Snowflake Inc (SNOW) 2026 Q3 法說會逐字稿

Good afternoon. Thank you for attending today's Snowflake Q3 Fiscal Year 2026 Earnings Call. My name is Jen, and I will be your moderator for today. (Operator Instructions)

午安.感謝您參加今天 Snowflake 2026 財年第三季財報電話會議。我叫珍，我將擔任今天的主持人。（操作說明）

At this time, I'd like to pass the conference over to our host, Katherine McCracken. Please proceed.

此時，我想把會議交給我們的主持人凱瑟琳·麥克拉肯。請繼續。

Good afternoon, and thank you for joining us on Snowflake's Q3 Fiscal 2026 Earnings Call. Joining me on the call today are Sridhar Ramaswamy, our Chief Executive Officer; and Brian Robins, our Chief Financial Officer.

下午好，感謝各位參加 Snowflake 2026 財年第三季財報電話會議。今天與我一起參加電話會議的有我們的執行長 Sridhar Ramaswamy 和我們的財務長 Brian Robins。

During today's call, we will review our financial results for the third quarter fiscal 2026 and discuss our guidance for the fourth quarter and full year fiscal 2026. 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.

在今天的電話會議上，我們將回顧 2026 財年第三季的財務業績，並討論我們對 2026 財年第四季和全年的業績展望。在今天的電話會議中，我們將發表一些前瞻性聲明，包括與我們的業務營運和財務表現相關的聲明。這些聲明存在風險和不確定性，可能導致其與我們的實際結果有重大差異。有關這些風險和不確定性的資訊可在我們的獲利新聞稿、最新的 10-K 表格和 10-Q 表格以及我們向美國證券交易委員會提交的其他報告中找到。

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 a reconciliation of GAAP to non-GAAP measures and business metric definitions, including 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.

我們所有聲明均基於我們目前掌握的信息，並截至今日發布。除法律另有規定外，我們不承擔更新任何此類聲明的義務。在今天的電話會議中，我們還將討論一些非GAAP財務指標，請參閱我們的投資者簡報，以了解GAAP與非GAAP指標的調節以及業務指標定義（包括採用情況）。獲利新聞稿和投資者簡報可在我們的網站 investors.snowflake.com 上查看。今天電話會議的錄音回放也將發佈在該網站上。

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

接下來，我將把電話交給斯里達爾。

Thanks, Katherine, and hi, everyone. Thank you all for joining us today. As every company transforms to embrace the AI era, Snowflake remains at the center of today's AI revolution. We have delivered yet another strong quarter, thanks to the hard work and dedication across our team to help our customers realize value through all their end-to-end data life cycle, and effectively harness AI's potential every step of the way. Our continued focus on operational rigor and close knit products and go-to-market execution has helped us maintain strength across our core business and innovate rapidly to bring new capabilities to market.

謝謝凱瑟琳，大家好。感謝各位今天蒞臨。隨著各公司轉型擁抱人工智慧時代，Snowflake 仍是當今人工智慧革命的核心。我們又取得了強勁的季度業績，這要歸功於我們團隊的辛勤工作和奉獻精神，幫助我們的客戶在整個端到端數據生命週期中實現價值，並在每一步有效地利用人工智慧的潛力。我們持續專注於營運的嚴謹性、緊密的產品和市場推廣執行，這幫助我們在核心業務方面保持了實力，並快速創新，將新的能力推向市場。

We are executing with urgency and focus and maintaining deep partnerships with our customers that enable us to capture the opportunity in front of us and sustain durable momentum. Product revenue in Q3 was $1.16 billion, up 29% year-over-year. Remaining performance obligations totaled $7.88 billion, with year-over-year growth accelerating to 37%. Our net revenue retention remained stable at a very healthy 125% and we added a record 615 new customers this quarter. As we continue to deliver strong revenue growth and healthy results, we are increasing our growth expectations for the year and reiterating our margin target.

我們正以緊迫感和專注力執行任務，並與客戶保持深入的合作關係，這使我們能夠抓住眼前的機會並保持持久的發展勢頭。第三季產品營收為 11.6 億美元，年增 29%。剩餘履約義務總額為 78.8 億美元，年成長加速至 37%。我們的淨收入留存率保持穩定，達到非常健康的 125%，並且本季新增客戶數量創下紀錄，達到 615 家。隨著我們持續實現強勁的營收成長和良好的業績，我們提高了今年的成長預期，並重申了我們的利潤率目標。

As I've shared, Snowflake is on a mission to empower every enterprise to achieve its full potential through data and AI and we are making incredible progress against that mission every day. We continue to double down on what makes Snowflake unique, delivering an AI data cloud that's truly enterprise-ready with a radical focus on our customers. Snowflake is intuitive and easy to use, seamlessly connected for collaboration and built with the security and governance that enterprises trust as their foundation. That's why customers like Coca-Cola Consolidated, PayPal and thousands more are transforming their businesses with Snowflake. And it's why more organizations than ever are going all in on Snowflake as their foundational data and AI platform.

正如我之前所說，Snowflake 的使命是透過數據和人工智慧賦能每個企業，使其充分發揮潛力，我們每天都在朝著這個使命取得令人矚目的進展。我們將繼續加倍投入，發揮 Snowflake 的獨特優勢，提供真正以企業為導向的 AI 資料雲，並始終以客戶為中心。Snowflake 直覺易用，可無縫連接以實現協作，並以企業信賴的安全性和治理能力為基礎建構而成。這就是為什麼像可口可樂聯合公司、PayPal 等成千上萬的客戶正在使用 Snowflake 來改變他們的業務。正因如此，越來越多的組織選擇 Snowflake 作為其基礎資料和人工智慧平台。

Already, Snowflake is the cornerstone for our customers' AI strategy. In Q3, more than 7,300 accounts are using our AI capabilities every week. Just recently, Morgan Stanley named Snowflake its Strategic Partner of the Year, recognizing how our AI Data Cloud is accelerating their transformation and driving AI innovation across one of the world's leading financial institutions. And with the general availability of Snowflake Intelligence, we are seeing the fastest ramp in product adoption in our company history. Already 1,200 customers are harnessing next-generation agentic AI capabilities to drive real business impact at scale.

目前，Snowflake 已成為我們客戶人工智慧策略的基石。第三季度，每周有超過 7300 個帳戶使用我們的人工智慧功能。就在不久前，摩根士丹利將 Snowflake 評為年度戰略合作夥伴，以表彰我們的 AI 數據雲如何加速其轉型，並推動這家世界領先金融機構之一的 AI 創新。隨著 Snowflake Intelligence 的全面推出，我們公司的產品採用速度達到了歷史最高水準。目前已有 1200 家客戶正在利用下一代智慧體 AI 功能大規模地推動真正的業務影響。

Snowflake Intelligence is transforming how businesses interact with their data, turning natural language into real-time actionable intelligence. For example, TS Imagine, a global SaaS platform for financial services, use Snowflake Intelligence to build an AI agent that now handles staff equal to 8.5 full-time employees. The agent helps users manage and query data, make faster trading and risk management decisions and automate customer case resolution, increasing transparency across teams and with clients. And Fanatics, the global leader in sports merchandise and e-commerce, uses Snowflake intelligence to connect billions of fan data points across shopping, collectibles and gaming platforms for more than 100 million fans worldwide. This Unified Data Foundation helps Fanatics better understand its customers, boost sales and grow its advertising business, powering the launch of new Fanatics advertising audience network this year.

Snowflake Intelligence 正在改變企業與其資料互動的方式，將自然語言轉化為即時可操作的智慧資訊。例如，面向金融服務的全球 SaaS 平台 TS Imagine 使用 Snowflake Intelligence 建立了一個 AI 代理，現在可以處理相當於 8.5 名全職員工的工作量。此代理程式可協助使用者管理和查詢數據，更快地做出交易和風險管理決策，並自動解決客戶案例，從而提高團隊之間以及與客戶之間的透明度。全球領先的體育用品和電子商務公司 Fanatics 利用 Snowflake 的智慧技術，將數十億粉絲數據點連接起來，涵蓋購物、收藏品和遊戲平台，服務全球超過 1 億的粉絲。這個統一數據基礎幫助 Fanatics 更了解其客戶，提升銷售額，發展其廣告業務，並為今年推出新的 Fanatics 廣告受眾網路提供了動力。

This momentum has enabled us to achieve a major milestone. $100 million in AI revenue run rate achieved 1 quarter earlier than anticipated, thanks to our pace of innovation, cross-functional collaboration and early adoption among many of our marquee customers. Because we operate as a consumption-based business, this number reflects real-world enterprise usage. It's a direct signal of how customers are using our AI capabilities in production to create value today. What's more, our AI capabilities are strengthening our customer relationships and expanding the value we deliver across every stage of the data life cycle.

這一勢頭使我們達成了一個重要的里程碑。由於我們的創新步伐、跨部門協作以及眾多重要客戶的早期採用，我們的人工智慧營收年化率提前一個季度實現了1億美元。因為我們是一家以消費量經營的企業，所以這個數字反映了企業的實際使用情況。這直接表明了客戶目前是如何在生產中使用我們的人工智慧功能來創造價值的。此外，我們的人工智慧能力正在加強我們與客戶的關係，並擴大我們在資料生命週期的每個階段所提供的價值。

AI is a key driver of the strength that we see in our core business. In Q3, we landed a record number of new logos and continue to build strong momentum with AI influencing 50% of the bookings signed this quarter. We also deepened relationships with existing customers as 28% of all use cases deployed during the quarter incorporated AI. But being enterprise-ready is not just about innovation, it's also about reliability. When a major cloud service provider experienced an outage this quarter, our disaster recovery capability seamlessly transferred more than 300 mission-critical workloads to backup systems, ensuring business continuity for our customers when it mattered the most.

人工智慧是我們核心業務實力成長的關鍵驅動力。第三季度，我們獲得了創紀錄數量的新客戶，而人工智慧繼續發揮著強勁的作用，本季簽署的訂單中有 50% 都受到了人工智慧的影響。我們也加深了與現有客戶的關係，因為本季部署的所有用例中有 28% 都採用了人工智慧。但具備企業級應用能力不僅關乎創新，也關乎可靠性。本季度，當一家大型雲端服務供應商發生故障時，我們的災難復原能力無縫地將 300 多個關鍵業務工作負載轉移到備份系統，在最關鍵的時刻確保了客戶的業務連續性。

Our commitment to making business critical capabilities, just work continues to resonate with our customers. And so we've built on this strength by expanding not only our product capabilities, but our ecosystem. This quarter alone, we announced new partnerships with Workday, Splunk, Palantir, UIPath and more to deepen integration, enable secure and seamless data access across the systems our customers use every day and unlock new innovations like agent to agent collaboration. More recently, we announced a landmark partnership with SAP to unite mission critical business data with the Snowflake AI Data Cloud. We are already supporting customers like AstraZeneca to access and analyze real-time data.

我們致力於讓業務關鍵能力成為切實可行的工作方式，這項承諾持續引起客戶的共鳴。因此，我們鞏固了這項優勢，不僅擴展了我們的產品能力，也擴展了我們的生態系統。僅本季度，我們就宣布與 Workday、Splunk、Palantir、UIPath 等公司建立新的合作夥伴關係，以深化集成，實現客戶每天使用的系統中安全無縫的數據訪問，並解鎖代理間協作等創新功能。最近，我們宣布與 SAP 建立里程碑式的合作夥伴關係，將關鍵業務資料與 Snowflake AI 資料雲結合。我們已經開始為像阿斯特捷利康這樣的客戶提供支持，幫助他們存取和分析即時數據。

These partnerships amplify our ability to deliver value to joint customers and extend our go-to-market reach. Our progress is clearly resonating with our global community. During Snowflake's annual world tour, over 40,000 customers, partners and prospects, joined us across 23 events, a record-breaking turnout representing a more than 40% year-on-year increase in participation from last year. More recently, our annual Build Developer Summit saw a 43% increase in attendance year-over-year underscoring the growing excitement and engagement across our global audience. Behind this incredible momentum is our relentless focus and continued delivery against our product strategy.

這些合作關係增強了我們為共同客戶創造價值的能力，並擴大了我們的市場覆蓋範圍。我們的進展顯然引起了國際社會的共鳴。在 Snowflake 的年度全球巡迴活動中，超過 40,000 名客戶、合作夥伴和潛在客戶參加了 23 場活動，創下了參與人數紀錄，比去年增加了 40% 以上。最近，我們的年度 Build 開發者高峰會的出席人數比上一年增長了 43%，這凸顯了我們全球觀眾日益增長的熱情和參與度。這驚人勢頭的背後，是我們始終如一的專注和對產品策略的持續執行。

Throughout the quarter, Snowflake maintained a rapid pace of innovation, bringing our total GA product capabilities to 370 year-to-date, a 35% increase over last year with AI being front and center. As I shared, Snowflake Intelligence continues to set the tone for enterprise-grade agentic AI. Just recently, we announced that Snowflake is the official data cloud provider for USA Bobsled/Skeleton, powering their journey to the upcoming Olympic Games. The team is using Snowflake Intelligence to unify and analyze data across its performance ecosystem to optimize, push performance and equip coaches with data-driven insights to create a competitive edge on the ice for a medal worthy performance. At the core of our AI philosophy is customer choice and flexibility, empowering organizations to leverage the world's leading models securely on their own enterprise data.

整個季度，Snowflake 保持了快速的創新步伐，使我們今年迄今為止的 GA 產品總數達到 370 項，比去年增長了 35%，其中人工智慧處於核心地位。正如我之前所說，Snowflake Intelligence 繼續引領企業級智慧人工智慧的發展方向。就在不久前，我們宣布 Snowflake 成為美國雪橇/鋼架雪車隊的官方數據雲供應商，幫助他們徵戰即將到來的奧運。該團隊正在使用 Snowflake Intelligence 來統一和分析其性能生態系統中的數據，以優化和提升性能，並為教練提供數據驅動的見解，從而在冰上創造競爭優勢，取得足以贏得獎牌的成績。我們的人工智慧理念的核心是客戶的選擇和靈活性，使組織能夠安全地利用世界領先的模型，並將其應用於自己的企業資料。

As you may have seen a few weeks ago, we announced our partnership with Google Cloud to make the latest Gemini models available to our 12,600-plus customers within Cortex AI and Snowflake Intelligence, further enhancing access and customer choice. To drive even more tailored innovation, we introduced Cortex AI for financial services, a comprehensive suite of AI capabilities and partnerships that empower financial services companies to unify their financial data ecosystem, deploy AI models, applications and agents securely and meet the rigorous security and compliance standards for regulated industry. Even as we supercharge the data life cycle with AI, we remain committed to strengthening our core data foundation. To ensure that Snowflake will continue to deliver the trusted, performant and scalable data platform our customers rely on every day. Key capabilities like Snowflake OpenFlow are making it easier than ever to bring in structured, unstructured, batch or streaming data into Snowflake.

幾週前，您可能已經看到，我們宣布與 Google Cloud 建立合作夥伴關係，將最新的 Gemini 模型提供給我們在 Cortex AI 和 Snowflake Intelligence 中的 12,600 多名客戶，從而進一步增強訪問權限和客戶選擇。為了推動更具針對性的創新，我們推出了面向金融服務的 Cortex AI，這是一套全面的 AI 功能和合作夥伴關係，旨在幫助金融服務公司統一其金融數據生態系統，安全地部署 AI 模型、應用程序和代理，並滿足受監管行業的嚴格安全和合規標準。即使我們利用人工智慧大幅提升了資料生命週期，我們仍然致力於加強我們的核心資料基礎。為了確保 Snowflake 能夠繼續提供我們的客戶每天所依賴的值得信賴、效能卓越且可擴展的數據平台。Snowflake OpenFlow 等關鍵功能使得將結構化、非結構化、批量或串流資料引入 Snowflake 變得比以往任何時候都更加容易。

Take EVgo, which is using OpenFlow to simplify and speed up how it ingests data across its EV charging network. By consolidating multiple data pipelines into Snowflake, EVgo has reduced latency, improved reliability and gained a more complete view of its customers and charging stations. And we are continuing to extend our value through strategic acquisitions. We recently acquired the technology behind Datometry software migration solution, which will enable our customers to move from legacy data warehouses to Snowflake at lower cost and with minimal disruption, further simplifying their journey to our AI data cloud. We have also agreed to acquire Select Star to enhance our Horizon catalog and deliver a more complete view of an enterprise's data estate.

以 EVgo 為例，該公司使用 OpenFlow 來簡化並加速其電動車充電網路中的資料收集方式。透過將多個資料管道整合到 Snowflake 中，EVgo 降低了延遲，提高了可靠性，並更全面地了解了其客戶和充電站的情況。我們正透過策略收購不斷提升自身價值。我們最近收購了 Datometry 軟體遷移解決方案背後的技術，這將使我們的客戶能夠以更低的成本和最小的干擾從傳統資料倉儲遷移到 Snowflake，進一步簡化他們向我們的 AI 資料雲端遷移的過程。我們也同意收購 Select Star，以增強我們的 Horizo​​n 目錄，並提供企業資料資產的更完整視圖。

We believe this richer context will empower agentic AI experiences like Snowflake Intelligence to better understand enterprise data and uncover deeper insight. As we scale the breadth and depth of our product capabilities, we continue to maintain tight integration across sales, marketing, product and engineering to effectively launch and scale new offerings and deepen our customer relationships. This alignment is driving tangible results. Q3 marked a strong bookings quarter underscored by accelerating RPO growth and healthy customer retention. At the same time, we are investing in and strengthening our strategic go-to-market partnerships.

我們相信，這種更豐富的脈絡將使 Snowflake Intelligence 等智慧 AI 體驗能夠更好地理解企業資料並發現更深層的洞察。隨著我們不斷擴大產品功能的廣度和深度，我們將繼續保持銷售、行銷、產品和工程部門之間的緊密合作，以有效地推出和擴大新產品，並加深與客戶的關係。這種協調一致正在帶來切實可見的成果。第三季預訂量強勁成長，RPO 成長加速，客戶留存率維持良好。同時，我們正在投資並加強我們的策略市場合作夥伴關係。

In addition to those I've already mentioned, today, we've announced an expanded partnership with Anthropic. This brings native model availability into Snowflake and also introduces a new joint go-to-market motion designed to accelerate enterprise AI adoption. We also continue to build our strong relationships with major cloud providers. In fact, Snowflake has already surpassed $2 billion in sales through AWS Marketplace in a single calendar year and was just recognized with 14 AWS Partner award wins, more than any other ISV provider. This underscores the extraordinary demand for Snowflake's AI data cloud.

除了我剛才提到的那些之外，今天我們也宣布與 Anthropic 擴大合作關係。這使得 Snowflake 能夠提供原生模型，同時也引入了一種新的聯合市場推廣策略，旨在加速企業採用 AI。我們同時也不斷加強與主要雲端服務供應商的合作關係。事實上，Snowflake 在一年內透過 AWS Marketplace 的銷售額已經超過 20 億美元，並且剛剛獲得了 14 項 AWS 合作夥伴獎，比其他 ISV 供應商都多。這凸顯了市場對 Snowflake 人工智慧資料雲的龐大需求。

Momentum is also accelerating with our global systems integrators. Accenture just launched a Snowflake Business Group, committing to train over 5,000 professionals on Snowflake solutions to help joint customers realize AI value faster. Already, Accenture and Snowflake are helping customers like Caterpillar, unlock the full value of their operational data. This collaboration is improving quality in manufacturing, providing timely insights for finance and helping teams share knowledge and solve complex challenges faster. As you can see, this was a milestone quarter for Snowflake defined by exceptional advances in product innovation and incredible customer momentum.

與全球系統整合商的合作動能也在加速增強。埃森哲剛成立了 Snowflake 業務集團，承諾培訓 5,000 多名 Snowflake 解決方案專業人員，以幫助共同客戶更快實現 AI 價值。埃森哲和 Snowflake 已經幫助像卡特彼勒這樣的客戶釋放其營運數據的全部價值。這種合作提高了製造質量，為財務部門提供了及時的見解，並幫助團隊共享知識，更快地解決複雜的挑戰。如您所見，對於 Snowflake 而言，這是一個具有里程碑意義的季度，其特點是產品創新方面取得了卓越的進步，以及客戶成長勢頭強勁。

As we deepen our strategic partnerships with the world's leading cloud service providers, AI model developers, SaaS providers and global system integrators, we're unlocking new levels of performance, accessibility and AI-driven insight for our customers while expanding the value and impact of the Snowflake platform across industry. I'm incredibly proud of our team for their efficiency and discipline they continue to demonstrate across the business. Our operational rhythm remains strong and as we invest strategically for long-term growth, we are building the foundation for sustained scale and high durable growth. To help lead us through this next phase, I am pleased to introduce Brian Robins as our new Chief Financial Officer. Brian brings extensive experience as a CFO across high-growth software companies and a deep understanding of scaling financial operations with discipline.

隨著我們與世界領先的雲端服務供應商、人工智慧模型開發人員、SaaS 供應商和全球系統整合商深化策略合作夥伴關係，我們正在為客戶解鎖更高水準的效能、可近性和人工智慧驅動的洞察力，同時擴大 Snowflake 平台在整個產業的價值和影響力。我為我們團隊在整個業務中持續展現出的高效能和自律感到無比自豪。我們的營運節奏仍然強勁，我們正在進行長期成長的策略性投資，為持續的規模化和持久的高成長奠定基礎。為了帶領我們進入下一個階段，我很高興地向大家介紹布萊恩‧羅賓斯擔任我們的新任財務長。Brian 在高成長軟體公司擔任財務長方面擁有豐富的經驗，並且對如何以嚴謹的方式擴展財務營運有著深刻的理解。

Brian, why don't you take us through some of the financial details.

布萊恩，能否為我們介紹一下財務細節？

Thank you, Sridhar. It's a truly exciting time for me to be at Snowflake. In Q3, we delivered strong results across revenue, bookings and margins. Our product revenue grew 29% year-over-year, fueled by durable growth in our core business and continued expansion into data engineering and AI workloads. Together, these factors contributed to a stable net retention rate of 125%.

謝謝你，斯里達爾。對我來說，現在加入 Snowflake 真是一個令人興奮的時刻。第三季度，我們在營收、訂單量和利潤率方面都取得了強勁的業績。在核心業務持續成長以及向資料工程和人工智慧工作負載的不斷擴張的推動下，我們的產品收入年增了 29%。這些因素共同促成了125%的穩定淨留存率。

Financial services and technology verticals led growth in Q3. We continue to see significant opportunity to expand within our existing customer base. Our Global 2000 customers now totaled 776 with each of these accounts spending, on average, $2.3 million on a trailing 12-month basis. Many of these customers are still in the early stages of their Snowflake journey with ample room for further growth. Q3 was an excellent quarter for go-to-market execution.

第三季度，金融服務和科技業引領了成長。我們仍然看到在現有客戶群中拓展業務的巨大機會。我們的全球 2000 強客戶總數達到 776 家，這些客戶在過去 12 個月中平均消費 230 萬美元。這些客戶中的許多人仍處於 Snowflake 使用之旅的早期階段，還有很大的成長空間。第三季是市場推廣執行表現非常出色的一個季度。

We achieved strong booking results, signing 4 9-figure deals. This represents a record number of large deals signed in a single quarter. Our focus on new customer acquisition continues to show yield. As Sridhar mentioned, it was a record quarter for new customer wins, adding over 600 new customers. Our ability to expand with existing customers and bring new ones onto the platform, underscores the strength of our business model.

我們取得了強勁的預訂業績，簽署了 4 筆九位數訂單。這標誌著單季簽署的大宗交易數量創下歷史新高。我們持續專注於獲取新客戶，並已取得成效。正如 Sridhar 所提到的，本季新增客戶數量創下歷史新高，新增客戶超過 600 人。我們能夠透過現有客戶拓展業務，並將新客戶引入平台，凸顯了我們商業模式的優勢。

Equally important, we continue to operate with financial discipline, delivering healthy margins as we scale. Q3 non-GAAP product gross margin was 75.9%. Non-GAAP operating margin expanded more than 450 basis points year-over-year to 11%, reflecting our continued focus on driving greater efficiency across the entire company. As a reminder, we intentionally front-loaded our sales and marketing hiring in the year. Non-GAAP adjusted free cash flow margin was 11%.

同樣重要的是，我們繼續秉持財務紀律運營，在擴大規模的同時保持健康的利潤率。第三季非GAAP產品毛利率為75.9%。非GAAP營業利潤率年增超過450個基點至11%，反映出我們持續致力於提高整個公司的效率。再次提醒，我們刻意將銷售和行銷方面的招募工作提前到了年初。非GAAP調整後的自由現金流利潤率為11%。

In Q3, we used $233 million to repurchase 1 million shares at a weighted average price per share of $223.35. We still have $1.3 billion remaining on our original authorization for $4.5 billion through March of 2027. We ended the quarter with $4.4 billion in cash, cash equivalents, short-term and long-term investments. Moving now to our outlook. For Q4, we expect product revenue between $1.195 billion and $1.2 billion, representing a 27% year-over-year growth.

第三季度，我們動用了 2.33 億美元，以每股 223.35 美元的加權平均價格回購了 100 萬股股票。截至 2027 年 3 月，我們最初授權的 45 億美元回購額度中，仍有 13 億美元剩餘。本季末，我們持有現金、現金等價物、短期和長期投資共 44 億美元。接下來談談我們的展望。我們預計第四季產品營收在 11.95 億美元至 12 億美元之間，年增 27%。

We expect non-GAAP operating margin of 7%. We are raising our FY26 product revenue guidance. We now expect product revenue of approximately $4.446 billion, representing 28% year-over-year growth. We are reiterating our FY26 margin targets. We expect non-GAAP product gross margin of 75%; non-GAAP operating margin of 9%; and non-GAAP adjusted free cash flow margin of 25%.

我們預期非GAAP營業利益率為7%。我們將上調2026財年產品營收預期。我們現在預計產品收入約為 44.46 億美元，年增 28%。我們重申2026財年的利潤率目標。我們預期非GAAP產品毛利率為75%；非GAAP營業利益率為9%；非GAAP調整後自由現金流利潤率為25%。

Before moving to Q&A, I'd like to share my perspective on my first 60 days here at Snowflake. Three key takeaways have truly stood out: First and foremost, I've been incredibly impressed by the caliber and energy of the Snowflake team. There's a sense of winning energy in every meeting and profound pride in their daily work. Specifically, the depth of the bench within our finance organization is exceptionally strong and really supports our next phase of growth. Second, I prioritize spending my initial weeks meeting with customers.

在進入問答環節之前，我想分享一下我在 Snowflake 工作的頭 60 天的感受。有三點讓我印象深刻：首先，Snowflake 團隊的素質和活力給我留下了極其深刻的印象。每次會議都洋溢著必勝的信念，他們對日常工作也深感自豪。具體來說，我們財務部門的人才儲備非常雄厚，能夠真正支持我們下一階段的成長。其次，我優先安排最初幾週的時間與客戶會面。

The customers I spoke with were fanatical about Snowflake and the transformational impact our platform has had on their business. They are placed in the AI data cloud at the absolute center of their strategic initiatives, underscoring our essential role in their future. Finally, the velocity of our product releases and innovation engine is world-class and consistently sets us apart. Snowflake sits at the intersection of a massive market opportunity and I could not be more excited to be part of scaling this phenomenal team and seizing the amazing growth ahead. As we look forward, my focus is on continuing to deliver efficient growth.

我接觸過的客戶都對 Snowflake 非常著迷，他們盛讚我們的平台對他們的業務產生了變革性的影響。它們被置於人工智慧資料雲中，處於其策略舉措的絕對核心位置，這凸顯了我們在它們未來發展中的關鍵作用。最後，我們的產品發布速度和創新引擎是世界一流的，這始終使我們脫穎而出。Snowflake 正處於一個巨大的市場機會的交匯點，我非常興奮能夠參與這個傑出團隊的擴張中，並抓住未來驚人的成長機會。展望未來，我的重點是持續實現高效成長。

I believe that continued alignment across our finance, go-to-market and product teams will enable us to balance growth with disciplined execution. With that, I'll now pass the call to operator for Q&A.

我相信，財務、市場推廣和產品團隊的持續協調一致，將使我們能夠在成長和嚴謹執行之間取得平衡。接下來，我將把電話轉接給接線生進行問答環節。

(Operator Instructions)

（操作說明）

Sanjit Singh, Morgan Stanley.

桑吉特辛格，摩根士丹利。

Yeah, thank you for taking the questions. I had one for Brian and one for Sridhar. Brian, first for you, when we look at the growth rates on product revenue this quarter, really attractive at 29%. It was just about 3% beat slightly below 3% beat versus the midpoint of guidance. But at the same time, when I look at your Q4 guide, it's probably the best sequential guide I've seen from the company in a couple of years. So I was wondering if you could help us square that.

是的，謝謝你回答這些問題。我為布萊恩準備了一份，為斯里達爾準備了一份。Brian，首先，我們來看本季產品收入的成長率，確實非常吸引人，達到了 29%。比預期中位數高出約3%，略低於高出3%。但同時，當我看到你們的第四季業績指南時，它可能是近年來我見過的貴公司最好的業績指南。所以我想知道您是否能幫我們解決這個問題。

And then for Sridhar, like really impressive in terms of getting to that $100 million AI revenue run rate. You mentioned on the press release that Snowflake Intelligence is one of the fastest adopting products. I was wondering if you could give us a color on the types of customers that are taking on your AI products, some of the use cases that Snowflake Intelligence is unlocking. And also if you could comment on kind of Cortex AI adoption. Thank you for the time.

對 Sridhar 來說，實現 1 億美元的 AI 營收年化率真的令人印象深刻。您在新聞稿中提到，Snowflake Intelligence 是普及速度最快的產品之一。我想請您介紹一下正在使用貴公司人工智慧產品的客戶類型，以及 Snowflake Intelligence 正在解鎖的一些應用案例。另外，您能否談談Cortex AI的應用情況？感謝您抽出時間。

Thanks, Sanjit. I'll answer the first part of the question on the financials. We're happy with the performance this quarter. We delivered 29% year-over-year revenue growth. The quarter pretty much played out as expected.

謝謝你，桑吉特。我將回答關於財務問題的第一部分。我們對本季的業績感到滿意。我們實現了29%的年收入成長。本季基本符合預期。

There was really only one surprise in the quarter, and that was the hyperscaler outage, which impacted our revenue approximately $1 million to $2 million within the quarter. I think it's really important with the consumption model that not to view quarterly beats as the best signal of the fundamentals within the business. The quarter, as you mentioned, we raised our fiscal year guidance by $51 million or $4.446 billion, and the FY guide is really the most meaningful signal. And I think the guide really reflects the underlying behavior that we see in our customer base going into the fourth quarter. Sridhar, over to you.

本季真正令人意外的只有一件事，那就是超大規模資料中心宕機，這導致我們本季的收入損失了大約 100 萬至 200 萬美元。我認為在消費模式下，非常重要的一點是，不要把季度業績超預期視為衡量企業基本面的最佳指標。正如您所提到的，本季我們將財年預期提高了 5,100 萬美元，即 44.46 億美元，而財年預期才是真正有意義的訊號。我認為該指南真實反映了我們在第四季度客戶群的潛在行為。斯里達爾，該你了。

Yes. Snowflake Intelligence amplifies the investments that our customers have made in putting high-quality data into Snowflake. To take our own example, we created a data agent on all of the sales information that matters for my sales team, whether it is consumption information or the Workday hierarchy itself of who is managing whom, information about customers, their use cases, and it's been a magical unlock for several thousand people because things that they needed to painfully find dashboards for, they can have answered immediately. Plus, you also get the benefit that unlike a dashboard, which is a 2D representation of a pretty complex space, you can ask questions that cut across any dimension, analyzed data in ways that previously were simply not possible before. And so we have a slew of customers, whether it is the USA Bobsled team or Fanatics or folks like ServiceNow or TS Imagine that are using this to create data agents specialized for some areas.

是的。Snowflake Intelligence 可以放大我們的客戶在 Snowflake 中投入的高品質資料的價值。以我們自己的例子來說，我們創建了一個資料代理，其中包含對我的銷售團隊來說重要的所有銷售信息，無論是消費信息還是 Workday 層級結構（誰管理誰）、客戶信息、他們的使用案例等等。這對幾千人來說簡直是神奇的解鎖，因為他們以前需要費力地在儀表板中查找的問題，現在可以立即得到答案。此外，與儀表板（它只是對相當複雜空間的二維表示）不同，您還可以提出跨越任何維度的問題，以以前根本不可能的方式分析資料。因此，我們有許多客戶，無論是美國雪橇隊、Fanatics，還是像 ServiceNow 或 TS Imagine 這樣的公司，他們都在利用這項技術創建專門針對某些領域的資料代理。

So anyone that is working in a particular function, for example, has all of the data that is relevant to them available from a single interface and right on their phone or laptop computer. It is that unlock of access to this data that is driving adoption. What I can tell you is like I -- whenever I have dinners with CIOs or with CEOs, and we are talking about them often, they turn out to be Snowflake customers and they end up showing off Snowflake Intelligence on my phone, usually to show them information that I -- we have about their companies, like how much they're spending, what use cases they have deployed. And the first thing that comes from them is they want this for their own business. That's the attraction of Snowflake Intelligence, which is it puts all of the data that matters to you right at your fingertips.

例如，任何從事特定職能的人員，都可以透過一個介面，直接在他們的手機或筆記型電腦上取得所有與他們相關的數據。正是這種對這些數據存取權限的開放推動了數據的普及。我可以告訴你的是，每當我與首席資訊長或執行長共進晚餐，並且我們經常談論他們時，他們最終都會發現是 Snowflake 的客戶，他們最終會在我的手機上展示 Snowflake Intelligence，通常是為了向他們展示我們掌握的關於他們公司的信息，例如他們的支出、他們部署了哪些用例。他們首先表達的就是，他們想把這個用於自己的生意。這就是 Snowflake Intelligence 的吸引力所在，它能將所有對你重要的數據直接呈現在你眼前。

And unlike before, this data is not confined to analysts. This is to every single business user within a company, and that's the big unlock for all.

與以往不同的是，這些數據不再僅限於分析師掌握。這適用於公司內的每位業務用戶，這是對所有人來說的重大突破。

Appreciate that thought.

感謝你的關心。

Kirk Materne, Evercore ISI.

Kirk Materne，Evercore ISI。

Yeah, thanks very much and thanks for taking the question, guys. Sridhar, I was wondering if you could just talk about the go-to-market. You guys mentioned you had a really nice quarter. And I was particularly interested in the 600 new customer wins. And I realize you all land small and then grow with your customers. But, with AI coming on in Snowflake Intelligence, are you landing with more products now, meaning is it still landing with the core data warehouse and then expanding?

是的，非常感謝，也感謝各位回答這個問題。Sridhar，我想請你談談市場推廣的問題。你們曾提到你們上個季度業績非常好。我尤其對新增的 600 位客戶感興趣。我知道你們都是從小規模起步，然後和客戶一起成長。但是，隨著 Snowflake Intelligence 中 AI 的出現，你們現在是否擁有更多產品？也就是說，它是否仍然以核心資料倉儲為基礎，然後逐步擴展？

Or are you all able to land with multiple products at once and then grow from there? I'm just kind of curious about whether your surface area is growing within some of these new customers. Thanks very much.

或者你們是否能夠同時推出多款產品，然後以此為基礎發展壯大？我只是有點好奇，在這些新客戶中，你們的市佔率是否有所成長。非常感謝。

Kirk, thanks for the question. Well, I think things like Snowflake Intelligence now play a key role in making the power of data come alive every single time you're pitching a new logo. One of the magic of recent advances in AI is our ability to do demos or POCs, proof of concepts, that are hyper customized for each customer. Often, we will generate a synthetic data set that say we will mimic an oil producer or a pharmaceutical company and show them the art of the possible. Previously when people got onto Snowflake, it was for an abstract need.

柯克，謝謝你的提問。我認為像 Snowflake Intelligence 這樣的工具現在每次你推銷新 logo 時，都能發揮數據力量的關鍵作用，讓數據真正發揮作用。人工智慧近期發展帶來的神奇之處在於，我們能夠為每位客戶進行高度客製化的演示或概念驗證。我們經常會產生一個合成資料集，聲稱我們將模擬一家石油生產商或製藥公司，並向他們展示各種可能性。以前人們使用 Snowflake 是為了滿足某種抽象的需求。

It was to make data more efficiently accessible so that you could do more analytics. Now we do the work to show them what is possible with a product like Snowflake Intelligence on top of their data. It just makes the value of the transition from previous systems on to Snowflake even more clear, and those are some of the stats that we've been sharing with you, which is AI having a helping hand. It's not the dominant thing but definitely having a helping hand in more than -- in close to 50% of the new logos that we acquired. I would say definitely opens up our aperture.

這是為了讓數據更容易獲取，以便進行更多分析。現在，我們要做的工作就是向他們展示，在他們的資料基礎上，借助 Snowflake Intelligence 這樣的產品，可以實現哪些功能。這更清楚地表明了從以前的系統過渡到 Snowflake 的價值，而這些正是我們一直與大家分享的一些統計數據，即人工智慧在提供幫助。雖然這不是主要因素，但在我們收購的新標誌中，它確實發揮了重要作用——在近 50% 的標誌設計中都起到了輔助作用。我認為這無疑拓寬了我們的視野。

On the other hand, I would add that products that are lower down the stack, products like OpenFlow are taking off because they actually help make the other side of the data life cycle more efficient. I've used OpenFlow. It's pretty magical to be able to sink data, whether it's from an Oracle OLTP system or from Google Drive onto Snowflake, I think shrewd investments like that are also helping us substantially in just accelerating what people do with us. Previously, we used to be just the analytics provider, but we can be there from soup to nuts with products starting with OpenFlow, but then things like Snowpark, obviously, our analytics engine, then ML and then AI. It's where this breadth of offering and the complete data offering will end up playing a larger and larger role.

另一方面，我想補充一點，像 OpenFlow 這樣的底層產品正在蓬勃發展，因為它們實際上有助於提高資料生命週期另一端的效率。我用過OpenFlow。能夠將資料（無論是來自 Oracle OLTP 系統還是來自 Google Drive）匯入 Snowflake 真是太神奇了，我認為像這樣的精明投資也大大幫助我們加快了人們與我們合作的速度。以前，我們只是分析供應商，但現在我們可以提供從頭到尾的全方位服務，產品包括 OpenFlow、Snowpark（顯然是我們的分析引擎）、機器學習和人工智慧等。正是在這裡，這種廣泛的產品和服務以及完整的數據產品將發揮越來越重要的作用。

Brent Thill, Jefferies.

布倫特‧蒂爾，傑富瑞集團。

Thanks, Sridhar. Good to hear the news on AI bookings influenced. I guess many are now turning to the go-lives. And when do you expect this batch of go-lives to go up that then helps re-influence the even more excitement on the platform? How do you think about the trajectory. Does that have a bigger ramification in the back half of '26 then as those deals go live?

謝謝你，斯里達爾。很高興聽到人工智慧對預訂產生影響的消息。我想現在很多人都轉向直播了。那麼，您預計這批即將上線的內容何時發布，進一步激發平台上的玩家熱情？你如何看待這一發展軌跡？那麼，隨著這些交易在 2026 年下半年生效，這是否會產生更大的影響？

Well, you're seeing it live, right? We gave guidance for Q4. It's a pretty hefty beat and raise. And that is driven by what we see in consumption trends. As you know, we tend to be pretty disciplined about how we forecast and guide.

你是現場觀看的，對吧？我們給了第四季業績指引。這是一次相當大的漲跌幅。而這主要受消費趨勢的影響。如您所知，我們在預測和指導方面往往非常嚴謹。

These are based on machine learning models, unsurprisingly, that predicts the future. And we are disciplined in following that. On the other hand, we track the other side, which is how many use cases are we winning? What is the time duration from a win over to a technical implementation over to a go-live and accelerating go-lives will continue to be a priority. And we're using AI pretty heavily in making some of these use cases go live a whole lot faster as well.

不出所料，這些預測未來是基於機器學習模型的。我們嚴格遵守這些規定。另一方面，我們還要關注另一方面，也就是我們贏得了多少用例？從贏得專案到技術實施再到正式上線需要多長時間？加快專案上線速度仍將是我們的首要任務。而且，我們也大量使用人工智慧來加快其中一些用例的上線速度。

And all of these feeds into the forecast and guides and the general optimism that we convey to you.

所有這些都影響著我們的預測、指南以及我們向大家傳達的整體樂觀情緒。

Great. And if I can just for Brian, on Anthropic, the $200 million partnership, great to see. Is that in backlog? Or what goes into backlog from that relationship?

偉大的。如果我能為布萊恩說一句，關於 Anthropic 的 2 億美元合作，真是太好了。那項任務在待辦事項嗎？或者說，這段關係會造成哪些積壓問題？

The $200 million is a buy side that we're buying from Anthropic.

這 2 億美元是我們從 Anthropic 收購的股份。

And in some ways, obviously, our confidence in being -- in having AI drive more and more of our revenue, it is a commitment. But as you see the front side of things like the AI consumption revenue ARR that we announced, the $100 million ARR, that's what gives us confidence that partnerships with Anthropic, which include a buy, but also a broader go-to-market motion will continue to accelerate the overall business.

顯然，在某種程度上，我們對人工智慧能夠為我們帶來越來越多的收入充滿信心，這是一種承諾。但正如你所看到的，例如我們宣布的 AI 消費收入 ARR 達到 1 億美元，這讓我們有信心與 Anthropic 建立合作關係（包括收購以及更廣泛的市場拓展），將繼續加速整體業務發展。

Brad Zelnick, Deutsche Bank.

布拉德‧澤爾尼克，德意志銀行。

Hey guys, this is Dan on for Brad. Just wanted to ask maybe Sridhar to start. Just if you can kind of help frame the impact that migrations had to product revenue this quarter versus last quarter? I know there were some kind of unique circumstance last quarter where some positive things came together to drive a pretty strong result. But just in general, I think across all of the cloud names, we've seen pretty strong momentum this year.

大家好，我是丹，代布拉德報道。我想問Sridhar能不能先開始。您能否幫忙分析一下，與上個季度相比，遷移對本季產品收入的影響為何？我知道上個季度出現了一些特殊情況，一些正面因素匯聚在一起，促成了相當強勁的業績。但總的來說，我認為今年所有雲端服務供應商都呈現出相當強勁的成長勢頭。

And just as you look at kind of the visibility and pacing here that you have into that maybe just the sustainability of what you're seeing on that side? And then maybe one for Brian, just on operating margins. I think 4Q operating margin was guided maybe a couple of points below where you guided 3Q and maybe a little down from what was implied in the guide last quarter. Anything just to unpack on op margins into Q4 for us to think about as we build our models. Thanks.

就像你觀察這裡的可見性和節奏，以及你所看到的這種永續性一樣？然後或許可以給布萊恩提個建議，就營運利潤率而言。我認為第四季的營業利潤率預期可能比你對第三季的預期低幾個百分點，也可能比上個季度的預期略低一些。任何能幫助我們分析第四季度營運利潤率的信息，以便我們在建立模型時進行思考。謝謝。

I'll start. We are super early with migrations. I think you folks heard Matt Garman say today that he thinks maybe like we are 15% to 20% of the way through kind of on-prem legacy migrations. And that's positive news for Snowflake. And I see AI, I see products like Snowflake Intelligence exert both a powerful tool because the data that's in Snowflake just became more valuable because it can be used to drive business a whole lot more effectively, but I also see AI play a big role in pushing migrations forward, in other words, making the act of migrating from legacy systems go faster.

我先來。我們很早就開始進行遷移了。我想你們今天都聽到了 Matt Garman 的話，他認為我們可能已經完成了本地傳統系統遷移的 15% 到 20%。這對 Snowflake 來說是個好消息。我認為人工智慧，以及像 Snowflake Intelligence 這樣的產品，既是強大的工具，因為 Snowflake 中的資料變得更有價值，可以更有效地推動業務發展；而且我還看到人工智慧在推動遷移方面發揮著重要作用，換句話說，就是加快從傳統系統遷移的過程。

And this is where tuck-in acquisitions like the acquisition of Datometry, which makes products that make migrations go faster, easier are also helpful. We keep a close watch on migrations through the entirety of the use case life cycle, and it's something that we are continuously looking to accelerate, bring better techniques. It's an area that I've been personally involved with throughout the year, and we continue to make very solid progress.

而像收購 Datometry 這樣的小規模收購也正是在這種情況下發揮作用的，Datometry 生產的產品能夠讓遷移變得更快、更容易。我們會在整個用例生命週期中密切關注遷移情況，我們會不斷尋求加快遷移速度，引入更好的技術。今年以來，我一直親自參與這個領域的工作，我們持續取得非常紮實的進展。

Yes. Just real quickly on the 4Q guidance. All I would say is that 4Q is a little tricky in the sense that you're giving the 4Q guidance and the annual guidance at the same time. And so don't read too much into that. There's nothing intended by meant to read into that.

是的。關於第四季業績指引，我簡單說幾句。我只想說，第四季的情況有點棘手，因為你同時給了第四季的業績指引和年度業績指引。所以不要過度解讀這件事。這句話並沒有任何深意。

Raimo Lenschow, Barclays.

雷莫·倫肖，巴克萊銀行。

Thank you. One question to stick to that one question rule. Sridhar, zero-copy comes up a lot in the conversation. And like every vendor is now talking about like, oh, we're doing zero-copy that helps to kind of -- help us play better with everyone else in the ecosystem, et cetera. How do you think that will impact you? Is it kind of -- does it drive more adoption? Does it impact how much you can monetize? Can you speak to that, please? Thank you.

謝謝。只提一個問題，遵守只提一個問題的規則。Sridhar，零拷貝技巧在討論中常被提及。現在每個供應商都在談論，哦，我們正在做零複製，這有助於——幫助我們更好地與生態系統中的其他人合作等等。你認為這會對你產生什麼影響？它是否在某種程度上——它是否能促進更多人採用？這會影響你的獲利能力嗎？您能談談這個問題嗎？謝謝。

Yes. Zero-copy generally comes up in the context of SaaS vendors who are under a lot of pressure from their customers to share data. Many of them are busy creating data products on top of the data as a way to monetize, and zero-copy or sometimes bidirectional data sharing agreements come up in that context as a faster, more efficient way for people to share data with each other. We see these as a win-win. We have these agreements with, let's say, ServiceNow, Salesforce, SAP with the recent partnership as well as Workday.

是的。零複製通常出現在 SaaS 供應商的脈絡中，因為他們面臨來自客戶的巨大壓力，需要共享資料。許多企業正忙於利用資料創建資料產品以實現盈利，而零拷貝或有時是雙向資料共享協議正是在這種背景下出現的，它為人們提供了一種更快、更有效率的方式來共享資料。我們認為這是雙贏之舉。我們與 ServiceNow、Salesforce、SAP（最近與 SAP 建立了合作關係）以及 Workday 等公司都達成了合作協議。

These products continue to drive our broader mission to be at the center of all of the data needs that our customers have. And they just make the process of data collaboration between the SaaS vendors and Snowflake just a whole lot easier. And we are very happy with these agreements. And what this means is that Snowflake will continue to be the place for our customers to get like that single -- that stable single pane of glass sort of view on everything that matters to them. And obviously, with agentic AI and agentic systems now, the value that you can get from the data is tremendous.

這些產品將繼續推動我們實現更廣泛的使命，即成為客戶所有數據需求的核心。它們讓 SaaS 供應商和 Snowflake 之間的資料協作流程變得容易得多。我們對這些協議非常滿意。這意味著 Snowflake 將繼續成為我們的客戶獲取對他們而言重要的一切資訊的單一、穩定的單一視圖的地方。顯然，隨著智慧人工智慧和智慧系統的出現，從數據中可以獲得巨大的價值。

I can tell you from personal experience that I'm not thinking when I'm looking at my sales data agent, about whether this data comes from Workday or from Salesforce or from our own systems, I can focus on the logic of what needs to get done. And the rest of the stuff works as though it is magic. And so zero-copy agreements just make data flow more smoothly and I think are a big step forward for everybody involved Snowflake, but most importantly, our customers.

我可以根據個人經驗告訴你，當我查看銷售資料代理時，我不會去想這些資料是來自 Workday、Salesforce 還是我們自己的系統，我可以專注於需要完成的邏輯。其餘的東西也都像變魔術一樣有效。因此，零拷貝協議使資料流更加順暢，我認為這對所有參與 Snowflake 的人來說都是一大進步，但最重要的是，對我們的客戶也是如此。

Okay, perfect. Thank you. Bye.

好的，完美。謝謝。再見。

Mark Murphy, JPMorgan.

馬克墨菲，摩根大通。

This is Arti on for Mark Murphy. Thanks for taking my question. Congrats on the strong quarter and continued momentum. I know you've touched on this thing throughout the call here, but we spoke to a Fortune 150 customer, recently and they described Snowflake as the most important piece of their AI and data strategy and explicitly stated that Snowflake budget is now tied to their AI budget, and they're kind of broadening their adoption of products on the Snowflake platform. So my question is, are you kind of seeing that sort of tying explicitly from customers of their Snowflake investments to the AI investments? And if so, how is this influencing the buying habits?

這是 Arti 為 Mark Murphy 所做的報導。謝謝您回答我的問題。祝賀公司本季業績強勁，並保持了持續成長勢頭。我知道你在整個通話過程中都提到了這一點，但我們最近與一家財富 150 強客戶進行了交談，他們將 Snowflake 描述為其人工智能和數據戰略中最重要的一部分，並明確表示 Snowflake 的預算現在與他們的人工智能預算掛鉤，他們正在擴大在 Snowflake 平台上的產品的應用範圍。所以我的問題是，您是否看到客戶將他們在 Snowflake 上的投資與在 AI 上的投資明確地連結起來？如果真是如此，這會對購買習慣產生怎樣的影響？

Are they entering into larger, longer-term contracts? Are they adopting more products or just any new customer patterns you're seeing emerge?

他們是否正在簽訂金額更大、期限更長的合約？他們是否採用了更多產品，或者只是出現了一些新的客戶模式？

Yes. The strongest pattern that we have had to work hard and earn this year is to be that genuine player when it comes to enterprise AI. And no amount of talking can make you that. You need project, products that produce the magic. And so building on earlier products like Cortex Analyst as well as Cortex Search, Snowflake Intelligence, the agentic platform that can use these different subproducts flexibly is the big unlock for us.

是的。今年我們努力奮鬥並贏得的最重要一點，就是成為企業人工智慧領域的真正參與者。再多的言語也無法讓你變成那樣。你需要能夠創造奇蹟的項目和產品。因此，在 Cortex Analyst 和 Cortex Search、Snowflake Intelligence 等早期產品的基礎上，能夠靈活使用這些不同子產品的代理平台對我們來說是一個重大突破。

And what you're also seeing is a number of these customers have tried to string together agent systems by, let's say, creating MCP servers on tables and sticking them into our foundation model. And then they realize that solutions like that don't actually work all that effectively. Part of what we provide are systems that can help them thoughtfully structure the data that then needs to be exposed to an AI agent and a careful amount of tuning that makes sure that these systems are failsafe, that they're reliable and can actually answer the questions they are supposed to. We also work with our customers on things like unheralded, but really important things like eval where they can judge ongoing performance so that they know that they're actually making their systems better. It's a combination of all of this expertise.

而且您還會看到，許多客戶嘗試將代理系統串聯起來，例如，在表上建立 MCP 伺服器，並將它們插入到我們的基礎模型中。然後他們意識到，像這樣的解決方案實際上並沒有那麼有效。我們提供的部分服務包括：幫助他們精心建構資料結構，然後將這些資料暴露給 AI 代理程式；以及進行細緻的調整，以確保這些系統安全可靠，能夠真正回答它們應該回答的問題。我們也會與客戶合作處理一些不為人知但非常重要的事情，例如評估，以便他們可以判斷持續的效能，從而知道他們實際上正在改進他們的系統。這是所有這些專業知識的綜合體現。

Yes, the partnership with the big foundation model providers to bring the best models as part of Snowflake, combined with our unrivaled expertise in data and modeling to help them create AI products that deliver value. And if you combine that with products like Snowflake Intelligence that now like are clearly valuable and useful for every business user, I think that's the narrative shift that you're seeing in a number of these companies.

是的，我們與大型基礎模型供應商合作，將最好的模型引入 Snowflake，結合我們在數據和建模方面無可比擬的專業知識，幫助他們創建能夠創造價值的 AI 產品。如果再加上像 Snowflake Intelligence 這樣的產品，它們現在顯然對每個企業用戶都很有價值和實用，我認為這就是你在許多這類公司中看到的敘事轉變。

And agentic AI is still evolving. We have a lot more to do. That's part of the reason why I keep repeating being in the center of enterprise AI because we are already the holders of the most valuable data that many of these enterprises have and then we are bringing the power of AI to get even more value from this data.

智能體人工智慧仍在不斷發展。我們還有很多工作要做。這就是我一直強調要成為企業人工智慧中心的原因之一，因為我們已經掌握了許多企業擁有的最有價值的數據，然後我們正在利用人工智慧的力量從這些數據中獲得更多價值。

Thanks so much, very insightful narrative shift. I think is a great way to describe it.

非常感謝，這個敘述轉變很有見地。我認為這是一個非常貼切的描述方式。

Kash Rangan, Goldman Sachs.

Kash Rangan，高盛集團。

Hey, this is Matt Martino on for Kash. This is Matt Martino on for Kash. Sridhar, I want to stick with the AI topic here. The number of customers leveraging Snowflake AI is accelerating very, very quickly within your installed base, and you're able to pull forward that $100 million in AI revenue, which very few of your peers have been able to do. From your perspective, what about the Snowflake platform is allowing customers to really accelerate their AI journeys? And maybe secondarily, do you see the market increasingly standardizing around a smaller subset of platforms to handle all their data requirements given your commentary about Snowflake really sitting at the center of the AI opportunity? Thanks a lot.

大家好，我是馬特·馬蒂諾，為您帶來卡什的報道。這是馬特馬蒂諾替補卡什上場。Sridhar，我想繼續討論人工智慧這個主題。在您的現有客戶群中，使用 Snowflake AI 的客戶數量正在迅速增長，您能夠提前實現 1 億美元的 AI 收入，而您的同行中很少有人能夠做到這一點。從您的角度來看，Snowflake 平台有哪些特質能真正幫助客戶加速其 AI 轉型之旅？其次，鑑於您曾評論 Snowflake 真正處於人工智慧機會的核心，您是否認為市場會越來越傾向於圍繞一小部分平台進行標準化，以處理其所有資料需求？多謝。

Yes. I think to take on your second question first, I think there is a lot of complexity in the data space. I know of the number of different tools that Snowflake, the company itself has had to use to have an effective data strategy and with things like Snowflake Intelligence and Streamlit, which we are very heavy users of, we are just able to do more with Snowflake. And again, investments like OpenFlow or even Postgres are going to expand the aperture of what we are able to tackle as the data platform.

是的。我認為應該先回答你的第二個問題，數據領域非常複雜。我知道 Snowflake 公司為了製定有效的資料策略，不得不使用許多不同的工具，而像 Snowflake Intelligence 和 Streamlit 這樣的工具（我們都是它們的重度使用者），讓我們能夠利用 Snowflake 做更多的事情。此外，像 OpenFlow 甚至 Postgres 這樣的投資將擴大我們作為資料平台能夠處理的資料範圍。

Alex Zukin, Wolfe Research.

Alex Zukin，Wolfe Research。

Yeah, hey, thanks guys for taking the question. Maybe for either of you, Brian or Sridhar, clearly, the momentum that you're describing is showing up in bookings. So I just maybe better understanding the confidence and conviction around -- and maybe the direction of travel for the expansion rate as we continue to see some of these go-lives and an explanation of how the consumption patterns, particularly as you start to see customers leverage the AI portfolio and the other -- and the multiproduct portfolio more broadly? And then, Brian, any timing elements? Last quarter, it seemed like there was a little bit more of a onetime bump or boost to product revenues from consumption from some very large deals in the quarter. But then this quarter, you also had 4 super large deals.

是的，嘿，謝謝各位回答這個問題。或許對你們倆，布萊恩或斯里達爾來說，你們所描述的這種勢頭顯然已經體現在預訂量上了。因此，我或許能更好地理解人們對擴張速度的信心和信念——以及擴張速度的發展方向，因為我們將繼續看到一些產品上線，並解釋消費模式是如何變化的，尤其是在您開始看到客戶利用人工智慧產品組合和其他產品組合——以及更廣泛的多產品組合的情況下？那麼，布萊恩，時間安排方面有什麼考量嗎？上個季度，由於一些非常大的交易，產品收入似乎出現了一次性的成長或提升。但本季，你們也達成了 4 筆超級大單。

So is there something where they maybe happened a little bit later. And last quarter, they happened a little bit earlier that maybe drove that beat magnitude cadence to be a little lower.

那麼，有沒有可能有些事情發生得稍晚呢？上個季度，這些事件發生的時間稍微提前了一些，這可能導致心跳幅度節奏略微降低。

I can start with the first one. The virtuous cycle of Snowflake customer is one in which they sign a deal. It has a certain amount of slack capacity that is built into it, that our teams then use to expand into use cases that can deliver value for our customers. And to actually address a previous question that I had left unaddressed, that was the first part of the previous question. Part of what drives broad adoption of AI with Snowflake is that we make it easy to do.

我可以從第一個開始。Snowflake 客戶的良性循環就是他們簽約的過程。它內建了一定的冗餘容量，我們的團隊可以利用這些容量擴展到可以為客戶創造價值的用例中。為了真正回答我之前未回答的一個問題，那就是上一個問題的第一部分。Snowflake之所以能被廣泛採用人工智慧，部分原因是它讓人工智慧的實作變得簡單。

It's not a brand-new system. You don't have to resolve the existing problems like governance and access controls. And we have made it super easy to first build chatbots and then to build more complex agentic systems like Snowflake Intelligence, which is why some 1,200% -- 1,200 customers are already using Snowflake Intelligence. And as we expand and deliver value, these then naturally result in more confidence in more conviction on the part of the customer that they're getting value from Snowflake. And remember, in all of this, they don't have to make any pre-commits towards AI.

這並非一個全新的系統。您無需解決現有的治理和存取控制等問題。我們讓建造聊天機器人變得極其簡單，然後讓建造像 Snowflake Intelligence 這樣更複雜的代理系統變得非常簡單，這就是為什麼大約 1200%——1200 家客戶已經在使用 Snowflake Intelligence 的原因。隨著我們不斷成長並創造價值，客戶自然會更加確信他們從 Snowflake 獲得了價值。請記住，在這一切中，他們無需對人工智慧做出任何預先承諾。

The value that they get is like it has to be delivered by the products that they build on top of Snowflake. This risk-free approach driven by our consumption model is what makes AI super attractive for our customers on top of Snowflake. We make it easy to use, we don't require them to commit and then they naturally expand out the ones that are creating value. And I'll just touch on the second part of your question, and I'll hand off to Brian. Large deals that we sign don't tend to have immediate impact on revenue within the quarter.

他們獲得的價值，就像必須透過他們在 Snowflake 之上建造的產品來實現一樣。這種由我們的消費模式驅動的無風險方法，使得 AI 在 Snowflake 之上對我們的客戶極具吸引力。我們讓它易於使用，我們不要求他們做出承諾，然後他們自然而然地擴展那些創造價值的功能。關於您問題的第二部分，我只簡單回答一下，然後就交給布萊恩了。我們簽署的大額交易通常不會立即對當季收入產生影響。

If anything, as soon as a large deal is signed, they typically get a better discount. So it tends to be slightly negative with respect to revenue. But as I said, these are long-term cycles. Our customers on average sign deals with us once every 2.5 to -- 2.5 years-ish on average. It's not really directly tied to consumption and within a quarter, and I would not read too much into timing constraints like that.

通常情況下，一旦簽訂大額合同，他們就能獲得更大的折扣。因此，就收入而言，它往往略微為負。但正如我所說，這些都是長期週期。我們的客戶平均每 2.5 到 2.5 年左右與我們簽訂合約。它與消費量和季度內的情況並沒有直接關係，我不會過度解讀這類時間限制。

Brian?

布萊恩？

Yes. Absolutely, Sridhar. I guess I would emphasize that product revenue is still the leading indicator of our business, and we saw that in really the migrations and increased use case wins. We're also happy with the developments in AI and also the data engineering workloads. We look at the consumption patterns up until today to inform our view of Q4.

是的。當然，斯里達爾。我想強調的是，產品收入仍然是我們業務的領先指標，這一點我們從用戶遷移和不斷增加的使用案例中也確實看到了。我們對人工智慧的發展以及資料工程的工作量也感到滿意。我們透過分析截至目前為止的消費模式來了解我們對第四季的看法。

The quarterly beats are less indicative, especially in a consumption model. I would really look at the FY guidance as the best indication of the long-term business trends for a consumption model. Our view of the business over the last 90 days has improved. And I think you can see that in our annual raise. This is also represented in the $7.9 billion in RPO, 37% year-over-year growth and all the new customer adds that we talked about in the prepared remarks.

季度業績超預期較不具指示意義，尤其是在消費模式下。我認為，財年績效指引是衡量消費模式長期業務趨勢的最佳指標。過去90天裡，我們對這家公司的看法有所改善。我認為這一點可以從我們每年的加薪中看出。這一點也體現在 RPO 的 79 億美元、年成長 37% 以及我們在準備好的發言稿中提到的所有新增客戶中。

Patrick Colville, Scotiabank.

派崔克‧科爾維爾，加拿大豐業銀行。

Thank you for having me on. I guess, Sridhar and Brian, one for both of you, please. You passed the $100 million consumption thresholds, really impressive to see that. I guess what do you see as the next milestone? And then could you just remind us what does that $100 million actually include? Is that equivalent to the Cortex suite?

謝謝邀請我參加節目。我想，Sridhar 和 Brian，請給你們兩個一個人一杯。你們的消費額突破了 1 億美元大關，真是令人印象深刻。我想問您，您認為下一個里程碑是什麼？那麼，您能否提醒我們一下，這1億美元究竟包含哪些內容呢？這和Cortex套件等效嗎？

Or are there other products that go into that $100 million of consumption that you achieved this quarter? Thank you.

或者，你們本季實現的 1 億美元消費額中，還有其他產品嗎？謝謝。

Yes. The $100 million is primarily the product suite, but it's the whole stack. It is Cortex AI and AI SQL. It's accessible from SQL also as a REST API. And then the products that stack up on top of that Cortex Search and Cortex Analyst, which are our unstructured and structured data products, respectively.

是的。這 1 億美元主要用於產品套件，但它涵蓋了整個技術堆疊。它是Cortex AI和AI SQL。它既可以透過 SQL 訪問，也可以作為 REST API 存取。然後，還有基於 Cortex Search 和 Cortex Analyst 的產品，它們分別是我們的非結構化資料產品和結構化資料產品。

And then Snowflake Intelligence, which builds on these building blocks to provide an agentic solution for data products. That's roughly the suite. In terms of the next milestone, I think much broader adoption of Snowflake Intelligence is certainly that we are driving. There is no reason for us to not have every single data set that is in Snowflake, be AI-ready. And you're already seeing this play out in the collaboration space where instead of sharing a data set, you can, in fact, share an agent on top of that data set so that the recipient on the other side can straight out just start asking business questions of this data without needing to build dashboards and so on.

然後是 Snowflake Intelligence，它基於這些建置模組，為資料產品提供智慧體解決方案。套房大致就是這樣。就下一個里程碑而言，我認為我們正在努力推動 Snowflake Intelligence 的更廣泛應用。我們完全有理由讓 Snowflake 中的每個資料集都做好人工智慧的準備。而且，在協作領域，我們已經看到這種情況正在發生：與其共享資料集，不如共享基於該資料集的代理，這樣另一端的接收者就可以直接開始詢問有關這些資料的業務問題，而無需建立儀表板等等。

Obviously, in many situations, this data flows through programmatically and will be combined with other data. But my point is making all data in Snowflake AI consumable and making the act of making that AI consumable is something that we will be -- honestly be spending a lot of time on. But the second and the third order impacts that I alluded to earlier, the pull-push analogy that I used. I think that's where the impact is going to be a lot more profound. I think migrating from legacy systems, bringing data into Snowflake using products like OpenFlow or being able to write data engineering workloads using our coding agents.

顯然，在許多情況下，這些數據是透過程式自動傳輸的，並將與其他數據結合起來。但我的重點是，讓 Snowflake AI 中的所有數據都變得易於使用，以及讓 AI 變得易於使用，是我們將會花費大量時間去做的事情。但我之前提到的第二和第三階影響，也就是我使用的拉推類比。我認為，這方面的影響將會更加深遠。我認為，從傳統系統遷移，使用 OpenFlow 等產品將資料導入 Snowflake，或能夠使用我們的編碼代理程式編寫資料工程工作負載。

All of those are going to get accelerated. I think that's where you're going to see like tremendous value that our customers can realize and tremendous potential for us as a business.

所有這些都將加速發展。我認為，這正是我們的客戶所能獲得的巨大價值，也是我們作為一家企業所擁有的巨大潛力所在。

Brad Reback, Stifel.

Brad Reback，Stifel。

Great, thanks very much. Sridhar, the results are very impressive. The booking is super great. The op margin obviously down-ticked on the first half sales and marketing investment. As we look forward into next year and beyond, how do you think about balancing the huge opportunity in front of you and the ability to drive margin expansion?

太好了，非常感謝。Sridhar，結果非常令人印象深刻。預訂體驗非常棒。由於上半年的銷售和行銷投入，營業利潤率明顯下降。展望明年及以後，您認為如何平衡眼前的巨大機會和推動利潤率成長的能力？

I think we live in fortunate times where this is not an either/or. We clearly invested pretty heavily in our sales and marketing teams in the first 2 quarters because we saw a tremendous opportunity. And what we're going through now is a maturation of the folks that are here, and we expect them to aid us substantially. But we have also invested equally heavily in how do we make sure that we upskill our own labor force, whether it is engineers or solution engineers. We have rolled out coding agents for the folks.

我認為我們生活在一個幸運的時代，這並非非此即彼的問題。前兩個季度，我們對銷售和行銷團隊進行了相當多的投資，因為我們看到了巨大的機會。我們現在正經歷的是團隊成員的成熟過程，我們希望他們能給予我們實質的幫助。但我們也同樣投入巨資，確保提升自身勞動力的技能，無論是工程師或解決方案工程師。我們已經為民眾推出了編碼代理。

I talked earlier about how we want to make it super easy for every single rep, every single solution engineer to be able to do custom demos, custom POCs for our customers. Obviously, we have a big services team as well, making then AI native is a big transformation. So the way we will -- the way we look at next year is, yes, we will continue to invest in the business, but I think there is also substantial gains to be had in just how efficient we are as a company. And I don't think of this as an either/or. We have had pretty healthy expansions in things like operating margin, but also things like SBC year-over-year, and we will continue to press hard on those things.

我之前說過，我們希望讓每位銷售代表、每位解決方案工程師都能非常輕鬆地為我們的客戶進行客製化簡報和客製化概念驗證。顯然，我們還有一個龐大的服務團隊，要讓AI原生化是一個巨大的轉變。所以，我們展望明年的方式是，是的，我們將繼續投資這項業務，但我認為，我們作為一家公司，在提高效率方面也有很大的提升空間。我不認為這是一個非此即彼的問題。我們在營業利潤率、SBC 等指標上都取得了相當健康的年成長，我們將繼續努力在這些方面取得進展。

Yes. I'll just echo what Sridhar said. We can do both. It's not one or the other. Obviously, it's a really big market, and we've delivered impressive growth and we'll continue to do innovations in our product to drive that revenue growth, but we'll do that responsibly.

是的。我只想重複一下斯里達爾的話。我們可以兩者兼顧。這不是二選一的問題。顯然，這是一個非常大的市場，我們已經取得了令人矚目的成長，我們將繼續在產品上進行創新以推動收入成長，但我們將以負責任的方式進行創新。

Mike Cikos, Needham.

麥克·西科斯，尼德姆。

Great. Thanks for taking the questions, guys. And Brian, congratulations again on the new role as CFO over at Snowflake. Looking forward to working together here. My question comes back to -- I think there's been a couple of different attempts throughout this call at understanding, frankly, the magnitude of the product revenue upside relative to the prior quarter, where, to be frank, last quarter was more significant, but I really -- I attributed it or I thought that you guys positioned it as last quarter really saw some very large customer migrations, which is outside your control. And so the question is, when we think about the increased confidence you're talking about for the year, the traction for data engineering and AI, is it fair to think that 3Q here was just a strong execution quarter but maybe a more normalized return to typical migration activity?

偉大的。謝謝各位回答問題。Brian，再次恭喜你榮任 Snowflake 的財務長一職。期待與大家共事。我的問題又回到了——我認為在這次通話中，你們已經嘗試了好幾次，坦白說，來理解產品收入相對於上一季的成長幅度。坦白講，上一季的成長更為顯著，但我真的——我將其歸因於，或者我認為你們將其定位為上一季確實出現了一些非常大規模的客戶遷移，而這超出了你們的控制範圍。因此，問題是，當我們考慮到您所說的今年信心增強、數據工程和人工智慧的發展勢頭時，是否可以認為第三季度只是一個強勁的執行季度，而是一個更加正常的回歸典型遷移活動的季度？

And then secondarily, just while we have everyone on the phone here, Brian, we would love to get your perspective on whether the guidance philosophy has changed at for the margin.

其次，布萊恩，既然大家都在電話裡，我們很想聽聽你對邊際指導理念是否有所改變的看法。

Yes. I'll start with the first one. We've consistently told all of you that we view a 3% beat as a very good beat. And anytime we do much better than that, we go back. Obviously, the ML models recalibrate and we calibrate ourselves back to the 3% beat.

是的。我先從第一個開始。我們一直告訴大家，我們認為 3% 的超預期收益已經是非常好的超預期收益了。每次我們做得比這好得多的時候，我們就會回到原來的狀態。顯然，機器學習模型會重新校準，我們也需要重新校準自己，以達到 3% 的節拍。

So -- and there is also a natural variability in a consumption business because this is literally the agglomeration of 12,000-plus enterprises deciding what they want to do with their data futures. And so I view the Q3 beat as actually still a very solid beat at some 2.5%. And yes, the Q2 beat, and we are upfront with you about it, had some large migrations that also had onetime activities, but we also have cautioned to you that large migrations are lumpy and not all that easy to predict. And that's roughly where we are. And as we look at things like Q4, we approach it the exact same way.

所以——消費業務也存在著自然的變數，因為這實際上是 12,000 多家企業決定如何處理其資料未來的集合體。因此，我認為第三季業績實際上仍然非常出色，超出預期約 2.5%。是的，第二季業績超乎預期，我們也坦誠地告訴大家，這其中包含一些大規模遷移以及一次性活動，但我們也提醒過大家，大規模遷移具有波動性，而且不容易預測。我們現在的情況大致就是這樣。當我們審視第四季等問題時，我們採取的方法也完全相同。

We do the best job that we can of trying to figure out where we are going to land and use pretty much the same guidance philosophy as we have before. Brian?

我們會盡最大努力弄清楚我們將要降落在哪裡，並採用與以前幾乎相同的指導理念。布萊恩？

Yes. Thanks, Sridhar. Just to echo what Sridhar said as well, the quarterly variability is not the right way to evaluate the consumption model. Companies that do migrations, they don't do those due to our quarterly earnings calls. They basically -- we snap the chalk line and where they're at and their migrations are at.

是的。謝謝你，斯里達爾。正如 Sridhar 所說，季度波動並不是評估消費模式的正確方法。那些從事遷移工作的公司，由於我們的季度財報電話會議，不會進行這些遷移。基本上，我們用粉筆線標出它們所在的位置以及它們的遷徙路線。

And so we really point you to the full year guide. And based on the behavior that we've seen up to the earnings call, we have the confidence to raise our full year guide to $51 million to 28% annual growth year-over-year. Just from a guidance philosophy perspective, there's a number of things that I did when I first joined, but one of the things that I did was spent a lot of time with the team that wrote all the AI models. It does the forecasting on a daily business of our revenue. Super impressive team, very detailed, and I can assure you that there will be no change to the guidance philosophy.

因此，我們強烈建議您參考全年指南。根據我們在財報電話會議之前所看到的表現，我們有信心將全年業績預期上調至 5,100 萬美元，年增 28%。從指導理念的角度來看，我剛加入公司時做了很多事情，其中​​一件事情就是花了很多時間和編寫所有人工智慧模型的團隊在一起。它負責預測我們每日的業務收入。團隊非常出色，做事非常細緻，我可以向你保證，指導理念不會有任何改變。

Matt Hedberg, RBC.

Matt Hedberg，RBC。

Great, thanks for taking my question. Just a quick one for Sridhar. The $100 million AI run rate is super impressive. Wondering if you could give us just a rough sense for how quickly that's growing? And then maybe more of a detailed question. On the heels of crunchy data, curious if you can comment about just now that you've had more time, how customers are thinking about that long-term balance of OLTP and OLAP within Snowflake?

太好了，謝謝你回答我的問題。給斯里達爾簡單提一下。人工智慧年收入達1億美元，這真是令人印象深刻。想請您大概介紹一下這個成長速度如何？然後或許會提出更詳細的問題。在掌握了大量數據之後，我很好奇您現在有了更多時間，能否談談客戶是如何看待 Snowflake 中 OLTP 和 OLAP 的長期平衡的？

Yes. As I said, AI revenue is predominantly driven by the Cortex product suite, including Snowflake Intelligence. This is among the fastest products to get adopted by our customers because, as I said, the value is very, very clear as soon as someone uses Snowflake Intelligence. So we expect to -- we expect for this to continue to grow quite well. We don't really want to guide to it or hint at that right now.

是的。正如我所說，人工智慧收入主要由 Cortex 產品套件（包括 Snowflake Intelligence）驅動。這是我們的客戶接受度最高的產品之一，因為正如我所說，一旦有人使用 Snowflake Intelligence，它的價值就非常非常明顯。因此我們預計——我們預計這種情況會繼續保持良好的成長勢頭。我們現在不想對此進行引導或暗示。

With respect to Crunchy Data, it will take us a couple of more months to get the product into GA, but all of the early conversations that we have is that customers are very welcoming of Postgres support within Snowflake. They view Snowflake as an incredibly robust and reliable data platform. And for many kinds of applications, having them be hosted as part of the overall Snowflake deployment makes perfect sense for these folks. For what it's worth, Unistore, which is our HTAP product, is also doing well. It addresses a different segment of the transactional data space.

關於 Crunchy Data，我們還需要幾個月的時間才能將產品正式發布，但我們早期的所有交流都表明，客戶非常歡迎 Snowflake 中對 Postgres 的支援。他們認為 Snowflake 是一個極為強大可靠的數據平台。對於許多類型的應用程式來說，將它們託管在 Snowflake 的整體部署中是完全合理的。值得一提的是，我們的 HTAP 產品 Unistore 也表現良好。它針對的是交易資料領域中不同的部分。

And we will continue to have both of these. But I think bringing Postgres to market will be an important step forward for us, especially for things like agentic solution that need an OLTP store to function effectively. So there are a number of those kinds of use cases that we are actively working with our customers on.

我們將繼續擁有這兩者。但我認為將 Postgres 推向市場對我們來說將是向前邁出的重要一步，特別是對於像代理解決方案這樣需要 OLTP 儲存才能有效運作的應用而言。因此，我們正在積極與客戶合作，研究許多此類應用案例。

Tyler Radke, Citi

泰勒‧拉德克，花旗銀行

Thanks very much for squeezing me in. Really impressive to see roughly $1 billion of RPO bookings in the quarter. I was hoping you could talk a little bit about the 3 9-figure deals that you added in the quarter. How are those sort of structured from a duration perspective? And are you expected to see significant growth in those deals? In other words, were they large expansions?

非常感謝您擠出時間幫我。本季RPO訂單金額接近10億美元，確實令人印象深刻。我希望您能談談本季新增的三筆九位數交易。從時長角度來看，這些內容是如何組織的？您預計這些交易會有顯著成長嗎？換句話說，它們是大規模擴張嗎？

And then just a follow-up for you, Brian. Anything we should be thinking about as it relates to FY27, whether it's headwinds or tailwinds in the model? I know you're not giving guidance, but just as we think about new products, optimization headwinds, anything you'd call out?

布萊恩，最後還有一個問題想問你。關於 2027 財年，我們應該考慮哪些因素？模型中是利好因素還是不利因素？我知道您不是在給予指導意見，但是當我們考慮新產品、優化方面的阻力時，您有什麼想指出的嗎？

Thank you very much.

非常感謝。

Well, as a matter of fact, Tyler, we had 4 9-figure deals this quarter. All of these folks are customers that are spending significantly with Snowflake and are very positive about additional value that they can bring. But bookings are an indicator of how much a customer thinks they're going to spend in the coming years. Product revenue is the best indicator of how our collective customers are going to be spending on Snowflake next quarter. And so that's the thing that I would look at.

事實上，泰勒，我們這個季度完成了 4 筆九位數的交易。這些人都是 Snowflake 的大客戶，他們對 Snowflake 能帶來的額外價值非常樂觀。但預訂量可以顯示顧客預計未來幾年會花多少錢。產品收入是衡量我們所有客戶下個季度在 Snowflake 上支出狀況的最佳指標。所以，這就是我會關注的地方。

Brian, do you want to take the last question?

布萊恩，你想回答最後一個問題嗎？

Yes. Tyler, as you mentioned, we'll guide to FY27 on our next call. But what's really important is consumption after the holiday season, is the most important input for FY guidance for next year. And so we'll need to see the consumption behavior unfold in January, February, and that will give us better visibility to deliver that on our next earnings call.

是的。泰勒，正如你所說，我們將在下次電話會議上展望2027財年。但真正重要的是節後消費狀況，這才是明年財政年度業績指引的最重要依據。因此，我們需要觀察一月和二月的消費行為，這將使我們在下次財報電話會議上更了解情況並給出相應的答案。

At this time, I'd like to pass the conference back over to our host, Sridhar, for closing remarks.

此時，我想把會議交還給我們的東道主斯里達爾，請他作閉幕致詞。

Thank you, everyone. Snowflake remains at the center of today's enterprise AI revolution. And we at Snowflake are focused on empowering our customers throughout the end-to-end life cycle for data. This is an incredibly exciting time for the company as we continue to reimagine what's possible with AI and push the boundaries of innovation to lead in this new era. We continue to execute strongly as evidenced by our product revenue growth and strong outlook for the remainder of fiscal '26, and we see a long runway of durable high growth and continued margin expansion ahead.

謝謝大家。Snowflake 仍然是當今企業人工智慧革命的核心。Snowflake 致力於在資料的端到端生命週期中賦能我們的客戶。對公司而言，這是一個令人無比令人興奮的時刻，我們將繼續重新構想人工智慧的可能性，並突破創新界限，引領這個新時代。我們的產品收入成長和對 2026 財年剩餘時間的強勁展望證明了我們持續強勁的執行力，我們看到未來將有很長一段時間保持持久的高成長和持續的利潤率擴張。

Thank you all.

謝謝大家。

That will conclude today's conference call. Thanks for your participation and enjoy the rest of your day.

今天的電話會議到此結束。感謝您的參與，祝您今天餘下的時間愉快。