Snowflake (SNOW.US) Q2 conference call: AI contributions increased by nearly half, CoCo accounts broke 9100, “AI flywheel” sped up

The Zhitong Finance App learned that the cloud data platform Snowflake (SNOW.US) announced results for the second fiscal quarter after the US stock market on Wednesday, and the stock price once jumped more than 20%. The company announced an increase in its annual product sales guidelines, which significantly exceeded Wall Street consensus expectations, and highlighted the rapid penetration capabilities of its AI-assisted coding tool Coco.

According to the data, as of the second fiscal quarter in July, product revenue increased 37% year over year to US$1.49 billion, higher than the market's general expectation of US$1.42 billion. Total revenue reached $1.55 billion, an increase of 35% over the previous year; however, remaining performance obligations (RPO, a measure of future contract amounts) recorded $9 billion, slightly lower than market expectations of $9.37 billion. The non-GAAP operating margin reached 15.3%, and the operating profit was 237 million. The AI product line became the biggest highlight of the season. The company revealed that its embedded AI programming assistant Coco added more than 2,000 new enterprise customer accounts during the quarter, bringing the total number of accounts used to 9,100.

Snowflake expects product revenue for the full year ending January 2026 to reach approximately US$6.07 billion, up from the previous value given in May, and far exceeding the analysts' average forecast of US$5.86 billion.

Here are the details of the Snowflake earnings call.

Participants

Company side:

Katherine McCracken — Head of Investor Relations

Sridhar Ramaswamy — CEO and Director

Brian Robins — Chief Financial Officer

Christian Kleinerman — Executive Vice President of Product Management

Analyst side:

Sanjit Singh — Morgan Stanley

Kirk Materne — Evercore ISI

Karl Keirstead — UBS

Raimo Lenschow — Barclays

Ryan MacWilliams — Wells Fargo

Matthew Hedberg — RBC Capital Markets

Koji Ikeda — Bank of America Securities

Brent Thill — Jefferies

Daniel Knauff — Deutsche Bank

Aleksandr Zukin — Wolfe Research

Tyler Radke — Citi

Samik Chatterjee — J.P. Morgan

Meeting minutes

Opening Remarks

Katherine McCracken (Head of Investor Relations):

Good afternoon everyone, and thank you for attending the Snowflake FY2027 second quarter results conference call. I am attending today's Q&A session with CEO Sridhar Ramaswamy, Chief Financial Officer Brian Robins, and Executive Vice President of Product Management Christian Kleinerman. Today's session will review financial results for the second quarter of fiscal year 2027 and discuss guidance for the third quarter and full year results.

This call contains forward-looking statements relating to business operations and financial performance. Subject to risk and uncertainty, actual results may vary materially. Relevant risk information can be found in our earnings press release, Form 10-K and 10-Q, and other reports submitted to the SEC. All statements are based on information we have as of today, and we are under no obligation to update them except as required by law.

We will also discuss some non-GAAP financial measures. See the investor presentation for details of the relevant definitions and reconciliation of GAAP and non-GAAP measures. Results press releases and presentations can be found on the official website investors.snowflake.com, and a replay of the conference call will also be available.

Now please Sridhar.

CEO Sridhar Ramaswamy speaks

Thank you Katherine, and thank you all for participating.

We are in the midst of a once-in-a-lifetime wave of technological change, and Snowflake has always been at the core of the enterprise AI revolution. AI is fundamentally changing the way businesses are built, operated, and made decisions. To remain competitive, every organization faces a new mission — to become an “Agentic Enterprise” (Agentic Enterprise) and achieve it quickly, safely, and efficiently.

Snowflake is making this transformation a reality. We bring together the core elements of an intelligent enterprise: a governed data foundation, access to leading AI models, deep application workflows, and a unified intelligent control plane, to collaborate these elements to transform intent into governed action. By putting intelligence to work at scale, our customers are speeding up the construction, efficient execution, and reshaping businesses in ways never before imagined.

Simply put, smart enterprises run on Snowflake. This trend is translating into strong business performance, and the Q2 results are the best proof.

Product revenue reached US$1.49 billion in the second quarter, and the year-on-year growth accelerated to 37%, achieving a record month-on-month increase in the dollar for the second consecutive quarter. After achieving 30% year-on-year growth in Q4 of the previous fiscal year, we increased our growth rate by 7 percentage points in just two quarters. Driven by a continued focus on execution and operational discipline, the Q2 non-GAAP operating margin increased by more than 400 basis points year over year to reach 15%. Thanks to all Snowflake employees for their hard work, you have made this achievement possible.

As these results fully demonstrate, AI is amplifying Snowflake's competitive advantage through three mutually reinforcing dynamics:

First, AI is bringing new workloads to the platform. To drive AI initiatives, enterprises need a governed, unified data and context foundation, and companies from all walks of life are choosing Snowflake to build this foundation.

Second, our self-developed AI products Cowork and Coco continue to be rapidly adopted. As customers build and deploy agents on Snowflake, we're expanding our role to the agent control plane, creating new opportunities for growth.

Third, AI activation continues to boost overall platform consumption. Customers using AI on Snowflake spend more on the data platform, creating a structural multiplier effect for our business.

Together, these three drivers show that smart companies have built a strong flywheel effect in our business, and this flywheel is running at an accelerated pace.

At the core of this momentum is the continued strength of our core business. Snowflake now provides data and AI infrastructure to 14,554 customers worldwide. Customers continue to choose Snowflake because our AI data cloud is easy to use, seamlessly connects to support collaboration, and can be trusted for enterprise-grade governance and security. There were 692 net new customers this quarter, including 14 Global 2000 companies, and the net number of new customers increased 32% year over year.

Meanwhile, world-renowned companies such as BlackRock and Block are deepening partnerships with Snowflake, migrating more mission-critical workloads to Snowflake, and using CoCo to accelerate delivery in multiple cases. The pattern is consistent: the more customers build on Snowflake, the deeper they go. At present, 65 customers have earned more than $10 million in product revenue in the past 12 months, fully reflecting the full investment of major customers in Snowflake.

One of our strengths is helping customers reach critical data outside of the organization. Currently, 43% of customers share data on Snowflake through at least one stable node, proving Snowflake's role as a modern enterprise “circular system” — we enable data, applications, and AI agents to flow securely and seamlessly not only within the organization, but across organizations. Reddit chose Snowflake because of our ability to share data, which now provides it with a privacy-safe advertising performance measurement service.

As customers rapidly modernize their data assets and establish a powerful contextual layer for AI, more and more customers are migrating workloads to our platform, and AI is vastly accelerating this process. For example, one of Australia's largest banks migrated its financial crime platform to Snowflake, processing 17 billion transactions and improving query performance by a factor of 10. Today, they're building AI agents on Snowflake to accelerate the migration of surplus data assets and automate legacy data discovery and mapping.

While strengthening core platform requirements, AI is also expanding Snowflake's opportunities to deliver next-generation AI-driven products and experiences. Because Snowflake is at the center of customer data, business context, AI models, and workflows, we are uniquely positioned to be the governed control plane for intelligent enterprises. Our groundbreaking AI products, Cowork and Coco, turn this vision into reality. They provide a layer of governance that allows all types of business users, from knowledge workers to developers, to use the full power of the corporate context for themselves through simple conversational language.

With CoWork and CoCo, customers are reshaping the most critical aspects of their business, from supply chain operations to enterprise-level sales processes. Sayari, who provides risk intelligence to Fortune 100 companies 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 to an AI-ready foundation. Adoption continues to grow as more customers see the potential of this technology.

CoWork has expanded to 5,800 accounts, an increase of nearly 11% over the previous month. CoCo continues to be rapidly adopted, and the number of accounts surpassed 9,100, with a net increase of more than 2,000 in this quarter alone. Security company 1Password (serving more than 200,000 businesses) chose Snowflake because of Coco's ability to quickly migrate critical data pipelines to Snowflake, laying the foundation for data and AI work. Indeed, the world's number one recruitment website, has fully deployed CoWork and COCO in its data team and integrated Snowflake into the core data architecture, indicating that cost and efficiency advantages continue to expand under the operating scale of more than 60 countries and 28 languages.

But the opportunities don't end with adoption. By using conversational language to enable enterprise data construction, collaboration, and interaction, CoWork and COCO are bringing a new user base to Snowflake. Among accounts using these products, we're seeing a jump in the number of users, and Snowflake is reaching new business units and expanding coverage within existing teams.

While continuing to develop CoWork and CoCo into intelligent control planes, we are also building a broader platform for enterprises to deploy AI on a large scale. Model selection gives customers flexibility, can choose from leading cutting-edge models and open source models, and adjust strategies as the market changes. Post-training allows customers to adjust models based on specific data and business contexts. Agent observability and analysis allow customers to understand AI behavior, performance, and costs at a glance. To help customers optimize cost, performance, and speed, we launched the Cortex AI Gateway, which can dynamically route each task to an appropriate model based on customer-defined policies and real-world performance data, with built-in cost and governance controls.

As economic benefits improve, customers can deploy AI more widely with greater confidence, creating new catalysts for adoption and consumption on Snowflake. Cortex AI Gateway also extends AI from insight to action through integration with Natoma — users can now send emails, summarize Slack conversations, create Jira work orders, and take action across the business without leaving Cowork or Coco.

We are also continuing to advance how AI agents can understand the unique context of an enterprise. At the Snowflake Summit, we launched Cortex Sense, which captures the business definitions and institutional knowledge required by AI agents and provides that context the moment questions are answered. This means Snowflake is providing AI with the ability to understand the context of the business and act on behalf of the business, with built-in enterprise-grade security, governance, and observability.

While driving AI transformation for our customers, we ourselves are also pioneering the use of Coco and CoWork to accelerate productivity and efficiency. For example, in the marketing department, CoCo helps internalize search engine optimization efforts, saving 400,000 dollars in external agency expenses each year, shortening keyword research from about 10 hours to 20 minutes, and content production from about 24 hours to just 2 hours. In the finance department, long-term planning used to require a team of 3 people and more than 50 spreadsheets; now it only requires 1 analyst and a series of models reflecting pricing structures and consumption dynamics.

In the sales team, we have automated expansion of over 125,000 contacts and leads, and 70% of the initial contact emails for inbound leads are now automatically generated by the system before the sales development representative steps in. We bring these proven use cases directly to market, while applying our operational experience to continuously upgrade our platform to quickly seize the AI opportunities that lie ahead. In the first half of this year alone, we officially released more than 330 product capabilities to the market, 35% more than in the same period last year, fully reflecting the speed of innovation and the breadth of Snowflake's platform expansion.

Our marketing organization is also continuing to execute, as evidenced by strong new customer growth. We deployed CoCo and CoWork in our sales team to analyze the pipeline, prepare customer conversations, and accelerate the onboarding of new sales representatives. The team uses these products every day, knows their capabilities firsthand, and brings these insights directly to customers. The result was a significant reduction in the time customers put Snowflake into use—the number of customer projects (use cases) deployed on Snowflake increased 89% year over year, and customers put more workloads into production. At the same time, the number of use cases won by each account manager increased by 43% year over year, reflecting both increased customer demand and increased sales productivity.

Combining growth investments with continued operating discipline, we are still on schedule to achieve GAAP profits in the fourth quarter of FY2028, and the operating leverage established in the process will reinforce the sustainability of this outcome. Taken together, the rapid pace of innovation, closer marketing execution, and operational discipline put us in a good position to seize the huge opportunities ahead.

This quarter proved that the transformation to a smart enterprise is accelerating, and Snowflake is at the center of it. The strength of AI agents depends on the data and business context they are based on and the governance surrounding them. Snowflake provides this trusted foundation while bringing together model choice and flexibility, access to critical applications, and a control plane that connects intelligence to the entire enterprise.

CoWork and Coco demonstrate what is possible with a governed architecture—enabling business users and developers to work faster and at a higher level of intelligence, while Snowflake manages complexity in the background. Importantly, customer success on AI directly translates into Snowflake's growth. AI brings new workloads to the platform, extends our reach to new users, and drives greater consumption across the business.

We are entering the second half of FY2027 with strong product momentum, and are seeing a long-term runway of sustained high growth and continued expansion in profit margins. Smart enterprises run on Snowflake, and this is just the beginning.

Next, let's ask Brian to introduce the financial details.

Chief Financial Officer Brian Robins Speaks

Thanks Sridhar. Product revenue accelerated again in the second quarter, growing 37% year over year. This is the third consecutive quarter of accelerated growth. Q2 benefits from the continued strength of the core data platform business and a significant increase in AI revenue. Our AI revenue reflects an ever-expanding portfolio of AI capabilities. Coco once again delivered outstanding quarterly performance. Cowork's consumption is growing on a large scale and contributing revenue, while also having a diverse portfolio of AI tools, from AI functions and document processing to machine learning and notebooks.

The marketing team continues to perform well in an environment of strong demand. As stated by Sridhar, the net number of new customers increased 32% year over year, with a net addition of 14 Global Top 2000 customers, bringing the total number to 829. Our AI data cloud now serves more than 41% of the world's top 2000 companies. Among existing customers, customer expansion remained healthy, and the net revenue retention rate was 126%. This expansion is supported by migration and the growth of AI use cases.

In Q2, 48 net new customers spent more than $1 million in the past 12 months. Currently, there are 828 customers who have spent more than the $1 million threshold. Remaining performance obligations (RPO) increased 30% year over year, totaling $9 billion. It's important to note that customers still tend to renew their contracts in the fourth quarter, so we expect orders to be increasingly concentrated in the fourth quarter. Of the $9 billion RPO, approximately 54% is expected to be recognized as revenue within the next 12 months, an increase of about 42% over the estimated value for the same period last year.

The Q2 results confirm our commitment to placing equal emphasis on growth and profit margin expansion. The Q2 non-GAAP operating margin increased by more than 400 basis points year over year to reach 15%. The performance exceeded expectations thanks to strong revenue growth and strict personnel management. The year-to-date net increase of 334 employees, including 173 from Observe's acquisition, compared to 935 in the same period last year. Cash, cash equivalents, and short- and long-term investments totaled $4.3 billion at the end of the quarter.

Regarding the performance outlook. As always, our predictions are based on observed consumption patterns, and there are no changes in forecasting methodology and guiding principles. In view of the strong momentum observed in both the core data platform business and the AI business, we have raised our annual product revenue guidance. Product revenue for the 2027 fiscal year is expected to be $6.07 billion, up 36% year over year. This includes Observe's growth contribution of approximately 1 percentage point, which is in line with the previous outlook. Product revenue for the third quarter is expected to be between $1,588 billion and $1,593 million, up 37% to 38% year over year.

About profit margins. Non-GAAP gross margin for FY2027 is expected to be 74%. This adjusted outlook includes a higher share of revenue from rapidly growing AI workloads, which currently have a lower margin of contribution. We are offsetting rising cloud costs by slowing the growth of personnel expenses and achieve continued expansion of operating margins. We raised our FY2027 non-GAAP operating margin guidance from 13.5% to 14.5%. The third quarter non-GAAP operating margin is expected to be 15.5%. The full year non-GAAP adjusted free cash flow margin guidance remained unchanged at 23%.

Finally, I would like to highlight two key goals for this year: first, to help the business achieve growth and profit margin expansion; second, to support continued excellence in marketing activities. AI is critical for us to achieve both of these goals — while helping customers modernize their data and business operations, AI is becoming a powerful driver of growth; internally, AI is unlocking greater productivity across the organization. From sales to engineering to finance, the use of AI is transforming our everyday work, driving greater efficiency and reducing reliance on workforce growth. The strong performance in Q2 fully reflects our progress in the two priority directions.

Next, we'll move on to the Q&A session.

Q&A session

Sanjit Singh (Morgan Stanley): Congratulations on another significantly accelerated quarter. My questions revolve around the “quality” of accelerated growth. Looking back at the time the company went public, growth was mainly driven by cloud investments, and customers included many cloud-native companies that might not be profitable. I want to ask from two dimensions: First, when it comes to winning customers, you mentioned supply chain and financial use cases. Why is CoCo and the platform the right solution for these use cases that go beyond traditional business analysis? Second, sustainability of growth — are you seeing any irrational consumption behavior or operational health issues?

Sridhar Ramaswamy (CEO): I'll answer first. First, acceleration comes from a very broad customer base, not focused on AI-native companies — they still account for a small portion of our revenue. Significantly different from the past, products like CoCo make optimization easier than ever — you can let CoCo debug queries that run too slowly, or let it work on the 10 most time-consuming queries or the most idle warehouses. Cost management is one of CoCo's top ten skills. We have also learned lessons from the pandemic and emphasized to every customer the need to drive consumption efficiently. This is also a concept actively adopted by the sales team itself — every time we help a customer do better, it is a process of building trust, which will ultimately pay off in the customer's new project. I'm very happy with that. Regarding the ability to win orders for business use cases — AI has drastically shortened the distance between data and value. As CEO, I can get value from data faster because of CoCo and CoWork. Smart tools are indeed a powerful weapon for solving all kinds of problems. We can use these tools to drive our own transformation, and then not just preach, but show our clients the results of our internal practices, which gives us great credibility in transformation conversations.

Brian Robins (CFO): From a sustainability perspective, our guidance is based on observed behavior and has been observed for several quarters. The sales team did a good job proving the value of the business, and new customer acquisition continues to be strong. Accounts using CoCo also spend more on core platforms; this is the flywheel effect. The number of CoCo accounts reached 9,100 this quarter, a significant increase from month to month, and the overall retention rate has remained relatively stable over the past few quarters. We now sell to a more diverse group of people — I have conversations with current or potential CFOs 3 to 5 times a week, and CFOs, CROs, CMOs, and CEOs are all involved in purchasing decisions.

Kirk Materne (Evercore ISI): Congratulations on a great start to the year. Can you help differentiate how much of the acceleration comes from new products that are rapidly being adopted, and how much comes from the flywheel effect of new products on the core?

Sridhar Ramaswamy (CEO): Roughly half each. AI products (not only CoCo and CoWork, but also AI functions, AI Gateway, etc.) contributed about half of the acceleration. But other products also performed well, including Notebooks and Streamlit/React apps deployed on Snowflake. The migration itself is also accelerating. I've talked about migration in every phone call for the past 6 quarters, and we continue to accelerate in this area. Recent advances in models and toolchains allow us to run long tasks at a scale and complexity that was previously unattainable. A major network equipment manufacturer is completing a Teradata migration in less than 3 quarters, which previously took 2 to 3 years.

Karl Keirstead (UBS): Please ask about model neutrality and model selection. I'm guessing most of the tasks CoCo has completed point to frontier labs but have changes in behavior been observed this quarter? If a model switch occurs, what impact will it have on Snowflake's gross margin? Is model neutrality a competitive barrier?

Sridhar Ramaswamy (CEO): As models become more powerful, cost is definitely a concern. The substantial change is that both leading model companies are deploying their own computing power, and it is also available on non-starter cloud platforms. We see that customers are very interested in switching between different models and optimizing costs. The open source model has also been developed over many generations. We support many within Snowflake, run our own reasoning with completely different economics, and provide optimization potential. In the toolchain, many customers request entry in an “automatic” mode, and we can match the task to the most appropriate model. This theme is in line with Snowflake's pioneering practices of providing superior capabilities in the early days of cross-cloud service providers.

Christian Kleinerman (EVP of Product Management): Another early but noteworthy trend is that customers want post-training on open source models; the training itself is an opportunity for us. On whether model neutrality is a competitive advantage — definitely. Many customers said they had made large promises to a model company and later discovered they should choose a different one, and the commitment to Snowflake gave them flexibility.

Brian Robins (CFO): When it comes to gross profit margin, the first principle we develop products is to build great products, followed by mass adoption and drive revenue before considering the impact on profit margins. Sridhar and I are very firm on overall operating leverage. The drop in gross margin of non-GAAP products to 74% was due to a sharp increase in the guidelines and changes in the composition ratio of AI products to the core, but the operating margin guidelines are improving. Currently, the most important thing is to give customers the best answers and the best business results, and continue to optimize profit margins. At the same time, cutting-edge models are also regularly reducing prices and introducing new versions, making costs relatively manageable compared to organizational usage.

Raimo Lenschow (Barclays): You're still relatively focused on North America. What do you see in other regions such as Europe and Asia? There seems to be a huge opportunity for expansion.

Brian Robins (CFO): Absolutely. This is not a region-specific issue. When I participated in the quarterly sales review a month ago, all regions were operating well, and the regional outlook was included in the guidelines.

Ryan MacWilliams (Wells Fargo): This looks like an AI moment for data. What has changed so that AI is now accelerating Snowflake revenue? Is Cortex Code helping users activate AI faster, or are other product improvements combined with better models to make AI use cases more appealing?

Sridhar Ramaswamy (CEO): The flywheel effect is a combination of factors. Products like CoWork clearly demonstrate the ability to quickly and flexibly extract value from data. The constant demonstration I've shown every CEO is querying their own company's data as a customer in Snowflake — this shows the power of data more vividly than abstract expressions. AI is a large-scale unlocker that sends data to the right people, and the data team is embracing this moment. Coco makes the entire Snowflake and sales team AI-native — when you run into a problem, just ask Coco how to solve it. We've seen a large number of customers and partners take over migrations and complete projects, which were never imagined when Cortex Code was written. This is the magic of coding agents. CoCo also makes it easier to create agents and derive value from data. We're tracking the “time to reach 80% of purchases” metric for new customers — a significant improvement in the latest customer group. This is the power of AI: projects are completed faster, and data generates value faster.

Matthew Hedberg (RBC Capital Markets): COCO and CoWork seem to be well-positioned for “intelligent” modern enterprises. What is the depth of use of CoCo among knowledge workers? Is Coco more like a sandbox and can be migrated to CoWork when the use cases mature?

Christian Kleinerman (EVP of Product Management): We use COCO and CoWork in almost every function and key business process within Snowflake, not only to improve products, but also to show customers how to become AI native and drive efficiency.

Brian Robins (CFO): In my organization, dealing desks, taxation, accounting, internal audit, financial planning, and fund management are all used. There are more than 150 “Snowflake on Snowflake” projects within the organization, and the adoption rate of financial organizations is close to 100%.

Koji Ikeda (Bank of America Securities): AI is a structural multiplier. How is the increase in consumption of AI users compared to non-adopters? How can you be sure that this reflects higher life cycle consumption rather than projects being brought forward?

Sridhar Ramaswamy (CEO): We can't share accurate improvement numbers yet, but we do measure group behavior. As the adoption of CoCo deepened, the effects were remarkable in all groups. Confidence comes from the breadth and depth of Coco and CoWork use cases — they enable users to perform complex operations that previously required specialized software. Our sales leadership team is using CoWork to conduct business reviews, which in the past required dedicated software, multi-quarter implementation, and phased promotion; now smart sales leaders can generate results for similar products by speaking English. Coupled with the range of use cases we're discussing with customers — from supply chain optimization to Sanofi's customer support systems to better fraud and risk detection — all give us confidence in the breadth and depth of AI in Snowflake.

Brent Thill (Jefferies): CoCo added 2,000 new accounts, how was the quarterly adoption different? Are you seeing larger initial deployments, more users, and higher initial spending?

Sridhar Ramaswamy (CEO): We have proven methods to measure CoCo penetration stages, from legal terms to everyday users. We have a range of tools to guide you from within the product within Snowsight to a 3-hour hands-on lab. While we can't be a lab for every customer, we're getting better at matching actions to drive results, and are better at sharing best practices across regions around the world. More importantly, Coco makes everything our customers do on Snowflake faster and better, and is one of the easiest products we sell.

Daniel Knauff (Deutsche Bank): Regarding model neutrality/optionality, in addition to open source and cutting-edge models, are there any models specially tuned for the Snowflake platform like Arctic? What's the strategy?

Christian Kleinerman (EVP of Product Management): The direction hasn't changed — we don't train cutting-edge models, but we continue to develop models for specific constraint tasks within the Arctic family to provide higher accuracy and efficiency for certain AI functions, document processing, and embedding. A mix of cutting-edge models, closed source models, open source weight models, proprietary models, and fine-tuning models is always our way to help our customers achieve “the right model, the right tasks, the right results, and the best efficiency”.

Aleksandr Zukin (Wolfe Research): We're still in the early days of smart companies, and you've seen a significant inflection point. How big of a budget can you reach now? Can you share some details about the FTE plan?

Sridhar Ramaswamy (CEO): AI has greatly brought business value closer to data assets. Recently, I spoke with an asset management company that manages tens of billions of dollars of assets. Every day, they receive large data sets, and decisions are scattered among hundreds or thousands of people — information distribution is mostly manual, and spreadsheets are sent back and forth. We're discussing how to build a “multiplexing/de-multiplexing” mechanism to process the most important information, which can significantly reduce risk and exposure. This is one of many conversations. We recruit talented people with industry expertise to answer questions such as “What 6 things have the greatest impact on a company's revenue and profits”. This is what the Frontier Engineering team is doing — combining data platform capabilities, tools such as Coco/CoWork with industry knowledge to drive meaningful results for customers. We've spoken publicly about our partnership with Sanofi, and we're also helping a major financial institution overhaul its digitalization and data strategy. Not only are we committed to delivering results (only receiving payments when results are delivered), but Snowflake is an open, well-known platform — compared to some proprietary vendors, we can continue to tell our customer data teams are fully capable of continuing to build on the projects they have already done.

Tyler Radke (Citi): What do you think of the trend of traditional SaaS companies cooperating with LLMs and becoming databases themselves? Should Snowflake take on more of a “system of record” data role?

Sridhar Ramaswamy (CEO): As software development becomes easier, the importance of data and semantics becomes more prominent. Our ability to discuss new value with our customers is driven both by the breadth of their data on Snowflake and by the tool chain (combined with the best models). I've always believed “having a user experience is important” for over two years, and COCo and CoWork are critical to our future. But we also know that we must live in an eco-friendly way; Snowflake is only one part of the customer's overall software assets. We provide interoperability at multiple levels.

Christian Kleinerman (EVP of Product Management): It's not a new trend for app providers to become “database players.” The CIO and CEO told us that if you use 3 apps, you won't be replicating data to 3 different platforms—it's easier to integrate on a single central platform like Snowflake. That's why we work with many of them with bi-directional zero-copy, and we've seen a large number of customers aligning data assets with Snowflake.

Sridhar Ramaswamy (CEO): Our investment in hosting applications on Snowflake, including analytical and operational applications, also puts us in a good position. There are already a lot of people inside building interesting apps that even Christian and I don't know about. Look forward to hearing more about Hybrid Tables and Postgres — they are the foundation for a new generation of smart applications.

Samik Chatterjee (J.P. Morgan Chase): How much do the core and AI each account for in the annual guidelines being raised? You mentioned that most of the increase last season came from CoCo. This season seems to be more balanced. Has there been a fundamental change in core consumption?

Brian Robins (CFO): The guidelines are based on observational behavior up to the conference call. Coco, Cowork, and AI functions drive incremental business, and those who adopt them also increase consumption at the core. This is a combination of strong AI products and a strong core layer.

End of meeting

Sridhar Ramaswamy (CEO): Thanks everyone. Smart enterprises run on Snowflake. We have just achieved 37% year-over-year growth, accelerating for three consecutive quarters while increasing our non-GAAP operating margin by 400 basis points to 15% year over year. AI has created a powerful flywheel effect in our business — strengthening platform demand, driving adoption of self-developed AI products, and thereby driving greater consumption. This flywheel is speeding up. Based on this strong momentum, we raised our FY2027 product revenue guidance by more than 500 basis points to a 36% increase. We are implementing with discipline and focus, and the opportunities ahead are huge. Thank you.