesn't make sense because at the end of the day, like we're trying to help drive great customer experiences that translate directly into dollars and revenue for the businesses that we serve.

Very helpful. And then maybe, Amanda, just a follow-up for you. On the OpEx growth, it slowed quite a bit. I was just -- especially R&D. Is that mostly like internal efficiencies? Was there anything timing related in terms of the timing of hiring? Just how should we think about that both in the quarter, what led to that slowdown? And then how we should think about it going forward?

Yes. Let me take that just to start qualitatively. So -- and Amanda will get to the numbers. Look, we've had an initiative internally about working AI-first since last fall. I'll tell you like I myself like just building product coding. I mean, the amount that you can do now with AI on 2 fronts. The first is -- the thing we -- ED and I in the early days, we moved very fast. We were great builders, but we're able to go much faster and not 20% faster. I'm talking 5, 10, 20x faster than the early days. It's fun. I can sit down now and build whole features, whole prototypes in an evening that used to take us weeks. And that's an advantage that our entire product and engineering team is embracing. And the second thing I'd say is everybody now gets to act as a developer. We have a number of internal systems that thousands of Klaviyo's rely upon that weren't built by engineers. Or they were built by engineers, but maybe not in the traditional sense. There are other folks at Klaviyo that are now doing engineering, thanks to the progress that we've seen with some of the coding agents. So look, we just said that, "Hey, all of that is stuff that we're really going to push on." And Amanda knows, like our roots were -- we were a small handful of folks that had great unit economics, great efficiencies in our early days. And now we're benchmarking ourselves against what the best AI companies are doing and looking at that as the new normal. But Amanda, anything else you want to add?

Yes. The only thing that I would add there, as we look forward into next year, we're really pleased with the trajectory that we're seeing on our overall operating margin. And I think that is because we are a high-growth company who's delivering expanding profitability while making high-return investments. And the great thing about where we are right now as an era and as an industry is, as AB said, we can make those high-return investments and be so much more efficient in how we're building and how we're scaling. So we are bullish about the ability that we have to continue to drive that profitability while continuing to expand our margins 100 basis points or more over the coming year.

Your next question comes from the line of Tyler Radke with Citi.

Chano, congrats on the co-CEO role. Nice to see you again. A question on go-to-market. Just given some of the leadership changes and go-to-market changes, how are you structuring and incentivizing go-to-market this year? Any other changes we should be thinking about? And I guess, Amanda, are you building in any conservatism just given the moving pieces on that front?

Yes. Thank you, Tyler. Great catching up. I mean, well, first of all, I'm lucky I found a great foundation here, both in terms of product innovation, customer satisfaction and I would say, talent. Obviously, as in any place, there are usually opportunities for improvement. And I think we saw an opportunity for improving in terms of the go-to-market, mainly across tidying up execution and rigor and discipline. And at the same time, potentially, as we say, there are many disjointed experiences that don't provide that full overview in CRM legacy systems that I think are ready to be ripped out, and we want to capture that opportunity, and we're greatly positioned to doing so. So that's why we're doing some of these changes from a go-to-market perspective. Me being right, that should show up over the next few quarters in terms of increased win ratios and clearly a reacceleration. And that is my expectation that's yet to be proved, right? Clearly, in terms of how we are aligning on incentives, I mean, at the end of the day, we try to align to what it makes sense for how our customers are seeing the benefits that we're bringing. So obviously, as we've said before, we are more a consumption and we're more an outcome-based. So that is as well how our teams are incentivized. So we're very focused on this Klaviyo's attributed value, Klaviyo's attributed revenue that last year came up to close to $80 billion that we are driving to our customers and very focused on producing those outcomes. That is what matters, right? Clearly, that as well should manifest if we're doing our homework that even the ratios being okay today in terms of having 60% of our ARR becoming multiproduct. In that case, it is 2 or more products. I believe that the opportunity for upsell and cross-sell, we're doing our homework is much better than that one. And I think we are having the right team to execute on it.

Thanks. And Tyler, as it regards to the model, we have the benefit that Chano has been involved in the business, both on the Board side as well as getting ramped up to speed as supporting us on an interim basis last fall. So he is coming in as a strong leader with clear plans, and we have strong visibility into the places where we're going to be making those changes and continuing to improve over the course of the coming year. So the main thing that I would point to as you think about the investments that we're making in go-to-market over the coming year is that they're going to be -- continue to be with a strong focus on unit economics that we've always had as a business. right? It goes back to AB and Ed. We tend to look at our CAC payback. Now as we're expanding into the mid-market and enterprise, we look at the LTV to CAC. But we see great opportunities ahead, and we see strong returns from those investments that we're making, and they're baked into the outlook that we have for the year of continuing to expand our operating margins.

Your next question comes from the line of Matthew VanVliet with Cantor Fitzgerald.

I guess as we look towards what the customer and marketing agents that you're offering could feed into the model, whether it's this year or into the next several years. Curious if you can share any sort of metrics around adoption by your customers? Just how many are using it in any capacity? Where do you expect that sort of percentage of your customers to at least be piloting and testing it out to be over the next several quarters?

Awesome, I can take that. So while we're not sharing the percentage of customers yet, I can tell you it's growing rapidly. And I mentioned some of the things that we're solving, especially for our entrepreneur and SMB customers to help them adopt faster. In a lot of cases, it boils down to, hey, "How do I train a model and how do I trust it?" But again, these are things we're really, really good at. If you go back to Klaviyo's early days, we had to teach people how to do more personalized messaging, marketing and make that almost out of the box. And now we're doing the same thing with agentic capabilities. I'll reference back to a data point that we watch a lot, which is like for people that are adopting our agent products, whether it's our Marketing Agent or our Customer Agent, how often are they coming back? And I'll go back to the data point we shared around our Marketing Agent, where for folks that try it and adopt it, we're finding that they're using it repeatedly. And in fact, it's starting to really eat into their share of how they generate and create marketing overall. So that -- it very much leads me to believe that like in the near future, we're going to have a large percentage of our customers who aren't going through and pointing and clicking and sort of doing creative design all on their own, they're going to do it in large part with AI. Just an example from just this week, I was talking to a customer that's playing with our Marketing Agent product. And he was telling me about how he was designing a campaign and his brand has a big Spanish-speaking audience. And the Marketing Agent actually understood that and generated versions that were not just in English, but in other languages as well. I mean these are the kinds of things that are now happening with agents that are running on top of our platform. And because we have the customer context to say what geography or locale somebody is in, their preferences, maybe things that they told us, in this case, about language, we're able to just automatically inject those and our agents are able to pick up on that. So I think you're going to see an increasing percentage of our customers are starting to really rely even for the majority of their usage of Klaviyo on our agents. And then also, we're going to see just a growing percentage, even give it a shot and become full-time adopters.

Yes. I think just to add on that one, we provided a couple of data points like customers that have taken or adopted Marketing Agent, 50% of their campaigns, they're already doing it with the Marketing Agent and seeing better results. Some examples back on the autonomous resolutions on service for customers like LifeStraw, 74% of those coming up. Or Harney & Sons, like 77% of autonomous resolutions within 60 days of adopting service agents. So there are already some great data points that, of course, as product keeps maturing and our go-to-market teams keep cranking on it, very exciting what's coming ahead.

We have time for one more question, and that question comes from Derrick Wood with TD Cowen.

This is Cole on for Derrick. The large customer adds was super strong. Can you just talk a little bit more about the strength there? And then specifically, how much of this is cross-sell versus actually going and landing new customers of this size?

Yes. Great question. Most of those customers tend to be net new lands in terms of the large-sized customers. And clearly, they expand over time, but it's more on the net new logo. I think it comes to 2 points, right? It's clearly how we're changing some of the practices in terms of focus more on the leading indicators like pipeline, qualification, and strength of that pipeline. Same as I said before, how we're tidying up those cadences as well with beyond sales themselves, with articulating the value proposition with product and engineering teams altogether, getting on to it. And clearly, I would say it's the teams that know to and to understand how to liaise and land these customers in terms of the complexity of the sales cycles. But I think, as I said before, qualifying the cycles it's very important, I think that we are on the right opportunities, and we understand those that we are into and we are winning, and we're just creating the sales place, increasing the activity, and certainly, as well as starting to come with partners like we mentioned before, Accenture and some others, which are great, right? I think it's important to highlight as well that, that is becoming across multi-verticals, and we're seeing customers now in Europe like Bayer or Lindt. So it's very international as well, TaylorMade and some others that are tremendous marquee names that are starting to trust in us. And I think that will create -- hopefully create a flywheel effect in terms of we're building our brand as we are ready now for enterprise and with the investments we're doing in compliance and governance, plus some data centers for some of the international customers, feel very strong about the value proposition and where we are going. So yes.

Ladies and gentlemen, that does conclude today's conference call. Thank you for your participation, and you may now disconnect.