Snowflake (SNOW) 2027财年第二季度业绩电话会：产品收入增长加速至37%

Snowflake 2027财年第二季度业绩电话会议强调了产品收入增长的加速、CoCo和CoWork更广泛的应用以及核心数据平台的持续扩张。管理层上调了全年产品收入和Non-GAAP营业利润率指引，同时下调了产品毛利率预期，以反映AI工作负载占比的提高。

核心要点

• 第二季度产品收入达到14.9亿美元，同比增长37%，标志着连续第三个季度实现增长加速。而在2026财年第四季度末，产品收入增速为30%。
• Non-GAAP营业利润率同比提升超400个基点至15%，这得益于收入增长和严格的人员规模管理。
• 管理层将2027财年产品收入指引上调至60.7亿美元，相当于同比增长36%。预计其中约1个百分点来自Observe。
• CoCo客户账户数突破9,100个，本季度净增超过2,000个新账户。CoWork扩展至5,800个账户，环比增长近11%。
• Snowflake第二季度末拥有14,554家客户，本季度净增692家新客户。净增客户数同比增长32%。
• 管理层估计，AI产品贡献了近期增长加速的大约一半，数据迁移和其他产品也对消费量构成了支撑。

核心财务数据

业务与运营表现

管理层将AI描述为Snowflake按需消费模式的“乘数”。AI项目将新的工作负载引入平台，Snowflake的第一方AI产品吸引了更多用户，而AI应用者对底层数据平台的消费量也更高。

CoCo和CoWork仍是这一战略的核心。Snowflake表示，客户正在数据迁移、供应链、财务、销售、风险管理及其他工作流中使用这些产品。在Snowflake内部，CoCo帮助将搜索优化收归内部，省去了每年40万美元的代理机构支出，并将关键词研究时间从约10小时缩短至20分钟。

管理层表示，最近的增长加速是广泛存在的，而非集中在AI原生企业。管理层将大约一半的加速归因于AI产品组合，包括CoCo、CoWork、AI功能、文档处理、机器学习、Notebooks以及Cortex AI Gateway。更快速的数据迁移以及使用Streamlit或React构建的应用也作出了贡献。

客户工作负载部署持续增加。部署在Snowflake上的客户用例数量同比增长89%，而每个客户经理负责的用例数增长43%。Snowflake还在2027财年上半年将330多项功能投入全面上架（GA），比上年同期增加35%。

数据共享仍然是平台的另一个驱动力。Snowflake表示，43%的客户与至少一个稳定边缘共享数据，支持跨组织协作。管理层还报告各区域表现健壮，未发现明显的区域差异。

管理层业绩指引

公司根据观察到的消费模式制定预测，并表示其预测方法未发生改变。管理层还重申了在2028财年第四季度实现GAAP盈利的目标。

风险与关注领域

• 快速增长的AI工作负载目前边际贡献率较低。这类收入占比的提升促使Snowflake将2027财年Non-GAAP产品毛利率指引设定为74%。
• AI模型成本仍是运营中需考量的因素。管理层正通过模型选择、开源模型和自动任务路由来平衡输出质量、成本与治理。
• Snowflake预计预订额（Bookings）将越来越集中于第四季度，因为客户继续倾向于在第四季度续约。
• 管理层表示，观察到AI采纳者的平台消费量更高，但尚未准备好量化具体的提升幅度。管理层的信心来源于群组行为（Cohort Behavior）和新用例的广度。
• 指引仍与观察到的消费行为相关联，这使得客户部署和扩展工作负载的节奏成为一个重要变量。

分析师问答亮点

增长的质量与持续性：管理层表示，增长加速涵盖广泛的客户群体，并非集中在AI原生企业。CoCo还包含成本管理功能，旨在帮助客户识别低效查询和闲置数据仓库。

AI与整体平台的贡献对比：Sridhar Ramaswamy估计AI产品贡献了大约一半的加速。其余部分反映了包括数据迁移、Notebooks、应用程序以及核心平台增长在内的领域。

模型中立性：Snowflake将获取前沿模型、开源模型和专有模型的能力视为竞争优势。Cortex AI Gateway可以根据客户策略和性能数据将任务路由至不同模型。管理层还看到了针对特定业务场景运行开源模型推理和后训练模型所带来的潜在效率提升。

加快客户激活速度：Snowflake跟踪新客户达到已购买消费额80%的速度。管理层表示，随着AI工具加速迁移和项目部署，近期客户群体的这一指标有明显改善。

企业工作流机遇：管理层表示，CoCo和CoWork正在将Snowflake的触角从传统数据团队扩展至CFO、CRO、CMO、CEO及其他业务用户。在Snowflake自身的财务部门内部，该产品的普及率几乎达到100%，涵盖交易审批、税务、会计、内部审计和资金管理。

数据架构与应用：管理层坚持认为，企业更倾向于将数据整合在中央平台上，而不是将其复制到多个应用数据库中。Snowflake计划通过包括混合表（Hybrid Tables）以及直接托管在平台上的应用程序等功能，支持更多分析和运营应用。

业绩电话会议完整文字实录

---

完整财报电话会议逐字稿

管理层陈述

Operator

Good day, and welcome to the Second Quarter FY '27 Snowflake Earnings Presentation. Today's conference is being recorded.

At this time, I would like to turn the conference over to Katherine McCracken. Please go ahead.

Katherine McCracken

Good afternoon, and thank you for joining us on Snowflake's Second Quarter Fiscal 2027 Earnings Call. Joining me on the call today are Sridhar Ramaswamy, our Chief Executive Officer; Brian Robins, our Chief Financial Officer; and Christian Kleinerman, our Executive Vice President of Product, who will participate in the Q&A session.

During today's call, we'll review our financial results for the second quarter fiscal 2027 and discuss our guidance for the third quarter and full year fiscal 2027. During today's call, we will make forward-looking statements, including statements related to our business operations and financial performance. These statements are subject to risks and uncertainties, which could cause them to differ materially from our actual results. Information concerning these risks and uncertainties is available in our earnings press release, our most recent Forms 10-K and 10-Q and our other SEC reports.

All our statements are made as of today based on information currently available to us. Except as required by law, we assume no obligation to update any such statements. During today's call, we will also discuss certain non-GAAP financial measures. See our investor presentation for the definitions of the non-GAAP financial measures and a reconciliation of GAAP to non-GAAP measures and business metric definitions, including customer count and adoption. The earnings press release and investor presentation are available on our website at investors.snowflake.com. A replay of today's call will also be posted on the website.

With that, I would now like to turn the call over to Sridhar.

Sridhar Ramaswamy

Thank you, Katherine, and thank you all for joining us today. We are in the midst of a once-in-a-lifetime technology shift and Snowflake remains at the center of the enterprise AI revolution. AI is fundamentally changing how enterprises build, operate, and make decisions. To stay competitive, every organization faces a new imperative, become an Agentic Enterprise and do it quickly, safely and cost efficiently.

Snowflake is making this transformation a reality. We bring together the core elements of an Agentic Enterprise, a governed data foundation, access to leading AI models, deep application workflows and the unifying agentic control plane that orchestrates across these elements to turn intent into governed action. By putting intelligence work at scale, our customers are building faster, executing more efficiently and reimagining their businesses in ways that weren't possible before. Put simply, the Agentic Enterprise runs on Snowflake. And the traction is translating into strong business performance as evidenced by our Q2 results.

Product revenue came in at $1.49 billion, with growth accelerating to 37% year-over-year, marking our second consecutive quarter of record sequential dollar growth. After exiting Q4 of last fiscal year, at 30% year-over-year growth, we have now added 7 points of acceleration in just 2 quarters. And with our continued focus on executing the discipline and operational rigor, our Q2 non-GAAP operating margin expanded by more than 400 basis points year-over-year to 15%. Thank you to all of our Snowflakes for the hard work and dedication that made this performance possible.

As these results convincingly demonstrate, AI is compounding Snowflakes advantage across 3 reinforcing dynamics. First, AI is bringing new workloads onto the platform. To power their AI initiatives, enterprises need a governed unified foundation for data in context and companies across industries are turning to Snowflake to power that foundation. Second, our first-party AI products, CoCo and CoWork continue to see rapid adoption. As customers build and deploy agents on Snowflake, we are expanding our role into the agentic control plane and creating new opportunities for growth.

Third, AI activation continues to lift overall platform consumption. Customers using AI on Snowflake consume more across the data platform, creating a structural multiplier for our business. Together, these dynamics show how the Agentic Enterprise has created a powerful flywheel across our business. And that flywheel is accelerating.

At the heart of this moment is the continued strength of our core business. Snowflake now provides the data and AI foundation for 14,554 customers around the world. Customers continue to turn to Snowflake because our AI data cloud is easy to use, seamlessly connected for collaboration and trusted with enterprise-grade governance and security. This quarter, we added 692 net new customers, including 14 from the Global 2000, representing a 32% increase in net new customer additions year-over-year.

At the same time, some of the world's most recognizable enterprises are deepening their relationships with Snowflake. Companies like BlackRock and Block are running more of their mission-critical work on Snowflake and in several cases, adopting CoCo to move faster. The pattern is consistent. The more our customers build on Snowflake, the more they lean in. In fact, 65 customers have now crossed $10 million in trailing 12-month product revenue, demonstrating how our largest customers continue to go all in on Snowflake.

Part of our strength is in extending our customers reach to the critical data that sits outside of their organization. Currently, 43% of our customers share data on Snowflake with at least one stable edge, demonstrating Snowflake's role as the circulatory system of the modern enterprise. We enable data, applications and AI agents to move securely and seamlessly not just within, but across organizations. In fact, credit chose snowflake for our data sharing capabilities, which now facilitate privacy-safe ads measurement.

And as customers move quickly to modernize their data estates and establish a strong contact player for AI, more and more customers are migrating workloads to our platform, a process now massively accelerated with AI. For example, one of the largest Australian banks migrated its financial crime platform to Snowflake, processing 17 billion transactions and delivering 10x faster credit performance. Now they're building AI agents on Snowflake to accelerate the migration of the rest of their data estate and automate legacy data discovery and mapping.

As AI strengthens demand for our core platform, it is also expanding Snowflake's opportunity to deliver a new generation of AI-powered products and experience. Because Snowflake sits at the center of our customers' data, business context, AI models and workflows, we are uniquely positioned to become the governed control plane for the Agentic Enterprise. Our breakout AI products, CoWork and CoCo bring that vision to life. They provide a government layer users across the business from knowledge workers to builders can put the full power of their enterprise context to work, all with simple conversational language. With CoWork and CoCo, customers are reimagining some of their most critical business processes from supply chain operations to enterprise-wide sales motion.

[ Saari ], whose risk intelligence supports Fortune 100 enterprises and national security agencies chose Snowflake to rebuild its global data infrastructure and cut costs by more than half. Its engineers are now using CoCo to accelerate the migration of 12 billion records into an AI-ready foundation. And as more customers see what's possible that this technology, adoption continues to build. CoWork expanded to 5,800 accounts, up nearly 11% quarter-over-quarter. Meanwhile, CoCo continues to see rapid adoption surpassing 9,100 accounts and adding more than 2,000 net new accounts in this quarter alone.

We have customers like 1Password, the security company trusted by more than 200,000 business, which choose Snowflake for our CoCo capabilities. CoCo enables their team to move key data pipelines into Snowflake quickly, playing foundation for their data and AI work. And the world's #1 job site, Indeed, has rolled out CoWork and CoCo across its data teams and integrated Snowflake into core data architecture, citing lower cost and greater efficiency, which compounds at the scale that they operate in over 60 countries and 28 languages.

But the opportunity goes beyond adoption. By making it possible to build, collaborate and interact with enterprise data through conversational language, CoWork and CoCo are bringing in entirely new users to Snowflake. Within accounts adopting these products, we see a step change in user growth as Snowflake reaches new lines of business and expands its footprint within existing teams. As we continue to develop CoWork and CoCo as agentic control plans, we are also building the broader platform enterprises need to put AI to work at scale. Model choice gives customers the flexibility to select from leading frontier and open models and evolve their approach as the market changes.

Post training lets them adopt models to their specific data and business context and agent observability analytics give customers full visibility into what their AI is doing, how it's performing and what it costs. And to help our customers optimize cost, performance and speed, we've introduced Cortex AI Gateway, which dynamically route each task to model based on customer defined policies and real-world performance data with cost and governance controls built in. As those economics improve, customers can deploy AI more broadly and with greater confidence, creating another catalyst for adoption and consumption on Snowflake.

Cortex AI Gateway also extends AI from insight to action through its integration of Natoma. Users can now send e-mails, summarize slot conversations, open [ Gira ] tickets and act across their business, all without leaving CoWork or CoCo. We've also continued to advance how our agents understand the unique context of a business. At Snowflake Summit, we introduced Cortex Sense, which captures the business definition and institutional knowledge and AI agent needs and provides that context at the moment it answers the question. This means Snowflake is giving AI both the context to understand the business and the ability to act on its behalf with enterprise security, governance and observability built-in.

As we drive this AI transformation for our customers, we are leading from the front using CoCo and CoWork throughout our own business to accelerate productivity and efficiency. For example, in our marketing organization, CoCo has helped bring search optimization in-house, eliminating $400,000 in annual agency spend, reducing keyword research from approximately 10 hours to 20 minutes and content production from an estimated 24 hours down to just 2. In finance, our long-range planning used to require a 3-person team and more than 50 spreadsheets. It now runs with 1 analyst and a series of models that reflect our pricing structure and consumption dynamics.

Within our sales teams, we have automated prospecting for over 125,000 contacts and leads, with 70% of initial outreach e-mails for inbound leads now being generated automatically before SDR involvement. We are bringing these proven use cases directly to market, while applying our operational learnings to continuously upgrade our platform, moving with speed to capture the AI opportunity in front of us. In the first half of this year alone, we launched over 330 product capabilities to general availability, 35% more than we did in the first half of last year, underscoring both the pace of our innovation and the breadth of platform expansion underway across Snowflake.

Our go-to-market organization also continues to execute as reflected in strong new customer growth. We have deployed CoCo and CoWork across the sales team to analyze pipelines, prepare for customer conversations and accelerate the onboarding of new reps. Our teams are using these products every day, learning firsthand what they can do and taking those insights directly to our customers. We're seeing the results in how quickly customers are putting Snowflake to work. The number of use cases, individual customer projects deployed on Snowflake increased 89% year-over-year as customers move more workloads into production. At the same time, use cases on per account executive increased 43% year-over-year, demonstrating both growing customer demand and strong sales productivity.

And we are pairing this investment in growth with continued operational discipline. We remain on track for GAAP profitability in Q4 fiscal '28 and the operating leverage we built along the way strengthens the durability of statute. Taken together, our rapid pace of innovation, fiber go-to-market execution and operational discipline positions us well to capture the huge opportunity ahead. This quarter demonstrated that the transition to the Agentic Enterprise is accelerating and Snowflake is at the center of it. AI agents are only as powerful as the data and business context there reason from under government's soundings. Snowflake provides that trusted foundation while bringing together model choice and flexibility, access to critical applications on the agentic-controlled plane that connects intelligence to action across the enterprise.

CoWork and CoCo demonstrate what governed architecture makes possible, enabling business users and builders to work with greater speed and intelligence while Snowflake manages the complexity underneath. And importantly, our customers' success with AI translates directly into growth for Snowflake. AI brings new workloads to the platform, extending our reach to new users and drive greater consumption across the business. We're into the second half of fiscal '27 with strong product momentum and we see a long runway for durable high-growth and continued margin extension. The Agentic Enterprise runs on Snowflake, and we're just getting started.

With that, I'll pass it to Brian to go through the financial details.

Brian Robins

Thank you, Sridhar. In Q2, product revenue once again accelerated to reach 37% year-over-year growth. This marks our third straight quarter of acceleration. Q2 benefited from continued strength in our core data platform business and a meaningful step-up in AI revenue.

Our AI revenue reflects a broadening portfolio of AI capabilities. CoCo delivered another standout quarter. Consumption of CoWork is scaling and driving revenue contribution alongside a diverse set of AI tools from AI functions and document assessing to machine learning and notebooks. Our go-to-market teams continue to execute well against a strong demand environment. As Sridhar mentioned, net new customer additions increased 32% year-over-year. We added 14 net new Global 2000 customers, bringing a total to 829. Our AI data cloud now supports over 41% of the Global 2000.

Within our existing base, customer expansion is healthy, as evidenced by our net revenue retention rate of 126%. This expansion is underpinned by growth in both migrations and AI use cases. In Q2, 48 net new customers surpassed $1 million in trailing 12-month spend. We now have 828 customers spending above the $1 million threshold. Remaining performance obligations grew 30% year-over-year, totaling $9 billion. As a reminder, we continue to see customers favor Q4 renewals. As a result, we expect bookings to be increasingly weighted towards the fourth quarter. Of the $9 billion RPO, we expect approximately 54% to be recognized revenue in the next 12 months. This represents an approximately 42% year-over-year growth compared to our estimate in the same quarter last year.

Our Q2 results reinforce our commitment to delivering both growth and margin expansion. In Q2, non-GAAP operating margin expanded over 400 basis points year-over-year to reach 15%. Our outperformance was driven by strong revenue growth and disciplined head count management. Year-to-date, we've added 334 employees, which includes 173 from our Observe acquisition. This compares to 935 added in the year ago period. We ended the quarter of $4.3 billion in cash, cash equivalents, short-term and long-term investments.

Moving to our outlook. As always, our forecast is based on observed consumption patterns. There are no changes to our forecast methodology or our guidance philosophy. Given the strength we've observed both in our core data platform business and AI business, we are raising our product revenue guidance for the year. For FY '27, we now expect product revenue of $6.07 billion, representing 36% year-over-year growth. This includes approximately 1 percentage point of growth from Observe, consistent with our previous outlook. In Q3, we expect product revenue between $1.588 billion and $1.593 billion, representing 37% to 38% year-over-year growth.

Turning to margins. For FY '27, we now expect 74% non-GAAP product gross margin. This revised outlook includes a higher revenue mix from fast-growing AI workloads, which carry a lower contribution margin today. We're delivering continued operating margin expansion as we offset growing cloud costs with slowing headcount expense. We are increasing our FY '27 non-GAAP operating margin guidance from 13.5% to 14.5%. For Q3, we expect non-GAAP operating margin of 15.5%. We're reiterating our full year non-GAAP adjusted free cash flow margin guide of 23%.

I'd like to close with my 2 key goals for the year: first, help the business to deliver growth and margin expansion; second, support ongoing excellence in our go-to-market motion. AI is fundamental to our progress against both goals. As we help our customers modernize their data and business operations, AI is becoming a powerful growth driver. Internally, AI is locking greater productivity. Across the organization, from sales to engineering to finance, our use of AI is transforming our daily work. AI is driving greater efficiency and reducing our reliance on head count growth. Our progress against both priorities is evident in the strength of our Q2 results.

With that, I'll pass the call to the operator for Q&A.

Operator

[Operator Instructions] We will take our first question from Sanjit Singh with Morgan Stanley.

分析师问答

Sanjit Singh

Congrats on the second quarter of a pretty material acceleration. The spirit of my question is around the quality of the acceleration that you're seeing and just sort of as a backdrop around when the time the company went public, growth was being driven by a lot of investment in cloud, cloud native companies that may have been unprofitable. And so I wanted to ask a question on the quality of the acceleration on bit of 2 levels.

First, on the right to win, in the script, you guys mentioned supply chain use cases and finance use cases. The question here is why is CoCo along with the platform, the right mousetrap for these use cases that kind of extend beyond classic kind of business analytics use cases? And then on sort of the durability of the growth, like, are you seeing any sort of irrational behavior or poor operational hygiene when it comes to consuming both CoCo and CoWork? So really sort of a question on the quality of the acceleration you're seeing.

Sridhar Ramaswamy

This is Sridhar. Let me take a first cut at this, other folks can add on since it's a pretty broad question. First, I think we see the acceleration come from a very broad swath of customers. It is not concentrated, for example, with, let's say, AI native companies. They continue to be a small and a small part of our overall revenue stream. And I think the thing that's also materially different this time around with folks that are investing is that products like CoCo make optimization far, far easier than before. You can point CoCo at a credit that's taking too long to run or you can basically have it debug the top 10 longest running queries are the most idle warehouses, things like that are a lot easier to do.

And in fact, our cost management, our cost management skill in CoCo is a top 10 skill. And it's also the case that as a company, we have learned the lessons of the pandemic and things that we stress with each and every 1 of our customers is the need to drive spend in an efficient way. And this is also a mantra that our sales team itself adopts pretty aggressively because they know that every such case where they go to a customer and point out things that they could be doing better is a trust-building exercise that is going to more than pay for itself in new projects that customers will implement on Snowflake.

So overall, I'm pretty happy with both the fact that our growth is coming from a very broad swath of our customers without a whole lot of concentration in any one particular sector. And also about the fact that the very tools that make it possible to do things quickly also come with a set of functions that make it pretty easy to optimize. And the final point, as I said, others will add on to it. The final point about our right to win for the kind of business use cases that perhaps we previously were not there in the conversation for, AI has massively shrunk the distance between data and value. I'm sure all of you have it in your day-to-day life. But certainly, I, as the CEO can get a whole lot of value out of data a lot faster because of tools like CoCo and CoWork.

And the agentic harness is need a very powerful weapon for solving many different kinds of problems. And it is our ability to take these powerful tools and drive our own transformation, whether it is in making SDRs more efficient or in making account planning work much more effectively at scale or in letting our sales leaders inspect and run their businesses a lot more effectively or our finance team under Brian to be a lot more effective with what they do. We are able to go to our customers and not just reach, but also demonstrate what we have shown for ourselves internally. That just gives us a lot of credibility going into these conversations about transformation.

Brian Robins

I'll add just a little onto what Sridhar said. From a durability perspective, we give our guidance based Observe behavior. So we've seen couple of quarters of this behavior. Our sales team is doing a great job with proving the business value of the use cases, and we're continuing to see great new logo additions. CoCo , when we look at CoCo, the accounts that are using CoCo are consuming more of the core as well. And so there's a flywheel effect that we talk about. We had 100 CoCo accounts this quarter. That's up significantly from last quarter. And the gross retention rate has been relatively flat across the last several quarters.

And then just want to emphasize what Sridhar has said as well is, we're actually selling into way more personas today. So in a given week, I have 3 to 5 conversations with CFOs of existing customers of ours or customers that want to be. And so the CFOs are now making the purchase decision, the CRO, CMO, CEOs. And so there's a lot more percentage that we're selling into this broader portfolio of products.

Operator

And we will go to next question from Stewart Materne with Evercore ISI.

S. Kirk Materne

Congrats on a great start to the year. I was wondering if you guys could try to separate out a little bit or give us a little bit of color on how we should think about what portion of the acceleration is coming from these newer products that are obviously getting really rapid adoption versus sort of the flywheel of those new products on the core? I assume just given the size of the core, it's the core growing faster is probably the bigger factor, but I was wondering if there's any way for us to sort of distill down what these new products are having maybe on their own account.

Sridhar Ramaswamy

I would roughly call it even. Our AI products, this is a pretty broad swath at this point. Absolutely, it's CoCo and CoWork. But it's also things like AI functions that make data operations proceed at an impressive scale or even newer products like the AI Gateway, they contributed approximately half of the acceleration that we are seeing.

But there are a lot of other products that are also demonstrating robust growth, and Brian touched on some of them. Whether it's Notebooks or applications written in Streamlit or React that are deployed into Snowflake. And of course, migrations themselves going faster. I have talked pretty much in every single earnings call over the past 6 quarters about migrations. And that is an area where we continue to get faster and faster. And some of the recent advances, both in models and harnesses are letting us run long-duration tasks of a scale and complexity that we haven't been able to do before.

And the rate at which workloads are coming on to Snowflake is also an important factor. And one anecdotal example, that a big network manufacturer is doing a Teradata migration in less than 3 quarters this year. And this is something that would have taken probably 2 to 3 years in any previous time. So these are some of the things that are contributing to our acceleration and [indiscernible]

Operator

And we will take our next question from Karl Keirstead with UBS.

Karl Keirstead

Okay. Great. Maybe I'll direct this to Sridhar and Christian. I'd love to ask about model neutrality and model choice. I'm guessing the bulk of tasks completed by CoCo are being directed to frontier labs. But I'm just curious, during the quarter, did you detect any interesting behavioral shift, let's say, a mix shift from open class models to sonet class models? And if that happens, Brian, is there any effect potentially positive on gross margins to Snowflake's financials? And Sridhar, is being model neutral, is that becoming a competitive advantage in cases where Snowflake competes directly with the prospect of a customer using one of the frontier labs stand-alone?

Sridhar Ramaswamy

I'll start. Christian will add on. As models have gotten more powerful, cost has absolutely become a concern. And all of you know this, at least as far as the frontier labs go, there used to be somewhat of a dichotomy where Anthropic was available extensively on AWS, while the OpenAI models tended to be more on Azure. The material change that's happened is that both the companies are deploying substantial capacity of their own, but it's also the case that they are available in other clouds than the ones that they started with.

And we're absolutely seeing a lot of interest in being able to switch between different models and also to optimize cost. And this is also where open source models come in. There's obviously been several generations of these open source models, and we support many of them within Snowflake. And yes, we have pretty different economics when it comes to open source models since we run the inference ourselves. So that offers a lot of potential for future optimization.

And within our harnesses, many of the requests that we get from customers come in this mode that we call auto, where we can pair up the task with the model that is most appropriate for that particular task. And that gives us a lot of leeway in being able to optimize task for our customers.

Christian Kleinerman

Yes, Karl, in addition to what Sridhar said, another interesting trend that I would call it early, but we're hearing from a number of customers is the desire to post-train open models, which the training itself is an opportunity for us, and we're starting to see a lot of interest.

And to your question on whether neutrality is a competitive advantage, absolutely, it is. We have heard from many, many customers that they made large commitments to one specific model company and later on are saying, oh, I should have wanted to do a different model, whereas the commitment to Snowflake gives them that flexibility. And as Sridhar said, automatic routing into what is the right model for the right task. So definitely a very strong advantage for us.

Sridhar Ramaswamy

This is a theme that clearly, Christian and early Snowflake pioneered in terms of being able to offer really great capability across the cloud service providers. To quote [ Yogibera ], it feels like the [indiscernible] all over again when it comes to model neutrality.

Brian Robins

Karl, just real quickly, I just wanted to hit on the margin aspect to your question. Going back to -- when we develop products, the #1 thing is we want to develop a great product. That is the key thing that we want to do. Secondly, we want to make sure that we have massive adoption through use cases and driving benefit to then, in turn, drive revenue. And then we'll work on sort of the margin implication of that. Sridhar and I are very committed to driving overall operating margin leverage in the business.

And so you saw our non-GAAP product gross margin go down to 74% because we've increased our guidance so much. And so the mix between our AI products and the course change a little, but we're still committed as we guided to increasing our overall operating margin. And so as we go through and do model choice and use different models, the best thing for us right now is to give our customers the best answer with the best business outcome, and then we'll continue to work on margins as we go forward. But we're committed to driving operating leverage in the model.

Christian Kleinerman

And one more thing on this one, Karl, which is even the frontier models have been revising prices down on a regular basis and have been introducing additional models to our families, which have kept cost somewhat intact relative to the usage of organizations.

Operator

We will take our next question from Raimo Lenschow with Barclays.

Raimo Lenschow

Congrats from me as well. If I look at the organization and if I look at where revenue is coming from at the moment, you're very -- you're still relatively index towards U.S. North America. And can you talk a little bit about what you're seeing in other regions like Europe, Asia? Because it does seems there's like a big opportunity to expand the footprint there?

Brian Robins

Yes, absolutely. I think there's -- this isn't region-specific. When set the sales QBR just a month ago and looked at sort of the performance and all regions are performing, and the outlook for our regions are factored into our guidance, but all regions are operating very well.

Operator

We will take our next question from Ryan MacWilliams with Wells Fargo.

Ryan MacWilliams

This really seems like the AI moment for the data space. What would you say is the biggest change on why AI is accelerating Snowflake revenues now? Is it Cortex Code helping users get activated on AI faster? Has it been some of your other product improvements in conjunction with better ad models now making AI as more attractive or customers just more ready for AI? What do you think has led to this AI moment for Snowflake?

Sridhar Ramaswamy

I spoke earlier about the flywheel. It's a lot of things coming together. What products like CoWork firmly demonstrated was the ability to get really flexible and quick value from data. The demo that I have unfailingly showed every CEO that I've met is the one in which I look up their company as a customer on Snowflake. It really brings a life the power of data in ways that abstract expressions never can. And there's this growing realization that AI is a massive unlock for getting the data to the right person.

And most data teams are embracing this moment because they see this as a way to get past the unending backlogs that they have had pretty much since time in memorial. That's a little bit of effect number one. And what CoCo has done for us in a super native way is it's made the entirety of Snowflake, Absolutely, our sales team AI native. They feel a lot more confident about being able to support any use case on Snowflake because the answer to most problems that a customer or you run into is to simply ask CoCo how you solve the problem. And in most cases, it can solve it by itself.

And so we see a lot of customers, a lot of partners take on migrations, get projects done that honestly, we would not even have conceived of when we originally wrote Cortex Code. That's the magic of these coding agents. And in a funny kind of way, CoCo also makes it far easier to create agents and get value from data itself. And this is the combination that makes Snowflake so attractive. And it's not just acquiring customers. We track this metric called like time to 80% of purchased consumption for new logos that we acquired. And it's -- and we measure it cohort by cohort.

Basically, of the customers that you acquired, let's say, in January, what fraction of them are consuming more than 80% of the purchased capacity, call it, 3 months after their purchase month. And this metric has very, very visibly improved for the newest cohorts of customers that we are acquiring. That's the part of AI. It's faster to get projects done. It's faster to get value from data. And that's the flywheel that we think is really driving the acceleration in our overall business. And as models continue to get smarter as our ability to run more long-duration things, agents in the cloud continue to mature. We expect this flywheel to accelerate even more.

Operator

We will take our next question from Matt Hedberg with RBC Capital Markets.

Matthew Hedberg

Congrats from me as well. I wanted to piggyback on the CoCo, CoWork line of questioning. It just seems increasingly that both products are really well positioned to identify the modern enterprise. And Sridhar, you mentioned you use it every day, your sales team is using it every day. I'm just kind of curious, how deep within your knowledge worker base is CoCo being used like things like procurement as an example. And is the right way to think about CoCo being more of a sandbox, if some of these use cases become more repeatable, that these can be brought over to CoWork as more turnkey use cases of agents?

Sridhar Ramaswamy

This is Christian's favorite question, so I'll let him answer it.

Christian Kleinerman

Absolutely. Like the pattern that we're seeing is we're leveraging CoCo and CoWork throughout pretty much every function and every key business process throughout Snowflake. And we're leveraging that not only to inform the quality and completeness of our products, but also go and engage with our customer, tell them, this is how you become AI native, this is how you go and drive efficiencies. And that continues to accelerate and inform one another.

Brian Robins

And you're talking about sort of how deep it's used by knowledge workers like just in my organization, we're using it in deal desk and tax and accounting, internal audit, SG&A, treasury. So we have over 150 Snowflake on Snow within the organization, where people are using CoCo to fundamentally change the way that they do work. And so the adoption within the finance organization is almost at 100%.

Christian Kleinerman

And it's true across functions.

Operator

We will take our next question from Koji Ikeda with Bank of America.

Koji Ikeda

So you described AI as a structural multiplier because customers using AI consume more across the broader Snowflake platform. And so what is the consumption uplift for AI adopters relative to comparable nonadopters? How has that developed across the earliest cohorts? And what evidence are you seeing? Or maybe what is giving you the confidence that all of this reflects higher lifetime consumption rather than projects just being pulled forward?

Sridhar Ramaswamy

Yes, I'll take a first cut, and Brian will add on. At this time, we aren't ready to share the exact uplift numbers, but we do measure cohort behavior. And as CoCo adoption gets deeper, more users within an account adopting and more accounts and more customers themselves adopting. The effect is pretty noticeable for all the different cohorts that we have worked with.

And what gives us confidence that this is not merely projects being pulled forward, is both the breadth and depth of use cases that are coming our way in terms of what people are doing with CoCo and CoWork. It is allowing people to do fairly sophisticated actions that previously would have required things like applications. Our own sales leadership teams, for example, have been experimenting a lot with their inspection process, how they can drive their business forward.

And something like that would have required a specialized piece of software, a multi-quarter implementation cycle and then a staged rollout, things like that are literally now a matter of a pretty smart sales leader saying things in English and having CoWork translate that into what looks like a product. This, combined with the fact that we are now having conversations with our customers about a set of use cases that honestly, we would not have been considered before. This is everything from supply chain optimization are much better support systems in the case of Sanofi, our much better fraud and risk detection systems. This is what gives us confidence that there is both breadth and depth in what AI is able to do for Snowflake.

Operator

We will take our next question from Brent Thill with Jefferies.

Brent Thill

Sridhar, on CoCo, good to see 2,000 accounts added. I guess when you start to see now quarter-over-quarter, is there a difference you're seeing in adoption? Are you getting bigger lands, more users, bigger consumption right out of the gate? Anything that you're seeing that's a trend line since the product is shipped?

Sridhar Ramaswamy

Yes. I work with the team that basically does go-to-market. This is the sales team, especially on the solution engineering side, our specialist team, but also the product team. And we have a pretty sophisticated methodology for measuring CoCo penetration from -- we need to get through legal terms all the way to there are a set of daily users of the product that are living inside CoCo. We have our own pipeline for what does this -- for the different stages of this penetration.

But more importantly, we also now have a suite of tools ranging from in-product guidance within Snowsight to hands-on labs that we run for 3 hours with our customers. And obviously, we have a lot of customers we can do hands-on labs with each and every one of them, but we are getting much better at matching our actions to the things that are going to drive outcomes. We are also doing a good job of sharing best practices across the different years in the globe.

All of this is driving just really positive momentum. And more importantly, this feels like a problem that is ours to solve and drive at scale for the simple reason that CoCo makes every single thing that a customer does with Snowflake go faster and better. So it's among the easiest sales that we have done to our customers. But I'm also pretty happy with how methodical and thorough we are being in driving CoCo adoption.

Operator

We will take our next question from Brad Zelnick with Deutsche Bank.

Daniel Knauff

This is Dan on for Brad. Congrats on a great quarter. I wanted to maybe go back to an earlier question on kind of model neutrality or optionality with open in frontier models now being offered. Maybe there's a third lag around models of your own like Arctic that might be specifically tuned for the Snowflake platform. I'd just be curious what the latest is in terms of your ambitions here and kind of fold into the overarching model strategy for CoCo and CoWork?

Christian Kleinerman

Yes. So Christian here, Brad. We have not changed the direction we've been on, which is we're not training models to go get into a frontier type of model. But we have continued developing models in the Arctic family for tasks that are more specifically, more constrained, that we can provide higher accuracy and more efficiency. We do that in some of the AI functions. We do that for some of the document processing. We do that for embedding, et cetera.

So we will continue doing that type of activity. And as you know, the mix and matching of Frontier, close models, open wave models and our own models with fine-tune models will continue to be part of how we help customers at the end of the day, deliver -- achieve what they want, which is what is the right model for the right task that gives the correct results at the best efficiency.

Operator

We will take our next question from Alex Zukin with Wolfe Research.

Aleksandr Zukin

Congrats on an exceptional quarter. I guess maybe Sridhar, it feels like we're still very early in the Agentic Enterprise experience. And yet you guys are already seeing a pretty meaningful inflection. And I appreciate that it's too maybe early to share the kind of ARPU expansion at some of these early adopters. But you talked about accessing larger kind of strategic priorities, maybe larger budgets. So maybe can you just talk about how much -- what is the embed of opportunity that you are now able to access and see in terms of budget dollars? And maybe we then -- we've heard some really exciting tails of your FTE program and some of the exceptional tracking that's getting out there in the marketplace, particularly on the outcome-based selling. So maybe just give us a sneak preview of that as well.

Sridhar Ramaswamy

Yes. As I was remarking earlier, AI has dramatically lowered the distance between business value that somebody sees -- I mean, that a company sees and the data estate that's next to it. And often, it's not as complicated as it sounds. Recently, I was talking to an asset manager that manages tens of billions of dollars of assets, and they have this problem where they get a very large number of data sets delivered to them every single day. They have a large pool of assets that they have and the set of decisions that they are in the process of making about new moves that they could be taking.

Obviously, this is distributed across hundreds, if not thousands of people. That act of distributing information effectively is basically manual is place. It's a sheets being passed on, someone has to download a spreadsheet and update a model that's probably sitting on their local PC. And we are talking to them about how do we construct effectively like a multiplex or demultiplexer for the most important information that is coming and that can meaningfully lower both their return and reduce their exposure because models just do a much better job of doing this kind of work.

And that's just one among many, many, many conversations that I end up having, which is pretty remarkable for a person effectively heating a data infrastructure company. We've also hired a set of exceptional folks that have industry expertise that can answer simple questions around what are the top 6 things that are going to make the biggest difference to a company's top line and bottom line? And is there a new perspective that we can offer to these. And this is what the frontier engineering team is doing. It is combining a knowledge of what is possible with the data platform with harnesses like CoCo and CoWork with the industry-specific knowledge needed to drive meaningful outcomes to our customers.

We have talked publicly about working with folks like Sanofi in our frontier engineering program. But this is an area where there is breadth and depth of adoption. We are, for example, helping a big financial institution effectively overhaul their digital and data strategy and bring it to the modern world in a way that is very, very sustainable for them. And the confidence that we have going into these kinds of engagements it's not just that we commit to delivering the outcome. Obviously, we get paid only when we deliver outcomes in situations like this, but it's also in the fact that Snowflake is an open, well-understood platform.

And compared to some pretty proprietary folks out there where you have to go back to them after you get the first outcome, we can confidently tell them that their data team is very, very capable of driving further engagement with the projects that they have done and building on top of it. It's the combination of these things, our ability to truly talk about business outcomes, commit to deliver but deliver it on a clean, open, well-understood architecture that makes the customer looks good and stay good that I'm most excited by.

Operator

We will take our next question from Tyler Radke with Citi.

Tyler Radke

Sridhar, I wanted to ask your take on some of the moves we've seen from traditional SaaS companies partnering with LLM and sort of becoming more of a database themselves if the LLM sort of take the UI layer. How do you see this playing out? Does it make sense for Snowflake to take on more of the system of record data? And how do you sort of anticipate that, that competitive overlap looks over time?

Sridhar Ramaswamy

I mean, the way I think about this is that as software gets easier and easier to create, it's the data and semantics that require more and more important. It isn't lost on any of us that our ability to talk about new value with our customers is driven both by the breadth of the data estates that many, many of our customers have on Snowflake combined with the power off the harness, obviously, using the best models. So I've been very, very consistent for now 2-plus years in my conviction, in our conviction that owning the user experience is critical.

And we see CoCo and CoWork as fundamental to our future because they demonstrate to us and to our customers what is possible. But on the other hand, we understand that we live in a world where we have to play nice. Snowflake is only a part of the overall software estate that our customers have. We offer interoperability at multiple levels. But we think our flagship products are very important to our future.

Christian Kleinerman

Yes, I'll add maybe the notion of some of these application providers becoming database players is not a new trend. And what we hear consistently from CIOs and CTOs is, if I use 3 applications, I'm not going to copy my data into 3 different platforms. It's easier to consolidate in a single central platform like Snowflake, which is why we have bidirectional zero-copy partnerships with many of them, and we see a lot of customers aligning their data estates with Snowflake.

Sridhar Ramaswamy

Yes. And our investments in which Christian has pioneered and spearheaded with the team for a very long time around being able to host applications in Snowflake, small and big also positions us exceptionally well for many applications, not just analytic ones, but also systems of record operational ones that can be built right on top of Snowflake.

And so internally, we have many projects, some of which Christian and I like don't even know of people that are building interesting applications on top of the analytic data and operational stores that they're setting up within Snowflake. You can definitely expect to hear a lot more about things like hybrid tables and progress because they are the foundation, we think, for a new generation of agent applications, some of which will have UI and some of which weren't on top of Snowflake.

Operator

We will take our next question from [ Dami Jaffjee ] with JPMorgan.

Unknown Analyst

Congrats from my end on the strong results here. Maybe if I can ask on the full year guide and trying to parse out the increase in the foot year guide between core increases on the core versus AI. I think that last quarter, you had mentioned most of the full year guide increase was on account of CoCo. This quarter, it sounds a lot more balanced between core and AI and your confidence in forecasting acceleration and product revenue growth also seems to be much higher. So just wondering if there's something fundamentally that changed during the quarter in terms of consumption of the core from your customers that's driving that higher visibility nnd raise to the full year? Or is it more just on account of visibility after having got through like half of the year at this point?

Brian Robins

Yes. This is Brian. Thanks for the question. We base our guidance based on Observe behavior up until the call that we have. And what we saw is that we talked about sort of CoCo, CoWork and all the AI functions driving additional business, but as well as the people who adopt them, they're also increasing business within the core. So it's a reflection of the strength that we're seeing in our AI products as well as the underlying strength that we're seeing in the core.

Operator

Thank you. This concludes today's question-and-answer session. I will now pass the call back to Snowflake for closing remarks.

Sridhar Ramaswamy

Thank you, everyone. The Agentic Enterprise runs on Snowflake. We have just achieved 37% year-over-year product revenue growth, marking our third straight quarter of acceleration, while expanding our non-GAAP operating margin 400 basis points year-over-year to 15%. AI has created a powerful flywheel effect across our business, strengthening platform demand, driving adoption of our native AI products, and in turn, fueling greater consumption across the business. And this flywheel is accelerating. Based on this strength, we have increased our fiscal '27 product revenue guidance by over 500 basis points to 36% year-over-year growth. We are executing with discipline and focus and see enormous opportunity ahead. Thank you.

Operator

Thank you. This does conclude today's call. Thank you for your participation. You may now disconnect.