Braze, Inc. (BRZE) Q2 2027 Earnings Report, Transcript and Summary

Welcome to the Braze Fiscal Second Quarter 2027 Earnings Conference Call. My name is Leila, and I'll be your operator for today's call. [Operator Instructions] I'll now turn the call over to Christopher Ferris, Vice President of Braze Investor Relations.

Thank you, operator. Good afternoon, and thank you for joining us today to review Braze's results for the fiscal second quarter 2027. I'm joined by our Co-Founder and Chief Executive Officer, Bill Magnuson; and our Interim Chief Financial Officer, Pankaj Malik. We announced our results in a press release issued after the market closed today. Please refer to the Investor Relations section of our website at investors.braze.com for more information and a supplemental presentation related to today's earnings announcement. During this call, we will make statements related to our business that are forward-looking under federal securities laws and the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements include, but are not limited to, statements regarding our financial outlook for the third quarter and fiscal year ended January 31, 2027, the anticipated benefits from and product advancements due to ongoing developments in BrazeAI technology; our anticipated customer behaviors, including vendor consolidation and replacement trends and their impact on Braze, our potential market opportunity and our ability to effectively execute on such opportunity; the execution and anticipated benefits of our share repurchase program and our long-term financial targets and goals, including our expectations regarding our profitability framework. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations and reflect our views only as of today. We assume no obligation to update any such forward-looking statements. For a discussion of the material risks and uncertainties that could affect our actual results, please refer to the risks identified in today's press release and our SEC filings, both available on the Investor Relations section of our website. I'd also like to remind you that today's call will include certain non-GAAP financial measures used by management to evaluate our ongoing operations and to aid investors in further understanding the company's fiscal second quarter 2027 performance in addition to the impact these items have on the financial results. Please refer to the reconciliations of our non-GAAP financial measures to the most directly comparable financial measures calculated in accordance with U.S. GAAP included in our earnings release under the Investor Relations section of our website. The non-GAAP financial measures provided should not be considered as a substitute for or superior to the measures of financial performance prepared in accordance with U.S. GAAP. And now I'd like to turn the call over to Bill.

Thank you, Chris, and good afternoon, everyone. We delivered a strong second quarter, generating $227 million of revenue, up 26% year-over-year and 8% from the prior quarter. We also continued to realize operating efficiencies, improving non-GAAP operating margin by over 600 basis points year-over-year. Dollar-based net retention for our large customer cohort, who spend at least $500,000 annually continued to inflect positively, rising 100 basis points to 112% in the quarter. And we generated a record second quarter free cash flow of $22 million. Our results, robust pipeline, channel expansion and customer adoption of AI solutions provide us with the confidence to raise our revenue guidance for the third quarter and the full year. We are also pleased to raise our operating income guidance for the full year and are on track to deliver at least 400 basis points of operating margin improvement, in line with our long-term profitability framework. Brands are adopting ever more sophisticated strategies and racing to deploy AI-driven solutions to leverage their first-party data and direct-to-consumer relationships, and this is reflected in our results. Bookings were strong in the quarter, driven by competitive takeaways from legacy marketing clouds, point solutions and a vendor consolidation motion that continues to gain traction. Net customer additions rose 76% sequentially, up 15% year-over-year, while customers spending at least $500,000 annually increased by 12% sequentially, up 28% year-over-year. Notable new business wins and existing customer expansions included Boots Thailand, Chime, David Jones, Foxtel Group, Insurify, Omaze U.K., Property Finder and Wilson Sporting Goods, along with many others. Brands across a diverse set of industries and geographies continue to graduate from legacy platforms to Braze. These included a global quick service restaurant, a bank in APAC, a European retailer and an American challenger bank. Upsell momentum remained robust, driven by expanded channel adoption, deeper data platform integrations and growing engagement with our rapidly advancing BrazeAI capabilities, including Agent Console, Decisioning Studio and Operator. While AI monetization is still early, enterprise customers are realizing ROI using our solutions and adoption is accelerating. Paid adoption of our AI tools, Decisioning Studio, Agent Console, AI Item Recommendations and our Predictive Suite reached roughly 1/3 of our large customer cohort in Q2, up about 900 basis points from Q1. And that adoption is changing how customers send messages and personalize product experiences. As teams put Agent Console and Decisioning Studio to work, they pair the added intelligence with upgraded engagement strategies, experimenting with more advanced cross-channel programs and deeper personalization. We expect this adoption trend to compound as marketers are pushed to get more sophisticated in an AI-driven world and as our composable AI architecture keeps making Braze smarter and easier to use. We also recently signed a 3-year strategic collaboration agreement with AWS. It establishes a dedicated co-sell motion, a joint go-to-market commitment and incentives for AWS sellers to bring Braze into their accounts. More of our customers are choosing to procure Braze through AWS Marketplace, drawing on cloud commitments they already have in place. And this agreement makes that path easier while extending our reach across international markets and industry verticals. It also deepens our work with AWS on data and AI, including the model integrations that let Agent Console run on Amazon Bedrock. You'll see us build on all of this over the coming quarters with more to come at Forge. That partnership lives at the infrastructure layer. But the question underneath most investor conversations right now is what sits above it. As AI models improve and converge, where does enduring value get created? The market's answer is coalescing around the Harness, the layer of context management, workflow orchestration, proprietary data, feedback loops, permissions and evaluations that turns raw model intelligence into reliable business outcomes. We agree, and the market has just found a word for what Braze has spent 15 years building for marketers. The 4 foundational strengths I've walked through on our last 2 calls, the Braze data platform as the foundation for first-party context, our vertically integrated data and decisioning stack, our composable AI ecosystem and our position as both a revenue engine and mission-critical infrastructure are the anatomy of an Agentic Harness. And the same vertical integration that made our stream processor, what we believe is the highest scale, most performant engine in our industry, has now become something bigger, a complete environment where marketer AI is built, deployed and trusted with production workloads. And when we talk about this in technical terms, for a marketer, the Harness has a much simpler purpose. It's what closes the gap between having a great idea and being able to ship it. Start with what the agents know. Personalization rooted in a brand's own data is the difference between a front desk reading from a script and a concierge who has known the guest for years. Engineering that knowledge securely and at massive scale is what the Braze data platform was built for. Among our large customers, 87% use the Braze data platform, 91% have adopted Canvas and 57% use cloud data ingestion, our reverse ETL suite, which enables bidirectional data flow between Braze and the major cloud data warehouses, including AWS Redshift, Databricks, Google BigQuery and Snowflake. The context system has 3 connected parts. The Braze data platform holds the first-party data, Canvas carries the live operational context of every customer journey and BrazeAI puts both to work. Agent Console agents executing on that context in the moment and Decisioning Studio learning from every outcome to optimize the decisions ahead. This quarter, we continued to widen those context streams. Cloud data ingestion now connects customer data warehouses directly into Decisioning Studio to power one-on-one personalization without middleware and our new knowledge sources feature adds semantic search to Agent Console. As this engine runs, we're not just ingesting data, we're also producing tremendous amounts of it. Currents keep expanding the scope of what it streams out with user profile updates now in beta and the entire BrazeAI catalog generating output all along the way, content variants, experiment results and observability data on every decision the AI makes. This data is so valuable that adoption of our export features is near universal. Amongst our large customers, 96% use our SDK, 98% use our rest APIs and 87% recapture the data that Braze generates through either Currents streaming or Snowflake data sharing. That depth flowing in both directions is what gives the context layer its power. Next is how the work gets done and how it gets better. We've said since inception that Canvas is a programming environment, and we built it like one, application context, a scalable distributed runtime, debugging and observability and full versioning and collaboration controls. For years, that depth has given Braze a higher ceiling than our competition. And Operator is now unleashing marketers to reach it, letting all of Braze shine through. And the timing could not be better because coding is the single skill that the foundation models have advanced most quickly. These models love to code, and we handed them a proper development environment to build in. Because we vertically integrated Operator skills with the design and metadata of the dashboard itself, for Braze workloads, Operator's bespoke intelligence is racing ahead of the foundational models that it's built on, providing proprietary advantage for our customers to further enhance their productivity. Operator now builds Canvases end-to-end, configuring the Agent Console and Content Optimizer steps inside them, and it can navigate the entire dashboard on its own. Describe your goal from any page and it goes where the work needs to happen. Virgin Media O2, one of the U.K.'s leading telecom providers, used Operator to cut their campaign quality assurance process from 6 hours down to mere seconds and their build iterations from 2 days to about 15 minutes. And the learning half of this loop is compounding just as fast. Content Optimizer is replacing manual A/B testing with always-on optimization across hundreds of message combinations all at once. Motorway, a car selling platform in the U.K., recently used it to mix and match subject lines, body copy and call to action in a single campaign, driving 114% lift in clicks and driving 37% more car valuations by their consumers, their key conversion metric. Meanwhile, Decisioning Studio is steadily taking responsibility for making compound decisions of what to send, when, on which channel and with which offer, all optimized against the precise business outcomes our customers choose. Then the part that decides whether any of this reaches production, trust. AI is compressing every step of the campaign launching path from brief to build to approval. The brief is becoming a prompt. The build is becoming a conversation with Operator, but compression only helps if your standards survive it. Well, you can dump a 100-page brand book and every historical campaign result into a giant context window and hope for the best, it's slow, expensive and produces lower quality, higher volatility outcomes. The answer is to make the standards part of the machine. That's Agentic Standards now in beta. Teams encode their brand guidelines, content standards, tracking mechanisms and compliance rules once. Every campaign then gets checked against them automatically with a pass or fail. Operator fixes the issues that it finds, and audit logs record every check and every response. The brand book becomes enforceable standards. The approval chain becomes an audit trail, automated nearly everywhere, but still with human sign-off wherever the customer wants it. Last September at Forge, we asked a room of senior executives to raise their hand if they let an AI system send content to their customers without review. No one was ready yet, but we knew that wasn't a technology gap. It's a trust gap and closing it is a product problem that we are solving. And as AI advances in cybersecurity capability, the challenge of trust is extending beyond the dashboard. In the post-Mythos communication environment, agentic traffic is rising fast. There's more noise for brands to cut through and more sophisticated threats probing their defenses. In an environment this adversarial, your pipes better not be dumb. They need to be secure, high performance and automatically optimizing. The major inbox providers are deploying increasingly stringent AI-driven filtering that rejects generic content. The answer is personalization built on context. Protecting deliverability is no longer a niche issue for the marketing team. It has become a whole company revenue protection challenge. The same is true for anti-fraud protections, where tighter sending controls and AI-enhanced countermeasures against traffic pumping have made us better watch guards for our customers. And our technical account managers are increasingly the guides for this new terrain, helping ensure our customers' data flows, product integrations and messaging architectures are designed securely, navigating them through the many governance and security gates required to ramp up their agentic workloads safely. That backdrop also explains the evolution of our services revenue, where 2 distinct trends are playing out. The first is mechanical. Starting with last year's pricing and packaging changes, we began breaking out services more explicitly in new business order forms and at renewal. Under our prior structure, customer success entitlements were primarily bundled into platform fees. As customers move to the new structure, some revenue that was previously bundled as subscription is now packaged as professional services and year-over-year comparisons of the 2 lines reflect this reclassification as cohorts migrate. As we stated on our last earnings call, approximately 90% of our professional services revenue is recurring, recognized ratably over the same term as subscription revenue. The remaining 10% of our services revenue comes from the minority of customers who opt for lightweight onboarding directly from Braze, although more than 80% of new Braze customers now leverage our growing ecosystem of Braze-certified agency partners as they kick off long-term services engagements that will maximize their ROI from Braze over the long term. The second driver of these trends is simply demand growth, and you just heard the reasons for it. As AI scales volume and risk together, customers are investing in the expertise that protects the ROI of their primary revenue channels. And where does the Harness go from here? We are excited to see AI acceleration of marketer productivity building along 3 dimensions as we invest further in Operator specifically and BrazeAI more generally. First, seeing Operator take on larger and more complex projects. The multistep work that used to require a specialist blocking off their week and extended collaboration with engineering is now built automatically. Second, moving from executing tasks to keeping watch. Think of Operator as a supervisor that never goes off shift, monitoring the AI systems, making optimization and decisioning choices, watching channel volumes, broad signals and deliverability trends and flagging anything that drifts off course. And third, AI that works with whole teams instead of individuals. Most AI productivity gains so far have gone to individual users, but customer engagement is a team sport that crosses marketing, data and engineering and supporting how those teams work together is where we see the next wave of value. This is composable intelligence, agents that do the work inside Braze while also working directly with the people and systems at the brands we serve. And customers are adopting these features with enthusiasm. In the last 90 days, almost 80% of Braze accounts engaged with Operator more than 10x and over half of those turned it into a habit, chatting with Operator more than 100 times in that same time period. Operator has also rapidly driven record declines in customer support tickets, letting our premium support stay focused on high complexity, high-value challenges, delivering stronger customer outcomes and higher post-sales efficiency at the same time. But this is not just a reflection of cases. Operator is helping our customers build, and it's already a sophisticated builder in its own right. At a hackathon that we hosted alongside City x City Sydney, a team from an Australian food and beverage brand put Operator through its paces. They were already using Braze Surveys to follow up with customers after their coffee purchases, and they wanted to upgrade those follow-ups with stronger personalization based on the feedback itself while introducing new channels to their most engaged customers. Operator took them through the whole setup. It found the right survey data, took them into Agent Console to build a sentiment agent and configured and tested the outputs. Then it drafted a new Canvas, added the agent step and templated the results into a follow-up text message. Customers who expressed negative sentiment were routed accordingly, and the agent summarized each customer's specific feedback, so the follow-up spoke to the exact concern they raised. If the coffee was cold, the message said so. By leveraging Operator for this multistep build, idea to implementation took only about 10 minutes. Marketing technology has always forced a trade-off. Every gain in power came with added complexity and every ambitious idea took specialists, time and patience. We believe that we are now breaking that trade-off. Braze is getting both more powerful and easier to use at the same time. When a marketer watches an idea, go from creative spark to working prototype in just minutes, they remember why they fell in love with this work. And that's the flywheel that we're building. The product itself inspires the builder spirit. Our hackathons, events and community feed it, and ambitious creative marketers bring their ideas to life faster, amplify their performance with leading-edge AI and carry their whole teams farther. Braze was born out of a hackathon. True to our roots, we've been hosting them all over the world, and they're driving great early adoption for Agent Console. Over 650 customers have now made their first agents with more being added every day. The hardest part of this industry has never been the idea. It was the distance between having one and being able to ship it. That distance is what this Harness collapses because a Harness after all, is not what keeps you on the ground, it's what lets you fly safely. Our technology is made for this moment, and so are the marketers using it. Our job is to teach them to fly again. They take it from there. We have a lot more to share about Operator, Content Optimizer, Agent Console and Decisioning Studio at Forge in just a few weeks, and we hope to see you there, either in person or on the live stream. While we won't be hosting a full Investor Day this fall, we will be welcoming investors for a reception on the evening of Tuesday, September 29. Contact Investor Relations for more details. Thank you for your interest and support in Braze. With that, I'll turn the call over to Pankaj.

As Bill stated, we reported a strong second quarter with revenue increasing 26% year-over-year to $227 million, driven by a combination of existing customer contract expansions, renewals and new business. BrazeAI Decisioning Studio contributed $6.6 million of revenue in the quarter, implying an organic year-over-year revenue growth rate of 24%. Subscription revenue remains the primary component of our total top line, contributing 91% of our second quarter revenue, while the remaining 9% represents a combination of recurring professional services and one-time configuration and onboarding fees. Approximately 90% of our professional services revenue is recurring with revenue recognized ratably over the life of the contract. As Bill noted, demand for these services is growing as AI adoption scales. Total customer count increased 15% year-over-year to 2,789 customers as of July 31, 2026, up 367 from the same period last year and up 76 from the prior quarter. Our total number of large customers, which we define as those spending at least $500,000 annually, grew 28% year-over-year to 361. And as of July 31, 2026, these customers contributed 65% to our total ARR compared to a 62% contribution as of the same quarter last year. Measured across all customers, trailing 12-month dollar-based net retention was 110%, while trailing 12-month dollar-based net retention for our large customers was 112%, up from 111% in the prior quarter. Expansion was again broadly distributed across industries and geographic regions. In the second quarter, our total remaining performance obligations was $1.1 billion, up 27% year-over-year. Current RPO was $691 million, up 24% year-over-year. The year-over-year increases were driven by contract renewals and upsells and the signing of new customer contracts. Non-GAAP gross profit in the quarter was $156 million, representing a non-GAAP gross margin of 68.6%. This compares to a non-GAAP gross profit of $125 million and non-GAAP gross margin of 69.3% in the second quarter of last year. The decrease in year-over-year margin was driven primarily by high premium messaging volumes and the addition of Decisioning Studio headcount attributable to cost of revenue. Total non-GAAP operating expenses were $134 million, 58.9% of revenue compared to $119 million or 66% of revenue in the prior year quarter. The improved efficiency reflects our disciplined approach as we effectively balance investing in our growth priorities while working towards achieving our long-term profitability targets. Non-GAAP operating income was $22 million or 9.7% of revenue compared to non-GAAP operating income of $6 million or 3.4% of revenue in the prior year quarter. Non-GAAP net income attributable to Braze shareholders in the quarter was $21 million or $0.19 per share compared to $17 million or $0.15 per share in the prior year quarter. Now turning to the balance sheet and cash flow statement. We ended the quarter with approximately $414 million in cash, cash equivalents, restricted cash and marketable securities. Cash provided by operations during the quarter was $24 million compared to cash provided by operations of $7 million in the prior year quarter. Including the cash impact of capitalized costs, we achieved record second quarter free cash flow of $22 million compared to $4 million in the prior year quarter. We expect our free cash flow will continue to fluctuate from quarter-to-quarter given the timing of customer and vendor payments. Finally, we are pleased to report that we completed our $50 million accelerated share repurchase program in August, buying back approximately 2.1 million shares. Approximately $50 million remains on the Board's authorization. We regularly review our capital allocation priorities and will look to return capital to shareholders when appropriate. Now turning to guidance. For the third quarter of fiscal 2027, we expect revenue to be in the range of $229 million to $230 million, which represents a year-over-year growth rate of approximately 20% at the midpoint. In line with normal seasonality patterns for our business, third-quarter operating income will be affected by the cost of Forge, our annual customer conference, as well as several global customer events scheduled during the quarter. Third quarter non-GAAP operating income is expected to be in the range of $16 million to $17 million at the midpoint. This implies a non-GAAP operating margin of approximately 7%. Third quarter non-GAAP net income is expected to be $15 million to $16 million and third quarter non-GAAP net income per share in the range of $0.13 to $0.14 per share based on approximately 114.5 million weighted average diluted shares outstanding during the period. For the full fiscal year 2027, we expect total revenue to be in the range of $910 million to $913 million, which represents a year-over-year growth rate of approximately 23% at the midpoint. Fiscal year 2027 non-GAAP operating income is expected to be in the range of $75.5 million to $76.5 million. At the midpoint, this implies a non-GAAP operating margin of 8%. Non-GAAP net income for the full fiscal year is expected to be in the range of $72.5 million to $73.5 million, and net income per share is expected to be $0.64 to $0.65 per share based on a full year weighted average diluted share count of approximately 114 million shares. By embedding AI-driven intelligence into the core of our platform, we continue to elevate what's possible in customer engagement. Looking ahead, Braze remains fully committed to an ambitious innovation road map while staying disciplined and firmly on track to meet our long-term financial targets. And with that, we'll now open the call for questions. Operator, please begin the Q&A.

[Operator Instructions] Our first question will come from Ryan MacWilliams with Wells Fargo.

It seems like the big shift this year in AI has been long-running agents, and it's great to see you guys take advantage as it's now generally available that Braze Operator can chain together a multistep work from a single prompt. But how do you think these more powerful models and agents can help unlock new use cases for Braze from here with customers?

Yes. Thanks. And we're super excited about this. I mentioned in the prepared remarks that I think it's the first time in our history that we're combining together both enhanced capability with enhanced ease of use and higher productivity, higher ability to build, et cetera. A lot of the structure and the capabilities that we put around the Canvas environment are being highly effectively utilized by Operator. And it has been fantastic to see as well that Operator is providing differentiated capability on top of the foundational models, abilities to use Braze in a more generic fashion. The way that, that is being driven is because of the vertical integration of the development environment into Operator skills and intelligence. So when Operator runs, it has a deep understanding of all the features available to it in its development environment, similar to how Cursor would integrate with an entire IDE. And it has full access to all of the metadata around how the -- our customers' integrations of Braze are set up in the first place. And so it can see all the different attributes and events and the full data schemas, and it understands what the goals of the campaigns are. And through Agentic Standards, it has increasing levels of understanding of compliance measures, brand voice, style, et cetera. When you pull all of that together in a vertically integrated environment like Braze, you just get this incredible building capability. And so on the one hand, we're really excited and seeing customers able to really unleash their builder spirit, bring their creative ideas to life more quickly. We also are really excited about what this means from a product-led growth standpoint. when you look at our pricing and packaging evolution over the last few years, a big part of what we've been trying to accomplish with the shift toward action credits and being able to use these more fungible credits more flexibly for customers is that it also allows for us to use Operator to point their adoption toward more of these more advanced features that are also monetized by Braze. And so whereas in the past, we would release a new feature or we would enhance a new channel, and we had to wait for the very kind of human manual process of being able to enable a broad field around that, enable the customer community around it, have them then build up the -- get over the activation energy of trying it out for the first time, potentially needing to build additional resources, et cetera. Instead, all of that happens instantaneously and automatically. When we build out new capabilities, when we do something like roll out Content Optimizer into beta, we're doing the same thing with Decisioning Studio Go right now, we're going to have a lot more to talk about at Forge. The ability for adoption to happen instantaneously upon those becoming available is unlike it has ever been before. And so we're really excited about Operator coming in, driving enhanced productivity, getting our customers all the way to our capability ceiling and really breaking through it with their creativity right out of the gate. And then that also being able to drive really strong monetization and adoption outcomes for us because it's effectively like product-led growth on steroids where Operator itself just drives the adoption agenda.

I appreciate that. That reflective cycle could be pretty powerful. And then I know you mentioned it's early for AI monetization. And look, you got really sophisticated customers that want a high degree of certainty before they put AI into production. But in terms of AI translating into like near-term revenue for Braze, are there certain types of customers you're looking at who are now more forward in deploying AI use cases? Or is it products like Operator that might get more adoption? Like how do you think we start to see more tangible results of AI increasing Braze consumption?

Well, Operator, we've been seeing week-over-week increases in penetration across the customer base and its stickiness is incredible. I mentioned the acceleration of a customer that tries it out 10x, more than half, and that's a number that's been increasing every week as well, accelerate to using it more than 100x. And we've got people that are doing all of their work in Operator now. And we're just seeing that as soon as people get a taste of it that it really drives immediate adoption and habit forming around that. And what that means is that we've also been able to start to build out Operator defaults that are driving additional adoption of these more advanced AI capabilities. That's something that we actually only started within the last few weeks. But I mentioned the build example during the prepared remarks from our City x City Sydney hackathon that we did back in August. And that -- in that example is the automatic creation of Agent Console. But we've also said for a long time that there's capabilities we have like Personalized Paths in the past and now Content Optimizer and Decisioning Studio Go, where it's just simply the case that basically every campaign that gets created should be using these AI optimization capabilities. But the reality is that in the past, due to bandwidth concerns or creative production requirements or just a lack of knowledge or confidence in utilizing these new features that adoption often lag the availability, but we're seeing that collapse down now as well. And so our ability to, for instance, roll out Content Optimizer and just turn that into the default way that a push notification campaign or an e-mail campaign gets built in the first place is -- that provides a new capability that we're just really excited is going to continue to accelerate adoption. And so I think not only do you get the full capabilities of Braze being on display for a higher percentage of our customer base more quickly, which helps with our differentiation and helps with customer ROI, but it just helps us drive the monetized adoption of these new capabilities as soon as we build them.

Your next question will come from DJ Hynes with Canaccord.

Bill, I want to ask about cRPO growth. I mean, obviously, it moderated a little bit this quarter, but dollar-based net retention remained quite healthy. Can you just help me unpack kind of the bookings mix this quarter? And how much of that cRPO deceleration reflects maybe a smaller renewal cohort versus any changes in new business or expansion activity?

Yes. Well, there is seasonality in cRPO. Q4 and Q1 are always our largest renewal quarters. Similarly, we lapped the OfferFit acquisition within the quarter. And so you're seeing that effect in cRPO as well. But overall, very happy with the quarter. I mentioned in the prepared remarks, it being a strong bookings quarter from a competitive win rate standpoint. We continue to see strong durability in the sales force's productivity. We're also seeing great diversity across the upsell motion with customers adopting new use cases and channels, expanding their action credits, allocations and continuing to drive new adoption of Decisioning Studio use cases. And we're seeing the full display of all the different dimensions of how Braze drives upsell across the customer base. And so when we look at the overall strength in the business, we were excited to be able to flow through the full beat and be able to raise both Q3 and raise the full year. And so excited to see the strength continue into the back half. We are really excited for the setup for the back half of the year as well, strong pipeline and just a lot of opportunity out there, and Forge is still to come. So we historically always see Forge as a strong accelerator of pipeline, and we're feeling good about what we're sitting on going into it. And so we're excited about the back half setup.

Yes. Good. And look forward to seeing you in Vegas. Maybe a follow-up just on the AWS agreement. Obviously, you talked about a dedicated co-sell motion and incentives for their sellers. How meaningful can these partnerships become as a source of pipeline and distribution? And how long does it take for that to ramp?

Yes. So we've been continuing to prioritize building out stronger alignment amongst, in particular, our very large technology partnerships. And you heard me mention all of the different Braze data platform partners across the major data warehouses and hyperscalers. With AWS, the growth in the AWS Marketplace purchasing across our customer base has been astounding over the course of the last couple of years. And so I think that what we see with the SCA is that we've got -- over the last couple of years, we've gotten a lot of the foundations built out, and now this provides the formalization that will help us further accelerate that. And it ties into good timing with the rest of our go-to-market verticalization efforts where we have driven a lot more organized vertical alignment within the Braze selling organization over the course of the last roughly 4 to 6 quarters. And that is creating better opportunities for us to align with the likes of AWS, Databricks, Snowflake, et cetera, who themselves are also vertically organized. And so I think that was something that created more of a partnership alignment challenge in the past, but it's one that we've gotten over. One other great thing about this quarter's results is that it was one of the most consistent regional performance sales results that we've seen in the last couple of years, we saw strong contributions from all of our major regions around the world. And one of the great things about strategic alignment partnership with organizations operating at the scale of AWS is that we're able to leverage the benefits of those globally. And you've seen us add new data centers over the course of the last couple of years as well in places like Korea, Indonesia, Australia, Japan. And those have all been done in partnership with AWS as well. And so bringing all of that together, we're excited about being able to accelerate on top of a strong foundation.

Next, we'll go to Raimo Lenschow with Barclays.

Bill, the -- on that cRPO and the bookings number, that's where I get most of the question. And we're living in this new AI world, like what are you seeing in terms of like pipeline, there's this argument of crowding out spending, et cetera. Listening to you, it doesn't sound like it, but could you just kind of double-click on that one more time, please?

Yes. I mean I think the biggest thing about our pipeline is that there's still a huge amount of addressable market that is living on legacy technology and Braze is their AI bet. And so we are -- we certainly see customers being cost conscious. As people are adopting Agent Console, they are certainly doing the calculations to figure out what it's going to cost them in advance. There's not just kind of a reckless running into the adoption of these technologies. But we built out capabilities like Agent Console being cognizant of the -- of customers' needs to budget and control these costs as well. We actually think that it's a critical component of the Harness. And I walked through all the different capabilities within Braze that make up the Agentic Harness for marketers and billing and cost controls and having the right observability and understanding to that is certainly an important input into adoption and scaling of new use cases. And so we are seeing that. But I think that the biggest underlying trend that we're seeing is that organizations are seeing that they need to have the combination of being able to have the full first-party context that is actionable with advanced AI within a platform that provides them the right guarantees and scalability and security and performance to be able to bring to life these more advanced AI capabilities and drive new strategies within their organizations. And that's exactly what Braze has to offer them. And so we're continuing to orient our go-to-market priorities around being able to drive that legacy replacement cycle. And we are excited to see more and more customers adopting Braze as exactly where they're making their AI investments. The pace at which customers are eagerly adopting and just the feedback that we get from customers around them being able to bring to life visions that they've had about how they would be able to run their customer engagement programs and what's now possible for them within Agent Console and how quickly they're able to build that utilizing Operator. We feel it compounding together, and we're excited about what that means.

Okay. Perfect. Yes, makes sense. That's really helpful, Bill. And then one quick number question. You were really much better this quarter on operating profitability. If I look at guidance, though, it looks like it's slightly below consensus. Some of that is timing. There's Forge as well in Q3. Is there any other factors that kind of drove the better this quarter and slightly below next quarter?

The only thing is, yes, I mean, the timing around Forge in Q3, we also are -- we're excited to be able to be investing with new sales capacity heading into next year as well. And so we're starting the hiring push to be able to have the right seller capacity as we start next year now, and we're investing in that. And so we didn't -- we were excited with the Q2 beat. We continue to see strong operating income results. We raised the full year operating income margin percentage for -- as a result of that Q2 beat. And we're going to continue to balance investment into the back half of the year with this continued strong performance and year-over-year improvements.

Our next question will come from Scott Berg with Needham.

Bill, you seem pretty pleased with the initial adoption of a variety of your AI modules here. You talked about some accelerating traction there. But as you've seen those customers adopt those modules, is it also driving expanded use of the core Braze platform? I mean when I look at things, whether it's Decisioning Studio, et cetera, I can see a real scenario where someone is using, it could be more profiles, more channels, more flex credits as they're able to create some of these campaigns more quickly. Just didn't know if you're seeing any of that tangential impact yet.

Yes, we definitely are. And one of the things that we're seeing in the hackathons as well is people adopting things like multichannel strategies as well as multi-variant experimentation strategies with more eagerness. Those are the types of strategies that I think historically have always been a good idea for people to adopt. But because they required more work or more creative production, et cetera, they just didn't see the adoption that their performance potential deserved. And so the most exciting thing, I think, that we're seeing from the Operator build-outs is just that there's been a lot that we've built in the last 5 to 8 years, in particular, that hasn't seen the level of adoption that it probably deserved to given what it had to offer to customers. And we're now seeing that gap get closed. And so that is resulting in customers utilizing more advanced features, expanding to more new channels more quickly and driving our automated experimentation capabilities. And as I mentioned before, we're in the early innings of starting to have those more advanced ways of building Canvases and of driving cross-channel strategies and then driving experiment automation as well as decisioning, just by default and making that be the way that new campaigns and new Canvases get built in the first place. And we're excited about what that means from a long-term retention as well as customer expansion standpoint because we already know that customers that adopt Braze with 3, 4, 5 channels as well as using our more advanced features that they have higher DBNR than other customer cohorts do, et cetera. And what Operator has the potential to, therefore, do is to make nearly every customer look like those that we already know have higher levels of retention and customer growth potential.

Very helpful. And then from a follow-up perspective, if I go back to my notes when you acquired the OfferFit a year ago, I think you all talked about an ASP uplift from that product being, I believe, it was $300,000 or $400,000 kind of an average level deal. Is that what you're still seeing today? Or is the uplift maybe something different than your initial expectations?

Yes. So the use case price range that we talked about around the acquisition is around $250,000 to $300,000, I believe. And we've seen the pricing power around that maintain itself. We are also building out Decisioning Studio Go, and we're going to have more to share about that at Forge. And that is, of course, that is a self-serve offering that will be able to be deployed more quickly and be able to be priced and paid for with action credits. And so that allows for a customer to adopt it without the larger upfront costs around it. Now of course, that is a more constrained deployment of the model. It doesn't have as much flexibility or as much power as Decisioning Studio Pro, but it still represents incredible capability to drive uplift for customers, especially if they're not using experiment optimization in various ways already today. And so our vision for the future of decisioning is one where we have multiple price points as well as different pairings of expert services on top of that in order to drive varying levels of performance optimization by taking advantage of other complexities that exist in the more advanced forms of the underlying reinforcement learning model. And a big part of the development road map for Decisioning Studio since the acquisition, of course, in addition to integrating it into the Braze data platform and into the Braze dashboard has been the build-out of Decisioning Studio Go, which is a deployment at a completely new price point with higher levels of accessibility around the customer base. And so we're excited to be able to have both ends of that spectrum available to customers relatively soon. You're going to hear more about that at Forge, as I mentioned. And that's going to allow for the opportunity for those high use case uplifts that we mentioned in the part of the customer base for whom that is the right performance ROI optimization for them and then to be able to introduce decisioning as a concept to the entire customer base very quickly through Decisioning Studio Go.

Next, we'll hear from Brian Peterson with Raymond James.

So Bill, I wanted to understand maybe more qualitatively on the uptick in AI adoption. Is there anything that you can share in terms of what types of customers, what industry, larger, smaller that are leaning into some of this AI adoption and what may still be on the come?

Yes. So we're seeing -- I think, in general, larger teams are adopting Operator in more cases at higher rates because of the fact that their build processes are already quite complex. However, what we're seeing in the trend line on that is that Operator is trending toward universal adoption. We certainly have run into a number of organizations either in different geographies that are a little bit more reticent to adopt AI quickly or large enterprises that might have AI adoption consoles or other sorts of heavier weight processes. But once people start to get a taste for it, they're really running at it. And so I think that we're -- within the next quarter or 2, we're just going to be talking about Operator, something that everyone is using and everyone is using at high rates. Within Agent Console, those are places where we're doing the work of getting customers into hackathons, building community around that, getting them to bring ideas to life, doing the education. And we're starting to see those efforts compound. Definitely, the kind of blockers that are in front of people to be able to get new use cases scaled and into production, they vary a lot from company to company. Some of them, it's cost concerns. Other places, it's access to data. In other places, it might be brand and kind of brand controls and things like that, really handing over that autonomy to be able to produce new content. But as we are rapidly iterating with customers that are building and they're doing it with us, we're getting that customer feedback back. The road map is moving really fast. We are able to build simultaneously with capabilities like Agent Console and Content Optimizer and Decisioning Studio Go with Operator involved as well. So as we get new feedback in and as we build new features, Operator becomes aware of those features immediately and is able to help customers adopt them more quickly. And so we're seeing a great feedback loop around that. It is still -- there are -- these are new concepts for a lot of people. And so it is still taking time. And in some cases, we're going customer by customer to train them on these capabilities. But man, when you see people's eyes light up as they get to bring an idea that they've had for 2 years, we just had a customer last week sent us an effusive note because they've had an idea sitting on the shelf for literally a decade, and they were able to build it in Agent Console and do it in just a few days, and it's driving a completely new campaign strategy for them. And so being able to see example after example like that has been really rewarding, I think, for the whole Braze team and has been an important catalyst around the Braze community, and we're really excited to see what Forge brings to continue to provide another boost to that adoption.

Great. Yes, that will be great to hear about that out in Vegas, Bill. But maybe just a follow-up. I know you've had a transition from some of the customer success revenue from subscription to professional services. Can you remind us how big was that impact? And any help on trying to kind of frame the growth rates over the next few quarters for ProServ?

Thanks for your question. Yes, I'd expect that the professional services contribution is going to range between 9% to 10% moving forward. We continue to see the migration of our customers. We still have them legging out of the legacy pricing and packaging into our new packages. We're about 50% through our customer base in the migration. I'd expect about 80% of the remaining to migrate into the new packages over the next 6 quarters. But also, as Bill indicated, some of the demand that you're seeing for pro services is a result of the AI adoption as customers start to hit more premium inboxes. There is the AI deliverability and cyber risks. And so there is a continued demand for AI services.

Our next question will come from Parker Lane with Stifel.

Bill, you talked about having really great diversity in the upsell motions of the business and higher DB&E characteristics or DBNR characteristics in the customers leaning into AI. Looking at total DBNR of 110%, I think that was consistent sequentially. How is that informing what you think that figure can look like going forward? Do you expect some of these new capabilities to help you remain at that level? Is there an opportunity to expand the corporate rate even higher than that?

Yes. We definitely think that customer growth has room to grow over time. We are seeing customers adopting a broader spectrum of channels and the opportunity for us to monetize a higher percentage of the sends that customers have by invoking more advanced AI in the messaging send flow is an exciting one. And it's how you see our road map oriented. We're closely tracking the percentage of sends across different channels that are invoking AI before those message sends happen. That's a number that is growing rapidly, and it's something that we're going to continue to keep as a North Star for the coming years. And we did a lot of work to be able to get ourselves into this action credit universe over the course of the last few years, and that really set the stage for us to be able to build and take advantage of this more quickly from here.

Got it. And I remember last quarter, you were talking about compressing the time lines for Decisioning Studio and AI launches through FB&E investments. You also mentioned greater demand for ProServ now. How do you feel from a capacity standpoint on where you're at today? And how does that flow through to your expectations for gross margins, in particular for the rest of the year?

Yes. So we're feeling good about the ability to deliver with Decisioning Studio. I also mentioned the Decisioning Studio Go has a lot lower requirement for services. It's nearly self-serve. Customers obviously need to be set up with Braze already in the first place to be able to utilize it. But after that, their ability to self-serve both the setup and be able to access the reporting and all that over time is something that doesn't require the forward deployed services to anywhere near the same level. And similarly, we've done a lot over the course of the last -- or like throughout this year in order to make sure that our recruiting pipeline, our training pipeline and being able to bring those field deployed data scientists up to speed quickly is all there. And those services are -- and the delivery personnel are also benefiting from productivity enhancements of their own through the capabilities that Operator is bringing them. And so you're seeing the -- as well as the integration of Decisioning Studio into the Braze data platform that I mentioned earlier. And so just a whole bunch of different dimensions that are helping us continue to improve the overall picture there. And across the BrazeAI product family, there is an accretive benefit to gross margin. You saw a strong gross margin percent result in the quarter. You should still expect to see the same seasonality that we have on an annual basis where the holiday quarter will certainly have a lower gross margin percent due to enhanced messaging volumes. And it is also the case that the headwinds from the growth in premium messaging is certainly there and mixes into the overall gross margin percent story. But there's a lot of work that we're doing to the positive as well. All the new product families that we've been working on across BrazeAI and over the last couple of years with new channel expansion as well as some of the BYO options in terms of how customers pay for certain channels as well as Agent Console, all of those provide gross margin accretive benefits. We've also done a lot of work with our finance team and our engineering teams to be able to do vendor optimization as well as performance optimization. I like to say that fast is my favorite feature because it provides strong -- it provides better outcomes for customers and better scalability, but it also, of course, saves us money on the gross margin side, and that has been a focus area for our R&D team throughout the year. And so you're seeing the puts and takes across all those different efforts as well as the continued underlying growth around the premium messaging channels. And so we're going to continue to work and focus on the gross margin percent. As I said, you should still expect the normal seasonality in Q4 that you've seen over the last several years, but it's going to continue to remain a focus area, and there's a lot of great work going on there.

[Operator Instructions] And our next question will come from Derrick Wood with TD Cowen.

Bill, the number of net new customers has bounced around the last several quarters. Could you update us on kind of how you're looking to balance targeting enterprise versus mid-market opportunities? And then just specifically around 7-figure deal activity. I mean, Q4 was very strong. I think Q1 was seasonally softer. How is Q2? And how are you feeling about large deal pipeline activity as you look into second half?

Yes. So generally, and you would have noticed this in the prepared remarks, that we've been really excited about the continued growth potential and acceleration in the $500,000-plus customer cohort. We continue to see Braze's differentiation really building huge amounts of value for these more complex and larger scale organizations. And I think that's something that you'll continue to see be part of our strategy. It's also well aligned with the opportunities that we have around the broader legacy replacement cycle. Seasonality is always going to trend Q4 to be a strong $1 million-plus quarter just due to enterprise budget cycles. We also, as a January 31 fiscal, we capture both the end of the calendar year and the beginning of the next calendar year's budget cycles in our Q4. And so you see the -- and that has a strong effect on that enterprise and large deal seasonality. And so I would expect to see the same thing as you -- as we get into the back half of this year. And like I said, really strong pipeline, and we're excited to see Forge be it -- provide its normal pipeline acceleration as we get into the back half of the year.

Our next question will come from Arjun Bhatia with William Blair.

Bill, it sounded like competitive dynamics are looking quite favorable. I think you called out good win rates and vendor consolidation. I'm curious how much you're hearing sort of your AI capabilities or AI in general come up in new customer RFPs. Is that sort of the primary driver that customers are looking at? And how often do you even hear customers think about DIY from an AI perspective as one of the alternatives they're considering?

Yes. I mean, first of all, it's -- it is in every single RFP, and it is a focus area in every single deal cycle. It is also an area where we are able to really shine because of our ability to deliver across all the different dimensions of marketer productivity and marketer experience as well as decisioning and all the great capability in Agent Console and Canvas. And now with launches like Agentic Standards as well, that's helping on the QA side of the house to allow for customers to be able to build more quickly and build with higher levels of confidence. And so all of that, I think, is coming together in strong ways to deliver on what the market is demanding. We also -- we've been doing a lot of work on our MCP server as well and some of the other agentic workflow capabilities that allow for customers who are doing more DIY to be able to utilize Braze as the infrastructure upon which they do that. We have a lot more to share about that at Forge coming up, which we're excited about. But I mentioned earlier that when you look at this problem space, I think there's a lot of conversation, which involves something like vibe coding on top of a set of dumb pipes, if you will. And I think that the reality in our environment, and this is becoming increasingly clear to more customers is that the idea of building on top of dumb pipes is just not one that's going to drive any amount of revenue for you because the space is not about sending messages. It's about customers actually seeing messages and then responding to those and that resulting in the business goals that you have. And that requires full end-to-end capability. It requires strong deliverability. It requires strong optimization. We had a customer actually in the quarter who had been moving ahead with a kind of vibe coded content delivery, et cetera. They quickly got themselves into full deliverability trouble, and we had to pair our deliverability services with them to ensure that they were actually -- that they were able to get their reputation back into a good place. And I think it's a great example where kind of driving into this more naively or with a more limited capability set is going to allow you to trigger messages or to get them sent. But for you to really accomplish your customer engagement goals is a more advanced set of requirements. And I think that we're obviously seeing more and more customers that are demanding that in the right ways. And there's always going to be a build versus buy conversation in marketing tech. There always has been, there always will be. But I think we feel really good about the offering that we're providing, the flexibility with which it can be implemented and wielded by organizations regardless of kind of where they are on that spectrum of how much they want to build versus how much they want to build on top of smart infrastructure. And Braze will provide -- we provide services across that spectrum. We always have. And I think that the conversation that we're seeing around build versus buy right now is it's a little bit elevated from the past, but not by much. And for every one example that I have where people are chasing their vibe coding dreams, we've got 10 more for whom Braze is their AI bet. And we're excited to see that opportunity, and that's where we're orienting our go-to-market.

There are no more questions at this time. I'd now like to turn the call over to Bill for closing remarks.

We're really excited to see you all in a few weeks at Forge. We've got a lot more to share with our customer community. And I thank you for joining us for today's earnings call, and we'll see you all later.