Confluent Inc (CFLT) 2025 Q1 法說會逐字稿

(audio in progress)

（音訊正在播放）

Thanks, Shane. Good afternoon, everyone, and welcome to our first quarter earnings call. We're pleased to start the year with solid momentum. Q1 subscription revenue grew 26% to $261 million. Confluent Cloud revenue grew 34% to $143 million and non-GAAP operating margin improved 6 percentage-points to 4%.

謝謝，肖恩。大家下午好，歡迎參加我們的第一季財報電話會議。我們很高興以強勁的勢頭開始新的一年。第一季訂閱營收成長 26%，達到 2.61 億美元。Confluent Cloud 營收成長 34% 至 1.43 億美元，非 GAAP 營業利潤率提高 6 個百分點至 4%。

Our Q1 results demonstrate the mission-critical nature of data streaming and our significant product leadership. We remain laser-focused on enabling our customers to cost efficiently build next-generation applications and win in the age of AI. On today's call, I'd like to walk through four key strategic drivers behind our resilience in the current environment. In fact, our resilience has been tested by multiple macro headwinds since Confluent's founding over 10 years ago, and it continues to support our path towards delivering long-term growth and profitability. First and most critically, data streaming sits at the heart of the mission-critical use cases our customers rely on every day.

我們的第一季業績證明了資料流的關鍵任務性質以及我們重要的產品領導地位。我們始終專注於幫助客戶以經濟高效的方式建立下一代應用程式並在人工智慧時代取得勝利。在今天的電話會議上，我想介紹我們在當前環境下保持韌性的四個關鍵策略驅動力。事實上，自 Confluent 成立十餘年以來，我們的韌性已經經受了多重宏觀逆風的考驗，並且它繼續支持我們實現長期成長和獲利的道路。首先，也是最關鍵的一點，資料流是我們客戶每天所依賴的關鍵任務案例的核心。

These are not lightweight experiments through the backbone of real production workloads. These applications play a critical role in the day-to-day business of our customers and can't be turned off without major disruption to core parts of their business. Confluent powers real-time fraud detection for financial institutions, inventory management for retailers, 5G networks for telcos and data pipelines that power businesses and countless industries. This dynamic has helped sustain gross retention rate above 90% despite multiple points of instability in the macro environment that reduced IT spend for some of our customers. The second driver of our resilience is the scale of the opportunity we're going after.

這些並不是透過實際生產工作負載主幹進行的輕量級實驗。這些應用程式在我們客戶的日常業務中發揮著至關重要的作用，如果關閉它們，則會對其業務的核心部分造成嚴重破壞。Confluent 為金融機構提供即時詐欺偵測，為零售商提供庫存管理，為電信公司提供 5G 網絡，並為企業和無數行業提供資料管道。儘管宏觀環境存在多個不穩定因素，導致部分客戶減少了 IT 支出，但此動態仍有助於維持 90% 以上的總留存率。我們的韌性的第二個驅動力是我們所追求的機會的規模。

Apache Kafka has become a foundational technology for data management. Today, it's used by more than 150,000 organizations, representing $100 billion-plus addressable market opportunity. Soaking up the countless use cases built on Kafka has been a core engine of our growth and continues to be a powerful tailwind. This quarter, we added 340 new customers, our highest net add in three years, and we continue to see robust growth with many of our largest customers as they expand to new use cases. In Q1, we added 16 new customers to our cohort of $1 million-plus ARR customers, our highest addition to that cohort ever.

Apache Kafka 已成為資料管理的基礎技術。如今，它已被超過 15 萬個組織使用，代表著超過 1,000 億美元的潛在市場機會。吸收基於 Kafka 構建的無數用例一直是我們成長的核心引擎，並將繼續成為強大的順風。本季度，我們增加了 340 個新客戶，這是三年來的最高淨增客戶數，隨著我們的許多最大客戶擴展到新的用例，我們繼續看到他們的強勁成長。在第一季度，我們增加了 16 位新客戶，加入了 100 萬美元以上 ARR 客戶群，這是我們有史以來對該客戶群的最高增幅。

Our customer base is diversified by industry and by geography, and no single customer accounts for more than 2% of our total ARR. This diversification further strengthens the resilience and durability of our business. Odyssey and Booking.com are two great examples of customers that started with open source and then converted to Confluent. Odyssey is a leading audio entertainment company with 200 million listeners across radio broadcast, podcasts and other digital content. As their audience grew and customer expectations rose, their old infrastructure began to hold them back.

我們的客戶群按行業和地理多樣化，沒有一個單一客戶占我們總 ARR 的 2% 以上。這種多樣化進一步增強了我們業務的彈性和持久性。Odyssey 和 Booking.com 就是兩個很好的例子，它們都是從開源開始，然後轉向 Confluent 的客戶。Odyssey 是一家領先的音訊娛樂公司，擁有 2 億廣播、播客和其他數位內容聽眾。隨著受眾的成長和客戶期望的提高，他們舊的基礎設施開始阻礙他們的發展。

Developers were spending most of their time wrestling with brittle point-to-point integrations built with open-source Kafka. These were difficult to manage and even harder to scale. Delivering new features at the pace the business needed became a major challenge. That's when Odyssey turned to Confluent. Our prebuilt connectors allow them to easily integrate data streaming into their existing tools and systems.

開發人員大部分時間都在努力解決使用開源 Kafka 建構的脆弱的點對點整合問題。這些都很難管理，甚至更難擴展。按照業務所需的速度提供新功能已成為一項重大挑戰。正是在那時，Odyssey 轉向了 Confluent。我們預先建立的連接器使他們能夠輕鬆地將資料流整合到他們現有的工具和系統中。

Stream Governance enables them to quickly and easily enforce data consistency across dozens of different systems and applications. With our complete data streaming platform, Odyssey freed up developer resources so they can focus on innovation instead of managing infrastructure. As a result, Odyssey accelerated feature development by 40%, delivering more personalized experience to keep their customers listening. This helps unlock new digital revenue opportunities across their platform. Booking.com is one of the world's largest online travel agencies.

流治理使他們能夠快速輕鬆地在數十個不同的系統和應用程式之間實現資料一致性。借助我們完整的資料流平台，Odyssey 釋放了開發人員資源，使他們能夠專注於創新而不是管理基礎設施。結果，Odyssey 將功能開發速度提高了 40%，提供了更個人化的體驗，讓客戶持續專注。這有助於在他們的平台上開啟新的數位收入機會。Booking.com 是全球最大的線上旅行社之一。

Its mobile app is one of the most downloaded travel agency apps in the world. Booking.com developed an in-house data streaming platform based on open-source Kafka. However, self-managing Kafka became increasingly burdensome as the company grew and introduced new use cases. Scaling clusters, handling updates and monitoring pipelines consumed significant resources. To alleviate the operational complexities of managing Kafka, Booking.com migrated each business unit's open-source clusters to Confluent platform.

其行動應用程式是全球下載次數最多的旅行社應用程式之一。Booking.com 開發了一個基於開源 Kafka 的內部資料流平台。然而，隨著公司的發展和新用例的引入，自我管理的 Kafka 變得越來越繁重。擴展叢集、處理更新和監控管道消耗了大量資源。為了減輕管理 Kafka 的操作複雜性，Booking.com 將各個業務部門的開源叢集遷移到了 Confluent 平台。

Our complete enterprise-grade solution provided enhanced reliability and out-of-the-box functionality. By spending less time managing infrastructure, Booking.com can now support various mission-critical use cases more efficiently, including marketing, payments, personalization and core booking processes. With a complete data streaming platform that is connected across their business, they were also able to deploy a connected trip experience. This allows customers to seamlessly book flights, accommodations, car rentals and experiences in one visit. The third driver of our resiliency is meeting customers wherever they are, whether that's on-prem, on the edge, in any cloud or hybrid environment.

我們完整的企業級解決方案提供了增強的可靠性和開箱即用的功能。透過減少管理基礎設施的時間，Booking.com 現在可以更有效率地支援各種關鍵任務案例，包括行銷、支付、個人化和核心預訂流程。透過一個貫穿整個業務的完整資料流平台，他們也能夠部署互聯的旅行體驗。這使得顧客可以在一次訪問中無縫預訂航班、住宿、汽車租賃和體驗。我們彈性的第三個驅動因素是無論客戶身在何處都能與他們會面，無論是在本地、在邊緣、在任何雲端還是在混合環境中。

That flexibility also provides another layer of resilience to our business. It means our growth strategy is less exposed to changes in cloud investment and provides a healthy mix of ratable and consumption-based revenue streams. Just as importantly, it lets us land and expand in environments where cloud is not an option, whether for regulatory reasons, company mandates or just customer preference. We continue to see strong momentum in this area. Our Confluent Platform business had a particularly strong quarter with revenue growth accelerating to 18% year over year, representing its strongest Q1 growth in three years.

這種靈活性也為我們的業務提供了另一層彈性。這意味著我們的成長策略不太受雲端投資變化的影響，並提供了可評估和基於消費的收入流的健康組合。同樣重要的是，它讓我們能夠在無法使用雲端的環境中著陸和擴展，無論是出於監管原因、公司要求還是只是客戶偏好。我們繼續看到該領域的強勁勢頭。我們的 Confluent 平台業務本季表現尤為強勁，營收年增 18%，創下三年來第一季最強勁的成長紀錄。

And finally, it's not just that our products are better, faster and more reliable, they're also more cost-effective. This is a strong differentiator that provides our customers with more value for less money across a wide range of use cases. Our low TCO enables us to expand usage within our existing customer base and also drive new conversions from open-source Kafka. It's one of the key levers that helps us retain customers, grow within our installed base and tap into the broader open-source community in a meaningful way. Key to capitalizing on this TCO advantage is offering pricing and packaging that fits full range of Kafka use cases from early projects to the most demanding production workloads.

最後，我們的產品不僅更好、更快、更可靠，而且更具成本效益。這是一個強大的差異化因素，它能夠在廣泛的使用案例中為我們的客戶提供更少的錢、更高的價值。我們的低 TCO 使我們能夠在現有客戶群中擴大使用範圍，並推動開源 Kafka 的新轉換。它是幫助我們留住客戶、擴大現有客戶群並以有意義的方式融入更廣泛的開源社群的關鍵槓桿之一。利用這項 TCO 優勢的關鍵是提供適合從早期專案到最苛刻的生產工作負載的所有 Kafka 用例的定價和包裝。

With new offerings like WarpStream and Freight Clusters, we're now able to serve high-throughput, low-latency workloads at attractive price points, enabling our customers to tackle a wider range of use cases. We continue to see strong traction with both offerings in Q1, including new customers like Lyftoff.io and the next wave of GenAI companies like Cursor and Thinking Machines. Here's an example of how our TCO advantage drives sustained growth with a top 20 global bank. This $5 million-plus ARR customer, who most recently increased their spend with us by over 30%, initially relied on open-source Kafka. However, the complexity and rising costs of self-managing Kafka quickly outweighed the value they receive.

借助 WarpStream 和 Freight Clusters 等新產品，我們現在能夠以極具吸引力的價格提供高吞吐量、低延遲的工作負載，使我們的客戶能夠應對更廣泛的用例。我們在第一季繼續看到這兩項產品強勁的成長勢頭，包括 Lyftoff.io 等新客戶以及 Cursor 和 Thinking Machines 等下一波 GenAI 公司。以下是我們的 TCO 優勢如何推動全球前 20 名銀行持續成長的一個例子。這位年收入超過 500 萬美元的客戶最近在我們這裡的支出增加了 30% 以上，最初依賴開源 Kafka。然而，自我管理的 Kafka 的複雜性和不斷上升的成本很快就超過了它們所獲得的價值。

More than five years ago, they migrated their first use case to Confluent. Since then, we've become a strategic partner as they've transitioned numerous legacy workloads to the cloud. Today, Confluent powers hundreds of use cases across their business like fraud detection, capital management, regulatory reporting of trade data and. By moving to Confluent, they have significantly reduced operational costs, turned their real-time data into a competitive advantage and lowered their TCO. In fact, the customer believes that for every dollar they spend with Confluent, they would otherwise spend $3 managing Kafka themselves.

五年多前，他們將第一個用例遷移到了 Confluent。從那時起，我們就成為了他們的策略夥伴，因為他們已經將大量遺留工作負載轉移到了雲端。如今，Confluent 為其業務中的數百個用例提供支持，例如欺詐檢測、資本管理、貿易數據的監管報告等。透過轉向 Confluent，他們顯著降低了營運成本，將即時數據轉化為競爭優勢並降低了 TCO。事實上，客戶認為他們在 Confluent 上花費的每一美元，他們自己都要花 3 美元來管理 Kafka。

Together, these four factors, mission-critical use cases, open-source conversion opportunities, hybrid business model and TCO advantages, have laid the foundation and made our business more resilient through multiple shifts in the macro environment. Additionally, we see continued adoption of our DSP components, which significantly outgrew our core cloud business. Confluent unifies everything organization speed to work with real-time data, the ability to Stream, Connect, Process and Govern continuously flowing streams of data all in one platform. This foundation is proving especially valuable as generative AI moves from experimentation to execution. In particular, we're seeing strong interest and adoption for Flink and Tableflow, two of the most recent additions to our DSP.

關鍵任務用例、開源轉換機會、混合業務模式和 TCO 優勢這四個因素共同奠定了基礎，使我們的業務在宏觀環境的多次轉變中更具彈性。此外，我們看到 DSP 元件的持續採用，其成長遠遠超過了我們的核心雲端業務。Confluent 將組織處理即時資料的速度、串流、連接、處理和管理連續流動的資料流的能力統一在一個平台上。隨著生成式人工智慧從實驗走向執行，這項基礎被證明是特別有價值的。特別是，我們看到了人們對 Flink 和 Tableflow 的濃厚興趣和採用，它們是我們 DSP 中最新添加的兩個功能。

Our complete platform is becoming the connective tissue that brings real-time context to our customers' GenAI applications so they can deliver trustworthy and actionable results that work in everyday operations. It's very exciting to see what our DSP enables our customers to do. For example, a leading luxury goods conglomerate with 75 brands and over 6,000 stores worldwide uses our Confluent Cloud for Apache Flink to power its real-time order management and drive e-commerce script. The company initially turned to Confluent Cloud to stream order management data to give internal teams accurate, real-time visibility into product availability. As Confluent proved its value the first use case, the customer consumes more of our platform, including our fully managed Flink service to prevent inaccurate stock information caused by duplicate data.

我們的完整平台正在成為為客戶的 GenAI 應用程式帶來即時背景的連接組織，以便他們能夠提供在日常營運中可信賴且可操作的結果。看到我們的 DSP 幫助客戶實現這些目標，我們感到非常興奮。例如，一家在全球擁有 75 個品牌和 6,000 多家門市的領先奢侈品集團使用我們的 Confluent Cloud for Apache Flink 來支援其實時訂單管理並推動電子商務腳本。該公司最初採用 Confluent Cloud 來傳輸訂單管理數據，以便內部團隊能夠準確、即時地了解產品可用性。由於 Confluent 在第一個用例中證明了其價值，客戶更多地使用了我們的平台，包括我們完全託管的 Flink 服務，以防止因重複資料而導致的庫存資訊不準確。

This customer uses Flink to filter out orders and inventory duplicates and to continuously analyze real-time product availability. When high demand items come back in stock, Flink automatically triggers real-time alerts to notify waiting customers, enhancing the customer experience and driving incremental revenue. Building on the success of their initial Flink use case, they are now exploring new ways to leverage the technology to streamline and scale product inventory management. Before closing, I'm excited to share two updates. Ryan Mac Ban has been promoted to Chief Revenue Officer at Confluent.

該客戶使用 Flink 過濾訂單和庫存重複項，並持續分析即時產品可用性。當高需求商品恢復庫存時，Flink 會自動觸發即時警報通知等待的客戶，從而增強客戶體驗並增加收入。在最初的 Flink 用例成功的基礎上，他們現在正在探索利用該技術簡化和擴展產品庫存管理的新方法。在結束之前，我很高興分享兩個更新。Ryan Mac Ban 已晉升為 Confluent 的首席營收長。

In this expanded role, Ryan will lead the global field strategy, bringing together sales, sales engineering, customer success and sales operations to help customers activate real-time data to build the next wave of intelligent applications. Ryan joined Confluent last year as Senior Vice President, Global Head of Sales. He brings over 20 years of sales leadership experience, building and leading top-performing teams around the world. Before Confluent, Ryan was President of UIPath Americas, where he drove significant growth across their multiproduct platform. He also held senior leadership roles at VMware and Cisco.

在這個擴展的職位上，Ryan 將領導全球現場策略，將銷售、銷售工程、客戶成功和銷售營運整合在一起，幫助客戶啟動即時數據，建立下一波智慧應用。Ryan 去年加入 Confluent，擔任資深副總裁兼全球銷售主管。他擁有超過 20 年的銷售領導經驗，在世界各地建立和領導表現最佳的團隊。在加入 Confluent 之前，Ryan 曾擔任 UIPath Americas 總裁，推動該公司多產品平台實現顯著成長。他也曾在 VMware 和 Cisco 擔任高階領導職務。

Second, we're honored to be named a Google Partner of the Year for the sixth time. This recognition is a reflection of the strong partnership we have with the leading CSPs. Together, we enable organizations to deliver the next-generation applications they need to thrive in the age of AI. In closing, we're proud of our strong start to the year. Our commitment to providing the industry's most complete data streaming platform, paired with a highly resilient business, uniquely positions Confluent to seize the $100 billion-plus data streaming market.

其次，我們很榮幸第六次被評為 Google 年度合作夥伴。這項認可體現了我們與領先的 CSP 之間的牢固合作關係。我們共同努力，幫助各組織提供在人工智慧時代蓬勃發展所需的下一代應用程式。最後，我們為今年的強勁開局感到自豪。我們致力於提供業界最完整的資料流平台，並結合高度彈性的業務，使 Confluent 擁有獨特的優勢，可以搶佔價值 1,000 億美元以上的資料流市場。

With that, I'll turn it over to Rohan.

說完這些，我就把話題交給羅漢 (Rohan)。

Thanks, Jay. Good afternoon, everyone, and thanks for joining our earnings call. Our first quarter performance underscores the strength of our mission-critical data streaming platform, the strategic value of our multi-cloud, multi-data destination and multi-deployment approach as well as flexibility of our well-diversified growth strategy. Turning to the results. Q1 subscription revenue grew 26% to $260.9 million, exceeding the high end of our guidance and representing 96% of total revenue.

謝謝，傑伊。大家下午好，感謝您參加我們的財報電話會議。我們第一季的業績凸顯了我們關鍵任務資料流平台的實力、多雲、多數據目的地和多部署方法的策略價值以及多元化成長策略的靈活性。談到結果。第一季訂閱營收成長 26%，達到 2.609 億美元，超過我們預期的高端，佔總營收的 96%。

Confluent Platform revenue reached a new record of $118.2 million with growth accelerating to 18%. This momentum was driven by early traction in our partner ecosystem where OEM showed particular strength internationally. Cloud revenue grew 34% to $142.7 million, representing 55% of subscription revenue. The revenue impact of lapping the leap year was approximately negative $1.6 million as Q1 this year had one fewer day for consumption.

Confluent 平台營收創下 1.182 億美元的新紀錄，成長率加速至 18%。這一勢頭是由我們合作夥伴生態系統的早期牽引力推動的，其中 OEM 在國際上表現出了特別的實力。雲端營收成長 34% 至 1.427 億美元，佔訂閱收入的 55%。由於今年第一季的消費天數減少了一天，閏年對營收的影響約為負 160 萬美元。

Turning to the geographical mix of total revenue. Revenue from the US grew 23% to $156.4 million. Revenue growth from outside the US grew 28% to $114.7 million.

轉向總收入的地理分佈。來自美國的營收成長23%，達到1.564億美元。美國以外地區的收入成長了 28%，達到 1.147 億美元。

Moving on to rest of the income statement. I'll be referring to non-GAAP results unless stated otherwise. While driving top line growth at scale, we continue to show significant operating leverage in our model. In Q1, subscription gross margin increased 100 basis points to 81.7% primarily driven by continued efficiency gain in Confluent Cloud, coupled with strength in Confluent Platform. Operating margin was 4.3%, exceeding our guidance of approximately 3% and was primarily driven by revenue and gross margin outperformance. Adjusted free cash flow margin, which excluded the nonrecurring impact of our compensation change in Q1, was 1.8%.

繼續討論損益表的其餘部分。除非另有說明，我將參考非 GAAP 結果。在大規模推動營收成長的同時，我們的模型中繼續顯示出顯著的經營槓桿。第一季度，訂閱毛利率增加 100 個基點至 81.7%，這主要得益於 Confluent Cloud 的持續效率提升以及 Confluent Platform 的強勁表​​現。營業利益率為 4.3%，超過我們約 3% 的預期，主要得益於收入和毛利率的優異表現。調整後的自由現金流利潤率為 1.8%，不包括第一季薪資變動的非經常性影響。

The impact of the change to free cash flow margin in Q1 was approximately 14 percentage-points. Net income per share was $0.08 using 367.8 million diluted weighted average shares outstanding. Fully diluted share count under the treasury stock method was approximately 380.9 million. We ended the first quarter with $1.92 billion in cash, cash equivalents and marketable securities. Turning now to customer metrics.

第一季自由現金流利潤率變動的影響約為 14 個百分點。以 3.678 億股稀釋加權平均流通股計算，每股淨收益為 0.08 美元。以庫存股法計算，完全稀釋股數約為 3.809 億股。第一季結束時，我們擁有 19.2 億美元的現金、現金等價物和有價證券。現在轉向客戶指標。

We ended Q1 with approximately 6,140 customers, representing a sequential increase of 340 customers, our highest sequential increase since Q1 of 2022. This robust growth underscores the strong value proposition of our multiproduct platform and our continued success in capturing the open-source conversion opportunity. As mentioned at Investor Day 2025, we will report total customer count on an annual basis and begin reporting $20,000-plus ARR customers on a quarterly basis. In Q1, $20,000-plus ARR customer count increased to 2,487, up 41 customers sequentially and represented more than 95% of our ARR. Our $100,000-plus ARR customer count increased to 1,412, up 31 customers sequentially and accounted for greater than 90% of our ARR.

截至第一季度，我們擁有約 6,140 名客戶，季增 340 名客戶，這是自 2022 年第一季以來的最高環比增幅。這種強勁的成長凸顯了我們多產品平台的強大價值主張以及我們在抓住開源轉換機會方面的持續成功。正如 2025 年投資者日所提到的，我們將按年報告客戶總數，並開始按季度報告 ARR 超過 20,000 美元的客戶。第一季度，年經常性收入 (ARR) 超過 20,000 美元的客戶數量增加到 2,487 人，比上一季度增加了 41 人，占我們 ARR 的 95% 以上。我們的 10 萬美元以上 ARR 客戶數量增加到 1,412 人，比上一季增加了 31 人，占我們 ARR 的 90% 以上。

Our $1 million-plus ARR customer count grew to 210, up 16 customers sequentially, our best quarter in net add for this cohort. NRR for the quarter remained stable at 117%, while GRR continued to be greater than 90%, demonstrating the mission-critical nature of our data streaming platform for our customers.

我們的 100 萬美元以上 ARR 客戶數量成長至 210 家，季增 16 家，這是我們本季淨增額最高的一個季度。本季的 NRR 保持穩定在 117%，而 GRR 繼續超過 90%，這表明我們的資料流平台對客戶而言具有關鍵任務性質。

Turning to our outlook. For the fiscal second quarter of 2025, we expect subscription revenue to be in the range of $267 million to $268 million, representing growth of approximately 19%, non-GAAP operating margin to be approximately 5% and non-GAAP income per diluted share to be in the range of $0.08 to $0.09. For fiscal year 2025, we expect subscription revenue to be in the range of $1.1 billion to $1.11 billion, representing growth of approximately 19% to 20%, non-GAAP operating margin to be approximately 6% and non-GAAP net income per diluted share to be approximately $0.36.

轉向我們的展望。對於 2025 財年第二季度，我們預計訂閱收入將在 2.67 億美元至 2.68 億美元之間，成長約 19%，非 GAAP 營業利潤率約為 5%，非 GAAP 每股攤薄收益將在 0.08 美元至 0.09 美元之間。對於 2025 財年，我們預計訂閱收入將在 11 億美元至 11.1 億美元之間，成長約 19% 至 20%，非 GAAP 營業利潤率約為 6%，非 GAAP 每股攤薄淨收益約為 0.36 美元。

Now I'd like to provide some additional context for our guidance, along with a few modeling points. In light of the uncertainties in the current environment, we are widening our revenue guidance range and embedding a modest decline in growth rates from Q2 through Q4. For our cloud business, some of our larger customers began slowing the pace of new use case addition and focusing on cost optimization efforts in March. In contrast, consumption activities in our smaller customer base remains stable. These trends have continued into April.

現在我想為我們的指導提供一些額外的背景信息，以及一些建模點。鑑於當前環境的不確定性，我們擴大了收入預期範圍，並預計第二季至第四季的成長率將略有下降。對於我們的雲端業務，我們的一些較大的客戶從 3 月開始放慢新用例添加的速度，並專注於成本優化工作。相比之下，我們較小客戶群的消費活動保持穩定。這些趨勢一直持續到四月。

Given the current macro conditions, we believe it's prudent to assume there will not be a near-term rebound in consumption. This contrasts with previous cycles where we typically saw a subsequent expansion following a period of slower consumption. As a result, we now expect cloud subscription revenue mix to be approximately 58% for Q4 '25 with a sequential mix increase of approximately 1 point each quarter. Finally, we expect fiscal year '25 adjusted free cash flow margin to be approximately 6%. This excludes, as previously discussed, a onetime headwind of 3 percentage-points to 4 percentage-points for the full year due to our compensation change in Q1.

鑑於當前的宏觀條件，我們認為謹慎的假設是消費短期內不會反彈。這與先前的周期形成了鮮明對比，在先前的周期中，我們通常會看到在一段消費放緩之後出現隨後的擴張。因此，我們現在預計 2025 年第四季雲端訂閱收入組合約為 58%，且每季將成長約 1 個百分點。最後，我們預期 25 財年調整後的自由現金流利潤率約為 6%。如前所述，這不包括由於我們第一季的薪資變化而導致的全年 3 個百分點至 4 個百分點的一次性不利因素。

We are confident that our updated guidance, coupled with multiple parts to growth, sets us up for success this year. Specifically, we see four key drivers of growth in our business. First, the core streaming conversion opportunity targeting an open-source installed base of over 150,000 organizations; second, our DSP upsell opportunities across Connect, Process, Govern and Tableflow; third, our highly strategic role in the age of AI. As traditional enterprises accelerate AI adoption, we believe this represents our largest monetization opportunity within AI; and fourth, the leverage we gain from our expanding partner ecosystem, including the OEMs, SIs, MSPs and other high-impact strategic partnerships that help us extend our global reach and accelerate go-to-market efforts. It's important to note, these aren't just cloud-only growth vectors.

我們相信，我們最新的指導方針，加上多個成長部分，將為我們今年的成功奠定基礎。具體來說，我們認為業務成長有四個關鍵驅動力。首先，核心串流媒體轉換機會針對超過 15 萬個組織的開源安裝基礎；第二，我們的 DSP 在 Connect、Process、Govern 和 Tableflow 方面有追加銷售機會；第三，我們在人工智慧時代具有高度策略性的地位。隨著傳統企業加速採用人工智慧，我們相信這是我們在人工智慧領域最大的獲利機會；第四，我們從不斷擴大的合作夥伴生態系統中獲得的優勢，包括 OEM、SI、MSP 和其他高影響力的策略合作夥伴關係，這些有助於我們擴大全球影響力並加快上市步伐。值得注意的是，這些不僅僅是雲端運算的成長載體。

We believe they represent durable growth drivers for both Confluent Cloud and Confluent Platform. For example, in Q1, revenue outperformance was driven by on-prem momentum with Confluent Platform and its critical role in closing large enterprise deals with our OEM partners. Our well-diversified growth strategy gives us resilience and flexibility, enabling us to continue to drive durable and profitable growth. In closing, our first quarter results are a testament to the resilience of our business and our ability to capture our market opportunity. With a large TAM, category-defining technology and an exceptional team, we are firmly positioned to sustain long-term growth and profitability.

我們相信它們代表了 Confluent Cloud 和 Confluent Platform 的持久成長動力。例如，在第一季度，收入的出色表現得益於 Confluent Platform 的內部部署勢頭及其在與我們的 OEM 合作夥伴達成大型企業交易中發揮的關鍵作用。我們多元化的成長策略賦予我們韌性和靈活性，使我們能夠繼續推動持久且獲利的成長。最後，我們的第一季業績證明了我們業務的韌性和抓住市場機會的能力。憑藉龐大的 TAM、定義類別的技術和卓越的團隊，我們完全有能力維持長期成長和獲利能力。

Now Jay and I will take your questions.

現在我和傑伊將回答大家的提問。

(Event Instructions)

（活動須知）

Pinjalim Bora, JPMorgan followed by RBC.

Pinjalim Bora，摩根大通，其次是加拿大皇家銀行。

One, obviously, back top of mind for everybody, I think you said you've seen some slowdown in new use cases across large customers. Wanted to ask you, what about existing kind of use cases, the consumption run rate there? Because I would assume that the existing use cases probably drives a bigger portion of revenue than kind of the ramping of new use cases, right? And also, from the ACV commitment standpoint, are you seeing any difference so far in April, which makes you to kind of look at the guidance?

首先，顯然，大家最關心的是，我想您說過，您已經看到大客戶的新用例有所放緩。想問您，現有的用例類型如何，那裡的消費運行率如何？因為我認為現有的用例可能會比新用例的增加帶來更大的收入份額，對嗎？此外，從 ACV 承諾的角度來看，到 4 月為止您是否看到任何差異，讓您開始專注於指導？

Yeah. Great question. So overall, what we said is that we did see a bit of lower consumption in our larger customers on the cloud. So overall, good quarter, strong growth in CP, solid growth in cloud. This was a combination of optimizations and slower addition of new use cases, I would say both. That wasn't across the customer base. In our smaller customers, we didn't see that pattern. And it wasn't specific to a single segment, either in terms of international distribution or industry. I wouldn't attribute that directly to macro.

是的。好問題。總的來說，我們確實看到雲端大客戶的消費有所下降。整體而言，本季業績表現良好，CP 成長強勁，雲端運算成長穩健。這是優化和緩慢添加新用例的結合，我想說兩者都有。這並不是針對整個客戶群的。在我們的較小客戶中，我們沒有看到這種模式。而且它並不是針對某一特定領域，無論是國際分銷還是產業。我不會將其直接歸因於宏觀。

I think the way I would explain it is this, right? We've seen a bit of an oscillating pattern in these larger customers, where they tend to have kind of a cycle of optimization and then additional growth and optimization and growth. In times that are a bit more certain, we would probably bake in the assumption that, that pattern would continue and we'd see more of a rebound. Given the current environment, we don't want to do that. We want to be a little bit more conservative in how we're looking at that and just kind of play it forward straight through the year. And that's what we've done when we've constructed the guidance for this year. Rohan, you may have more to add to that.

我想我應該這樣解釋，對嗎？我們在這些較大的客戶中看到了一種波動模式，他們傾向於經歷一種最佳化週期，然後是額外的成長、最佳化和成長。在更確定的時期，我們可能會假設這種模式將會持續下去，並且我們會看到更多的反彈。鑑於當前的環境，我們不想這樣做。我們希望以更保守的態度看待這個問題，並在今年繼續前進。這就是我們在製定今年的指導方針時所做的。Rohan，你可能還有更多要補充的。

No. I think you covered it, Jay. Pinjalim, I'll take the second part of your question. On the commitments, as you can see from our RPO numbers, Q1 was strong, both on the CP side where we saw multiyear deals, customers committing multiple material deals; also on the cloud side, where our cloud motion continues to be consumption-first, but we've seen a lot of customers actually come and commit to multiyear deals on the cloud side as well. So the ACV commitments in Q1 was strong. That's how I'd categorize it.

不。我想你已經講清楚了，傑伊。Pinjalim，我來回答你問題的第二部分。在承諾方面，正如您從我們的 RPO 數據中看到的那樣，第一季度表現強勁，在 CP 方面，我們看到了多年期交易，客戶承諾了多項重大交易；同樣在雲端，我們的雲端運算動向仍然是以消費為先，但我們已經看到很多客戶實際上也來這裡並承諾在雲端進行多年的交易。因此第一季的 ACV 承諾很強勁。這就是我對其的分類。

Yeah. Got it. Jay, one follow-up for you and just more of a high level. We have seen some enthusiasm in the open-source community around this idea of a discless Kafka. It seems like with the new IP around there. Obviously, that's writing directly to object storage, and WarpStream has been doing that for some time now in your own WarpStream. So I want to ask you, how do you kind of see the future of Confluent? Is Confluent overall as a whole, not just WarpStream, could it move to a discless future at some point? Or the latency issues kind of with object storage kind of might --

是的。知道了。傑伊，還有一件事想跟進，那就是更高層次的問題。我們看到開源社群對無盤 Kafka 這個想法表現出一定的熱情。好像那裡有新的 IP。顯然，這是直接寫入物件存儲，而 WarpStream 已經在您自己的 WarpStream 中執行此操作一段時間了。所以我想問您，您如何看待 Confluent 的未來？Confluent 作為一個整體（不僅僅是 WarpStream）是否能在某個時候走向無盤未來？或者物件儲存的延遲問題可能--

Yeah. We've actually -- it's a great question. Yes, we've actually done that in both our cloud and with WarpStream. We're actually very intelligent about the use of storage. And so our Freight Clusters use kind of a pure discless approach. Our other clusters use a combination of storage technologies across the memory hierarchy to optimize it. There's a fair amount of sophistication layer, and it actually matters a lot when you think about the cost characteristics and the kind of performance. It's definitely an area we put a fair amount of R&D effort.

是的。我們實際上——這是一個很好的問題。是的，我們實際上已經在雲端和 WarpStream 中做到了這一點。事實上，我們對儲存的使用非常明智。因此，我們的貨運集群採用一種純無盤方法。我們的其他叢集使用跨記憶體層次結構的儲存技術組合來對其進行最佳化。有相當多的複雜層，當你考慮成本特性和性能類型時，它實際上非常重要。這絕對是我們投入大量研發精力的領域。

Matt Hedberg, RBC followed by Wolfe.

RBC 的 Matt Hedberg 和 Wolfe 緊隨其後。

Congrats on the quarter environment at all. Jay, it was great to hear the success in DSP. I think you called out Flink and Tableflow in particular. I'm just sort of curious, I think you have 4Q UP was 13% of your cloud business. I'm wondering if you can sort of quantify how that improved sequentially. And then maybe just a little bit more on Flink, it seems to be coming up more and more in our partner checks. I'm wondering where you're seeing some of the option there.

對本季的環境表示祝賀。傑伊，很高興聽到 DSP 取得成功。我認為您特別提到了 Flink 和 Tableflow。我只是有點好奇，我認為你們第四季的雲端業務成長了 13%。我想知道您是否可以量化其連續改進的情況。然後也許對 Flink 再多說一點，它似乎在我們的合作夥伴檢查中出現得越來越多。我想知道您在哪裡看到了一些選項。

Yeah. I'll give you some characterization. Our intention with that kind of DSP percentage is to come back to that periodically but not every single quarter. And so here's what we're seeing. So ton of enthusiasm. Overall like sentiment in the company is the DSP offerings are on great success trajectory, significantly outgrowing the kind of core cloud business. And we've been through this before. As that stuff starts to get to scale, it moves the overall numbers. And a lot of enthusiasm from customers. We just released Tableflow as GA in AWS.

是的。我將給你一些特徵描述。我們採用這種 DSP 百分比的目的是定期（但不是每個季度）恢復到這個水平。這就是我們所看到的。如此充滿熱情。整體而言，公司內部的情緒是 DSP 產品正處於巨大的成功軌跡，其成長速度遠遠超過了核心雲端業務。我們以前也經歷過這樣的事。隨著這些東西開始規模化，它就會改變整體數字。顧客們也表現出了極大的熱情。我們剛剛在 AWS 上發布了 Tableflow 作為 GA 版本。

Great early signs for that. And then Flink is succeeding really nicely both in the cloud as well as in CP. We've added this to our CP offering and picked up some really sizable customers on that doing awesome use cases. And so I think this is still the early innings for the offerings on both sides but great initial success. In the cloud, these kind of serverless offerings, you start with kind of a broad base of simple use cases, then you start to build to the more mission-critical, large conversions where people are moving dozens or hundreds of different processing jobs over.

這是非常好的早期跡象。Flink 在雲端和 CP 方面都取得了巨大的成功。我們已將其添加到我們的 CP 產品中，並吸引了一些真正規模可觀的客戶，並實現了出色的用例。因此，我認為這仍是雙方努力的早期階段，但已取得巨大的初步成功。在雲端，這些無伺服器產品，您可以從廣泛的簡單用例開始，然後開始建立更關鍵的任務，大型轉換，人們在其中移動數十或數百個不同的處理作業。

And that's, of course, when you start to see the -- and we're starting to win some of those kind of larger customer engagements. We're putting in a ton of work to make sure we can really deliver against these big migrations, and that's a real focus in the team. But overall, do you think that there's a significant push inside companies towards real time in the use of data? We're seeing this pattern around, we would call it shifts left where a lot of what customers were doing with data is moving upstream into the real-time area so that it can go to multiple destinations, so that it can land in their analytics environments faster and so that it can be part of the use of AI applications, which often need that. And for all those reasons, I think we have great tailwinds in that area.

當然，當你開始看到——我們開始贏得一些更大的客戶參與時。我們正在投入大量工作以確保我們能夠真正實現這些大規模遷移，這也是團隊真正關注的重點。但整體而言，您是否認為公司內部在即時使用數據方面有顯著的推動力？我們看到這種模式，我們稱之為左移，其中客戶對資料的大量操作正在向上游移動到即時區域，以便它可以到達多個目的地，以便它可以更快地進入他們的分析環境，並且可以成為人工智慧應用程式使用的一部分，而人工智慧應用程式通常需要這樣做。出於所有這些原因，我認為我們在該領域擁有巨大的順風優勢。

We ultimately don't make a ton of that product success into our guidance. It's often hard to forecast these new things, something like tableflow that's just coming out to GA right now. We would have internal goals, but when we think about our financial plan for the year, we wouldn't have something attached to that until we really see it perform. But the initial signs are very promising in terms of build the pipeline and customer conversations.

我們最終不會將大量產品的成功納入我們的指導範圍。通常很難預測這些新事物，例如現在剛進入 GA 的 tableflow。我們會有內部目標，但是當我們考慮年度財務計劃時，我們不會對其附加任何內容，直到我們真正看到它發揮作用。但就建立管道和客戶對話而言，初步跡象非常有希望。

That's fantastic. And then maybe just one for Rohan. I think the reduced guidance makes a lot of sense to us just given the macro. I appreciate the conservatism there. Coming off of a record quarter, at least from the past several years of new customer adds, I'm wondering how you're thinking about the relative rate of customer adds versus NRR. I mean, are you assuming that, that NRR comes down a little bit, new customer adds? Just kind of the components of that would be helpful.

太棒了。然後也許只給羅翰一個。我認為，從宏觀角度來看，降低指導對我們來說很有意義。我欣賞那裡的保守主義。在經歷了創紀錄的一個季度之後，至少從過去幾年新增客戶的數量來看，我想知道您如何看待客戶增加率與 NRR 的相對比率。我的意思是，您是否假設，NRR 會下降一點，新客戶會增加？其中的一些組成部分將會很有幫助。

Yeah. Matt, sure thing. I mean I'll make a couple of comments on guidance, and then I'll get to your question on NRR and customer count. I mean, as we called out, our guidance philosophy has been very consistent. And how we've approached it is disciplined approach and set prudent targets.

是的。馬特，當然可以。我的意思是我會對指導發表一些評論，然後我會回答你關於 NRR 和客戶數量的問題。我的意思是，正如我們所呼籲的，我們的指導理念一直非常一致。我們採取的是嚴謹的方法並設定審慎的目標。

And Jay talked about the dynamic around consumption. On the CP side, we did see good momentum in Q1. Our visibility for Q2 remains good. So that's kind of driving our outlook for Q2 and rest of the year. But if I take a step back and look at the guidance, what I'll tell you is we're looking at a guidance of 19% to 20% for the year at $1 billion-plus scale.

傑伊談到了消費方面的動態。在 CP 方面，我們確實看到了第一季的良好勢頭。我們對第二季的可見度仍然良好。這在某種程度上影響了我們對第二季及今年剩餘時間的展望。但如果我退一步來看看指導方針，我會告訴你，我們預計今年的指導方針是 19% 到 20%，規模超過 10 億美元。

And we're doing that with what I would call a fairly prudent setup for us. And we have multiple vectors of growth as we look at it, which includes streaming, which includes DSP, the AI as well as the partner ecosystem. So that's a little bit about the guidance and why we feel good with the setup that we have right now. To your question around customer count, that's right. I mean that entire ecosystem of customer count, not only total customers on one bookend, but also $1 million-plus customers, we had the best quarter from $1 million-plus customers, that tells you that top of the funnel, obviously, we are adding a lot of customers into the Confluent ecosystem.

我們正在採取我認為對我們來說相當審慎的措施來做到這一點。我們看到有多個成長方向，包括串流媒體、DSP、AI 以及合作夥伴生態系統。這就是有關指導的一點信息，以及為什麼我們對目前的設置感到滿意。對於您關於客戶數量的問題，答案是肯定的。我的意思是整個生態系統的客戶數量，不僅是書擋上的總客戶數，還有 100 萬美元以上的客戶，我們擁有來自 100 萬美元以上客戶的最佳季度，這告訴你，在漏斗的頂部，顯然，我們正在向 Confluent 生態系統中添加大量客戶。

Our $20,000 customer tell you the amount of, I would say, high-intensity, high-frequency customers who can spend a lot with us. That shows the core customer count. And $1 million-plus customers is just how we are pushing them through the funnel. So overall, we feel good. NRR last couple of quarters has been, give or take, around 117% range.

我們的 20,000 美元客戶會告訴你，我想說，高強度、高頻率的客戶可以在我們這裡花很多錢。這顯示了核心客戶數量。而價值 100 萬美元以上的客戶正是我們推動他們進入銷售管道的方式。所以總體來說，我們感覺很好。過去幾季的 NRR 大約在 117% 左右。

And what supports that NRR is our consistency on the GRR side. That's call-out number one. We've been greater than 90% since we've been a public company. And on the expansion side, the dynamics that we called out on the consumption side could have an impact. But if I think about NRR, it should be in that ZIP code of 117% as we look ahead.

支持 NRR 的是我們在 GRR 方面的一致性。這是第一個呼叫。自我們成為上市公司以來，我們的盈利率已超過 90%。在擴張方面，我們在消費方面所呼籲的動態可能會產生影響。但如果我考慮 NRR，展望未來，它應該位於 117% 的郵遞區號中。

Michael Turrin, Wells Fargo follow by Morgan Stanley.

圖林 (Michael Turrin)、富國銀行 (Wells Fargo) 和摩根士丹利 (Morgan Stanley)。

Jay, I want to start with something that was, I think, very topical at the Investor Day, which was just some of the AI-related use cases. You've mentioned Cursor a couple of times. Investor sentiment on this topic, certainly, it seems to ebb and flow. But can you just help add some context around what you're seeing in terms of AI-related demand, if there are certain use cases you're finding you'll have strong product market fit as those evolve and just how we should start to expect some of this to evolve over a longer period of time?

傑伊，我想從投資者日上非常熱門的話題開始，也就是一些與人工智慧相關的用例。您曾多次提到 Cursor。投資人對這個話題的看法似乎確實時好時壞。但是，您能否幫助我們補充一些關於您所看到的與人工智慧相關的需求的背景信息，如果您發現某些用例，那麼隨著這些用例的發展，您將擁有強大的產品市場契合度，以及我們應該如何開始預期其中一些用例在更長的時間內會發展？

Yeah, absolutely. So we've seen success kind of in two dimensions in the AI. I mean, first, selling to AI companies like Cursor, as you mentioned. The other is the kind of enterprise uses of AI, where we're really fulfilling two roles. One is this data supply chain for getting the data that fulfills the context for some of these AI applications.

是的，絕對是如此。因此，我們看到了人工智慧在兩個維度上的成功。我的意思是，首先，正如您所提到的，向 Cursor 這樣的人工智慧公司銷售。另一種是人工智慧的企業用途，我們實際上扮演著兩個角色。一個是數據供應鏈，用於獲取滿足某些人工智慧應用環境的數據。

The second is for these agents that are actually acting on things in real time. The usage -- the more sophisticated use of data has often been some batch job at the end of the day. But the point of these agents is to get into the kind of real-time flow of things. And we are starting to see those use cases emerge. I would say that's earlier of the two.

第二種是針對那些實際上是即時對事物採取行動的代理。用途－資料的更複雜用途通常是在一天結束時進行一些批次作業。但這些代理人的目的在於進入事物的即時流動。我們開始看到這些用例的出現。我想說，這是兩者中較早的一個。

But I think it's a pattern that's quite durable. We've done a fair amount of writing and kind of thought leadership in this area. And we're seeing kind of a community of interest forming around this in our customer base and well beyond in the larger community. So I think that's very promising.

但我認為這是一個相當持久的模式。我們在這個領域做了大量的寫作工作並發揮了思想領導作用。我們看到在我們的客戶群中以及在更大的社區中圍繞這一點形成了一種利益共同體。所以我認為這是非常有前景的。

Just as a follow-up, if I may. Just beyond the consumption patterns in terms of some of the things you're contemplating with guidance, it sounds like newer products are just lower assumptions in the earlier stage. But are there certain milestones we should be thinking about in terms of go-live with some of the newer products? And then on the go-to-market transition, maybe just help us with what you've seen from Ryan and how you're thinking about the shape of productivity you could see throughout the year.

如果可以的話，這只是一個後續問題。除了您在指導中考慮的一些事情的消費模式之外，聽起來新產品只是早期階段的較低假設。但是，在某些新產品的上線方面，我們是否應該考慮某些里程碑？然後，在市場轉型方面，也許您可以幫助我們了解 Ryan 的看法，以及您對全年生產力形態的看法。

Yeah. So I'm really excited to be working with Ryan. Any of these executive roles, we obviously do very broad outside search and look inside as well and met a lot of people in the process. And I felt internally, we had somebody who had a ton of strength in execution.

是的。所以我很高興能與瑞安合作。對於任何這些高階主管職位，我們顯然都會進行非常廣泛的外部搜索，也會進行內部調查，並在過程中結識了很多人。我內心裡感覺到，我們有一個在執行上非常強大的人。

He was a strategic thinker and understood what we were doing as a business, the space we were operating in. He would be an awesome part of the executive team and leadership of the company. So I feel really good about the changes. Obviously, also less disruptive because you have somebody who's already managing the full set of sales reps, just kind of taking on additional responsibility. And you have somebody who kind of proven themselves on the field versus taking a shot on somebody new coming in and adjusting to our space.

他是一位策略思想家，了解我們作為一家企業所做的事情以及我們所經營的領域。他將成為公司執行團隊和領導階層中出色的一員。所以我對這些改變感到非常滿意。顯然，這也不太會造成乾擾，因為你已經有人管理所有銷售代表，只是承擔了額外的責任。而且，你擁有的是一個已經在球場上證明過自己的球員，而不是一個剛進入球場並適應我們場地的新人。

So I feel really good about that and excited about what we can do in the year ahead. Obviously, less changes overall on the go-to-market side this year than last year as last year, we were doing this bigger set of consumption changes. So I think overall, that's a positive setup for us. And then remind me of the other part of the question.

因此我對此感到非常高興，並對我們在來年可以做的事情感到興奮。顯然，今年在市場進入方面的整體變化比去年要少，因為去年我們進行了更大範圍的消費變化。所以我認為總的來說，這對我們來說是一個積極的安排。然後提醒我問題的另一部分。

Just on some of the new products. So are there time lines we should be thinking about where those could start to see a steeper adoption ramp just based on the --

僅介紹一些新產品。那麼，我們是否應該考慮一下，什麼時候這些技術可以開始出現更快速的採用率？--

Yeah. The two things that I think are in our mind, we're -- we've started to get this Connect business scale, the business around the connectors. We're now able to do much more of the kind of lift and shift of existing use of open-source Kafka Connect and bring that into our platform. We feel pretty good about that motion.

是的。我認為我們考慮的兩件事是——我們已經開始擴大這個 Connect 業務規模，即圍繞連接器的業務。現在，我們能夠對現有的開源 Kafka Connect 進行更多的提升和轉變，並將其引入我們的平台。我們對該動議感到非常滿意。

So now it's kind of taken that to scale. For Flink, it really is about these larger customers and use cases. Taking the -- any kind of cloud layer that does data processing is basically a programming framework where you're writing effectively complete program, SQL or Java, and the ability for customers to take things that they've done to migrate in and really get to scale with that across big complicated production use cases. As we start to make more of those successful more quickly, that's kind of the motion to start bringing more and more onto the cloud offering there. CP Flink, I think, is of a nice start with customers.

所以現在它已經達到一定規模了。對於 Flink 來說，它確實與這些更大的客戶和用例有關。採取－任何進行資料處理的雲層基本上都是一個程式框架，您可以在其中有效地編寫完整的程式、SQL 或 Java，並且客戶能夠將他們已經完成的工作遷移進來，並真正實現跨大型複雜生產用例的擴展。隨著我們開始更快地取得更多成功，這是一種開始將越來越多的內容引入雲端的動力。我認為 CP Flink 對於客戶來說是一個好的開始。

That's kind of an easier process of lift and shift. And then Tableflow is just out of the starting gates. So the big milestones there are getting it across all the clouds, opening up the integration with Delta and Unity, the Databricks technologies that we've kind of promised to our customers and are working with Databricks around. Both of those are exciting milestones. I think we'll start to see that get out to early customers and have a better sense of the trajectory in the coming quarters.

這是一種更簡單的提升和轉變過程。然後 Tableflow 就剛起步。因此，最大的里程碑是將其擴展到所有雲端，開放與 Delta 和 Unity 的集成，這些是 Databricks 技術，我們已經向客戶承諾過，並且正在與 Databricks 合作。這兩個都是令人興奮的里程碑。我認為我們將開始看到早期客戶對此有所了解，並能更了解未來幾季的發展軌跡。

Sanjit Singh, Morgan Stanley.

摩根士丹利的 Sanjit Singh。

As we sort of enter this period of uncertainty, it sounds like the business coming in is quite healthy, and really impressive customer add metrics across the board. One of the things I'm trying to get a handle or we're all trying to get a handle on is that if optimization activity comes back online, how to compare it maybe do what we saw in the back half of 2022, throughout 2023, when there was kind of a more widespread optimization activity across the customer base. So in that spirit, Jay, I was wondering if you could talk about maybe the profile of the business back then in the second half of '22 and '23 versus what's driving growth in your customer base today. And if we did see another round of optimization activity, how would you compare it versus the last cycle?

當我們進入這個不確定的時期時，聽起來業務進展相當健康，而且各個方面的客戶增加指標確實令人印象深刻。我正在嘗試掌握或我們都在嘗試掌握的事情之一是，如果優化活動重新上線，如何將其與我們在 2022 年下半年和整個 2023 年看到的情況進行比較，當時在整個客戶群中開展了更廣泛的優化活動。因此，傑伊，本著這種精神，我想知道您是否可以談談 2022 年下半年和 2023 年的業務狀況以及當前推動客戶群成長的因素。如果我們確實看到另一輪優化活動，您會如何將其與上一輪進行比較？

Yeah. I would say that probably the biggest change in the business between those two time frames is I do think customers have done a lot of this. When you think about effectively what we went through in '19, '20, '21, there's kind of a run-up in cloud spend and a big push on just build, build, build, new stuff, new stuff, new stuff. And in that environment, you do get a lot of effectively unoptimized cloud usage in a variety of ways. And probably the companies with a consumption model as we have for our cloud saw the runup of that the fastest.

是的。我想說，這兩個時間段內業務發生的最大變化可能是，我確實認為客戶已經做了很多這樣的事情。當你認真思考我們在 2019 年、2020 年和 21 年所經歷的事情時，你會發現雲端運算支出增加，我們大力推動新事物的建置、建置、建置、新事物。在這種環境下，您確實會以各種方式獲得大量未優化的雲端使用。那些擁有像我們這樣的雲端運算消費模式的公司可能最快地見證了這一成長。

And then, of course, also in that consumption model, it's easier then for -- the customers should be able to make sure they're really getting the full value out of it, and that will take down the spend. I think the difference now is a lot of that has been done, right? So we do see refinement in customer usage. But I don't think that there's nearly as much of a -- I think there's a lot more tightness in the base of the cloud usage than there was at that time period.

當然，在這種消費模式下，客戶應該能夠確保他們真正從中獲得全部價值，這將減少支出。我認為現在的不同之處在於很多事情已經完成，對嗎？因此我們確實看到客戶使用情況有所改善。但我並不認為現在的情況與那個時期相比有太大差別——我認為現在雲端使用的基礎比那個時期更緊密。

So maybe the range of outcomes is maybe narrower than it was a couple --

因此，結果的範圍可能比一對夫妻的結果範圍要窄--

I think that's probably right.

我認為這可能是對的。

Yeah. And so picking up on the strong Confluent Platform performance, the acceleration of 18%, I was wondering if you sort of view that as a structural change as some of the knock-on effects from some of the policies -- the macro policies that have -- the policy debate around customers toggling a little bit more to self-managed given potential actions on digital services, those types of things. And so do you see Confluent Platform and some of the innovation around CP Flink now more of a better grower for the product portfolio, in your view?

是的。因此，鑑於 Confluent 平台的強勁表現，即 18% 的加速，我想知道您是否認為這是一種結構性變化，是某些政策（宏觀政策）產生的連鎖反應，圍繞客戶轉向自我管理的政策辯論，考慮到數位服務方面的潛在行動，諸如此類的事情。那麼，您認為 Confluent Platform 和 CP Flink 周圍的一些創新現在是否會對產品組合產生更好的促進作用？

Yeah. I think it's too early to say to make that. What we saw in Q1 was actually just really good performance in CP. I don't think -- those deals obviously have some time line to build. So it wasn't directly in relation to tariffs or anything like that, as you that quickly.

是的。我認為現在下結論還為時過早。我們在第一季看到的實際上是 CP 的良好表現。我不認為——這些交易顯然需要一定的時間才能達成。所以它與關稅或類似的東西沒有直接關係，正如你所說。

So what we have seen, I think, is probably a ton of exuberance about cloud-heavy investment and I would say maybe a more measured cloud and on-premise strategy in some of the larger enterprises. And what you're seeing is relevant, right, which is, hey, if companies are internationally are more concerned about the interaction with the US, could there be more pressure on cloud usage? It's certainly possible. I mean I don't see any kind of major pullback from the cloud, but it's certainly possible. The one thing I would say is Confluent has intentionally built to be highly resilient across that. And so to some extent, we're not taking a hard bet on one side or the other.

因此，我認為，我們看到的可能是人們對雲端運算密集型投資的極大熱情，我想說，一些大型企業可能正在採取更謹慎的雲端運算和內部部署策略。您所看到的是相關的，對吧，那就是，嘿，如果國際上的公司更加關注與美國的互動，那麼雲端使用會面臨更大的壓力嗎？這當然是有可能的。我的意思是，我沒有看到雲端運算出現任何重大回調，但這肯定是可能的。我想說的是，Confluent 有意將其建構得具有高度的彈性。因此從某種程度上來說，我們不會偏向任何一方。

We've continued to invest in Confluent Platform. We've continued to add functionality. There wasn't as much growth in data center usage for a period of time. If there's a bit more, CP will probably grow a bit faster. If that's not the case, cloud will grow faster.

我們一直在對 Confluent 平台進行投資。我們不斷添加功能。在一段時間內，資料中心的使用量並沒有太大的成長。如果再多一點，CP 可能會成長得更快一些。如果情況並非如此，雲端運算將會成長得更快。

We're basically happy to support customers wherever they are. I don't -- I wouldn't expect a large scale pullback from the cloud just because we've, as an industry, kind of come to just rely so heavily on it. So I think it'd be premature to call any of that. But the key point for us, making sure we're in a great place with customers one way or the other.

無論客戶身在何處，我們都樂意為他們提供支援。我不認為雲端運算會出現大規模的撤退，因為我們作為一個產業已經變得非常依賴雲端運算。所以我認為現在做出任何這樣的判斷都為時過早。但對我們來說，關鍵在於確保我們能以任何方式與客戶保持良好的關係。

And I will tell you that this ability to serve customers multi-cloud and hybrid has become really top of mind for a lot of these companies as they are just starting to think about the balance of investment, what they're doing, what's happening with some of the legacy stacks, et cetera, I think that's turned into a real asset and with it. Maybe four years ago, it was more hypothetical. Now it's just a very concrete must-have.

我要告訴你們，這種為客戶提供多雲和混合雲服務的能力已經成為許多公司最關心的問題，因為他們才剛開始考慮投資平衡、他們在做什麼、一些遺留堆疊發生了什麼等等，我認為這已經成為一項真正的資產。也許四年前，這還只是假設而已。現在它只是一個非常具體的必需品。

Sanjit, I'll add another, I would say, slightly different lens to what Jay just shared. When you look at our $1 million-plus customers today and we added net 16 customers, the highest we've had, the mix of that is very balanced across our portfolio. For example, like seven of those customers were driven by CP expansions are driven by CP. And then 50% of them are currently hybrid customers. And 13 out of those 16 customers actually had DSP in it. So fairly balanced. And I think that's something that we feel good about.

Sanjit，我想補充一點，我想說，與 Jay 剛剛分享的內容略有不同。如果您看我們今天的 100 多萬美元客戶，您會發現我們淨增了 16 位客戶，這是我們有史以來的最高紀錄，而我們的投資組合中的客戶組合非常均衡。例如，其中七個客戶是由 CP 擴張推動的。其中 50% 目前是混合客戶。在這 16 位客戶中，有 13 位實際上擁有 DSP。所以相當平衡。我認為這是我們感到高興的事情。

Yeah. I think another way to look at this is as we were going public, I think there was concern because our cloud revenue base was solid, like, hey, can these guys really be successful in the cloud? Can they really take a leadership position there? At least personally, I think that question is kind of answered. At this point, we really have a leadership -- product leadership, I think of the cloud, product leadership on-premise and a really good hybrid story. And so we feel at this point, it's kind of the customer's choice how they want to buy.

是的。我認為從另一個角度來看，當我們上市時，由於我們的雲端收入基礎穩固，所以人們會擔心，例如，嘿，這些人真的能在雲端運算領域取得成功嗎？他們真的能在那裡佔據領導地位嗎？至少就我個人而言，我認為這個問題已經得到了一定的答案。此時，我們確實擁有領導力——產品領導力，我認為是雲端、內部產品領導力和一個非常好的混合故事。因此，我們認為，就這一點而言，這在某種程度上是顧客如何購買的選擇。

Brad Zelnick, Deutsche Bank, followed by William Blair.

德意志銀行的布拉德‧澤爾尼克 (Brad Zelnick)，其次是威廉‧布萊爾 (William Blair)。

I wanted to follow up on Sanjit's question. I mean such a strong CP quarter out of the gate here in Q1. And I think within that term license is up like 60% year on year. And if I listen carefully to your comments, Rohan, in the prepared remarks, you mentioned the OEM relationships in international. I don't typically think of you guys as a strong OEM company, maybe not like Microsoft, but maybe I'm just not listening well enough. Can you just expand a little bit more on the nature of these types of relationships, the visibility you have to these deals, the typical term length? And what is the better-together story with these type of relationships?

我想跟進 Sanjit 的問題。我的意思是，第一季的 CP 表現非常強勁。我認為在這個期限內許可證數量將比去年同期增加 60%。羅漢，如果我仔細聽你的評論，在準備好的發言中，你提到了國際上的 OEM 關係。我通常不認為你們是一家強大的 OEM 公司，可能不像微軟，但也許我只是聽得不夠多。您能否更詳細地闡述這些類型的關係的性質、您對這些交易的了解程度以及典型的期限長度？那麼，在這種類型的關係中，更好的相處方式是怎麼樣的呢？

Yeah. Happy to take that, Brad. Yeah. From an overall CP business perspective, as you rightly called out, Q1 was a strong quarter, 18% growth. But I've said this before for CP. I think looking at this business over a 12-month period is actually a much better indicator. Typically, timing of some of these larger deals, timing of these renewals which are large could have an impact on revenue. But if you look at it from a 12-month lens, you'll see the consistency in the CP performance. And historically, we've called out strength from regulated industries. And over the last couple of quarters, we've been talking about our focus around the partner ecosystem.

是的。很高興接受這個，布拉德。是的。從整體 CP 業務角度來看，正如您正確指出的那樣，第一季表現強勁，成長了 18%。但我以前針對 CP 說過這一點。我認為以 12 個月的時間範圍來觀察這項業務實際上是一個更好的指標。通常，一些較大交易的時間、大型續約的時間可能會對收入產生影響。但如果從 12 個月的角度來看，您會看到 CP 表現的一致性。從歷史上看，我們一直呼籲受監管行業發揮力量。在過去的幾個季度中，我們一直在談論合作夥伴生態系統的重點。

And some of these OEM deals, I think, in a very simple manner, it just helps us amplify our message and send basically sell through as well through a lot of our OEM partners. So this particular OEM that we called out was primarily on the international side, which was strength. But it's not just from OEM. It's also from multiyear deals that we also had in Q1. And as we look at the rest of the year, I think the visibility is we'll see a good momentum in CP for rest of the year primarily because of the reasons that Jay just laid out.

我認為，其中一些 OEM 交易以一種非常簡單的方式幫助我們擴大我們的訊息，並透過我們的許多 OEM 合作夥伴實現銷售。因此，我們所提到的這個特定的 OEM 主要集中在國際領域，這是其優勢所在。但它不僅僅來自 OEM。這也是我們在第一季達成的多年期交易。當我們展望今年剩餘時間時，我認為我們會看到 CP 在今年剩餘時間內呈現良好勢頭，這主要是因為 Jay 剛剛提出的原因。

That's very helpful. If I could just follow up. Thank you for continuing to call out the impact from the change in comp on free cash flow. Anything here as we think about the remainder that would swing the growth and trajectory of free cash flow versus net income growth that we might want to keep in mind?

這非常有幫助。如果我可以跟進的話。感謝您持續強調薪資變動對自由現金流的影響。當我們考慮剩餘部分時，是否有任何因素會影響自由現金流與淨收入成長的成長和軌跡，我們可能需要記住這一點？

No. Our adjusted free cash flow goals for the year is approximately 6%. And outside of this onetime impact, we don't see any kind of adjustments out of our free cash flow number for rest of the year. So historically, what we've said outside of these timing of comp payments or timing of commission payments, generally, the free cash flow tends to go in line with the operating margin. And for the full year, for sure, kind of they're pretty closely related. And that's what you're going to see for rest of the year.

不。我們今年調整後的自由現金流目標約為 6%。除了這一次性影響之外，我們認為今年剩餘時間內我們的自由現金流數字不會有任何調整。因此，從歷史上看，除了補償金支付時間或佣金支付時間之外，我們所說的自由現金流一般傾向於與營業利潤率保持一致。就全年而言，它們肯定是密切相關的。這就是今年餘下時間你們將會看到的情況。

Jason Ader, William Blair, followed by Needham.

傑森‧阿德 (Jason Ader)、威廉‧布萊爾 (William Blair)，其次是尼德漢姆

How do you explain the strong net adds in Q1? Is it just the go-to-market changes that you made last year? Or is there anything else going on?

您如何解釋第一季強勁的淨增幅？這僅僅是您去年做出的行銷改變嗎？或者有其他事情發生嗎？

Yeah. There's a set of things. You have to look at it a little bit, as Rohan said, by the tier of customers. The folks entering the $1 million-plus is obviously a different explanation from that kind of top-of-the-funnel total customer count adds. At that top-of-the-funnel side, I would say we've made a set of improvements in the product-led side of the business, and we feel like that has kind of picked up a little bit of steam.

是的。有一系列的事情。正如羅漢所說，你必須根據客戶的層次來稍微看一下。進入 100 萬美元以上的人群顯然與漏斗頂部的總客戶數量增加有不同的解釋。在漏斗頂端，我想說我們在產品主導的業務方面做出了一系列改進，我們感覺這方面已經獲得了一些進展。

On the $1 million-plus side, it really is just customers continuing to progress in their streaming journey and helping them along. It's rare that companies start there. They usually start a bit smaller and grow into it. So that's just an evolution usually of work we've been doing with them over a year, multiyear, really depending on the case.

在 100 萬美元以上方面，這實際上只是客戶在串流媒體之旅中不斷進步並幫助他們前進。很少有公司從那裡起步。它們通常一開始較小，然後逐漸長大。所以這只是我們與他們合作一年或多年的工作的演變，實際上取決於具體情況。

Okay. And then a quick follow-up on -- for you, Jay. Just do you think the macro uncertainty will actually stretch out AI deployment time lines for customers?

好的。然後快速跟進一下——傑伊。您是否認為宏觀不確定性實際上會延長客戶的 AI 部署時間？

Yeah. It's a good question. It's a hard one to answer. We -- like what have we actually seen so far? So it's not the case that we've seen a bunch of projects put on hold or some sizable change yet. But it definitely is on the minds of customers. We have customers who are car companies or retailers, and they're thinking about this. And I think the answer is they're not sure, and I haven't seen the action yet. But certainly, when we're thinking about our guidance, the trajectory for the rest of the year, it makes us more cautious and conservative in what we're doing.

是的。這是個好問題。這是一個很難回答的問題。我們——到目前為止我們實際上看到了什麼？因此，我們還沒有看到大量項目被擱置或出現重大變化的情況。但這絕對是顧客關心的問題。我們的客戶是汽車公司或零售商，他們正在考慮這個問題。我認為答案是他們不確定，而且我還沒有看到行動。但當然，當我們考慮我們的指導方針、今年剩餘時間的發展軌跡時，我們會對所做的事情更加謹慎和保守。

Mike Cikos, Needham, followed by Mizuho.

麥克·西科斯（Mike Cikos）、尼德姆（Needham），其次是瑞穗（Mizuho）。

If I just come back to the consumption dynamics that we're talking to here, I know we've seen typical oscillations from larger customers before. With what played out in 1Q, just curious, can you give us any more color on when that started to evidence itself? Was it towards the back half of March? Was it the final couple of days? And then now that April is essentially in the rearview mirror, has that, in any way, deteriorated further and stable from where we were ending March?

如果我回到我們在這裡談論的消費動態，我知道我們之前已經看到來自較大客戶的典型波動。考慮到第一季發生的情況，我只是好奇，您能否告訴我們更多關於這種情況何時開始顯現的資訊？是在三月下半月嗎？是最後幾天嗎？那麼，四月基本上已經過去了，與三月底相比，情況是否進一步惡化並趨於穩定？

Yeah. So we saw this in March, what wasn't just the last few days. We certainly saw some impact through March. And then we've seen stability, but not an immediate rebound in April. And that's kind of thinking about that, thinking about just the larger environment, that's what made us more cautious overall in the cloud guidance.

是的。所以我們在三月就看到了這種情況，而不僅僅是最近幾天。我們確實在三月看到了一些影響。然後我們看到了穩定，但四月並沒有立即出現反彈。正是考慮到這一點，考慮到更大的環境，才使得我們在雲端指導方面整體上更加謹慎。

Got it. And for the follow-up, I think I went to Rohan with the prepared comments, but one of the reasons you attributed some of the strength for Confluent platform was specifically tied to the partners in international. Is it fair to think that with your -- the traction you're seeing with your partner ecosystem, CP, this is, call it, an incremental growth vector here just because certain geographies have a preference or a greater desire to go for CP over cloud?

知道了。至於後續問題，我想我已經帶著準備好的評論去了 Rohan，但您認為 Confluent 平台的部分優勢之一與國際合作夥伴密切相關。您是否認為，就您所看到的合作夥伴生態系統 CP 的吸引力而言，這是一個增量增長向量，僅僅是因為某些地區對 CP 而不是雲有偏好或更大的願望？

Yeah. Mike, I mean, when I think about the growth drivers for us, called out four growth drivers. The streaming opportunity DSP, which is not only net new but also expansion to our existing customers, AI as well as the partner ecosystem, which includes SIs, GSIs, OEMs, these are all not only Confluent Cloud, but they are also Confluent Platform opportunities. So I think that's probably one thing that I wanted to call out, I also called out in my prepared remarks. So short answer to your question is yes, it's across the board. And our strategy around being the partner ecosystem, Switzerland, where multidata destination, multi-cloud and also multiform factor is helping in the current environment.

是的。麥克，我的意思是，當我考慮我們的成長動力時，我列舉了四個成長動力。串流媒體機會 DSP 不僅是全新的，而且還擴展到我們現有的客戶、AI 以及合作夥伴生態系統，其中包括 SI、GSI、OEM，這些不僅是 Confluent Cloud，也是 Confluent Platform 的機會。所以我想這可能是我想說的一件事，我在準備好的發言中也提到了這一點。所以對你的問題的簡短答案是肯定的，它是全面的。我們的策略圍繞著成為瑞士的合作夥伴生態系統，其中多數據目的地、多雲和多形式因素正在為當前環境提供幫助。

Gregg Moskowitz, Mizuho, followed by Cowen.

格雷格‧莫斯科維茲 (Gregg Moskowitz)，瑞穗 (Mizuho)，其次是考恩 (Cowen)。

Jay, wondering if you could give us an update on WarpStream. Since acquiring the company, have you been able to close any meaningful transactions? And if so, has all of that come from the WarpStream pipe? Or was some of this Confluent-generated as you start to integrate it into the fold of the rest of the go-to-market, so to speak?

傑伊，想知道你是否可以向我們提供有關 WarpStream 的最新消息。自從收購該公司以來，您是否已經完成任何有意義的交易？如果是這樣，所有這些都來自 WarpStream 管道嗎？或者說，當您開始將其整合到其他行銷中時，其中一些是否是由 Confluent 產生的？

Yeah. So we've definitely had some really nice strong wins. I think we've shared a few of those. We've mentioned Cursor, but there's been a set of others that we brought forward in some of the past earnings. And yeah, those are a mixture of some things that were -- a few that were already in the pipeline, but we're actually seeing great activity and a set of wins across the larger team. And it's not surprising. This is something that was very close to home, fill the kind of natural niche in our product portfolio and was ready to sell. And so we kind of knew coming in to be one that could potentially hit the ground running, and that's what we've seen.

是的。所以我們確實取得了一些非常出色的勝利。我想我們已經分享了其中的一些。我們提到了 Cursor，但我們在過去的一些收益報告中也提到了其他一些公司。是的，這些都是一些事情的混合體——一些已經在籌備中，但我們實際上看到了整個團隊的出色活動和一系列勝利。這並不奇怪。這是一種與我們息息相關的產品，填補了我們產品組合中的自然空白，並且準備出售。因此，我們知道這是一個有可能迅速取得成功的項目，而這正是我們所看到的。

All right. Fantastic. And then just for Rohan, as part of the revised 2025 guide, you mentioned you're not assuming a near-term rebound in consumption. But maybe if we were to look at the other side of the coin, I'm just curious if any thought was given to assuming a further weakening of these trends just given greater unpredictability of the macro. Or from where you sit, from your perspective, does it just seems difficult to envision things degrading from here?

好的。極好的。然後就羅漢而言，作為修訂後的 2025 年指南的一部分，您提到您不認為消費會在短期內反彈。但如果我們看一下硬幣的另一面，我只是好奇是否有人考慮過假設這些趨勢會進一步減弱，因為宏觀經濟的不可預測性更大。或者從您所處的位置、從您的角度來看，是否很難想像事情會從現在開始惡化？

Yeah. I think the way we thought about it was, again, make sure that we have a prudent setup for rest of the year. So what are the two big assumptions behind it? We spoke about it, but I'll reiterate it because it's important. First, for our larger cloud customers, we know we did see some cost optimization slowdown in net new use cases.

是的。我認為我們的想法是，再次確保我們為今年剩餘時間做好審慎的準備。那麼這背後的兩個大假設是什麼呢？我們討論過這個問題，但我要重申這一點，因為它很重要。首先，對於我們較大的雲端客戶，我們知道我們確實看到淨新用例中的成本優化有所放緩。

We are not assuming an immediate near-term rebound in their consumption patterns. If you compare and contrast that with history, we've seen that -- historically, we've seen a rebound after optimization pattern. And you've got us call out the sawtooth pattern that we see. We're not assuming that same shape. That by itself provides some prudence to our guide as we look forward the year.

我們並不認為他們的消費模式會在短期內立即反彈。如果將其與歷史進行比較，我們會看到——從歷史上看，我們看到了優化模式後的反彈。你讓我們說出我們看到的鋸齒狀圖案。我們並不假設相同的形狀。這本身就為我們展望新的一年提供了一些謹慎的指導。

For Confluent Platform, it's slightly different because, of course, as Jay mentioned, there are deals that are already in the works. You have visibility into the pipeline. You have -- so our guide is based on a lot of visibility in the pipeline that we have from Confluent Platform. And I think the final takeaway from my perspective is if we take a step back and you look at our guide, we're calling 19% to 20% at a greater than $1 billion run rate. But we're not assuming any form of acceleration in the back half of the year. I think that setup makes us feel good as we head into what I'd call uncertain times, yes.

對於 Confluent Platform 來說，情況略有不同，因為當然，正如 Jay 提到的，有一些交易已經在進行中。您可以看到管道的情況。您有 — — 因此我們的指南是基於我們從 Confluent Platform 獲得的管道可見性。我認為從我的角度來看，最終的結論是，如果我們退一步看看我們的指南，我們預計運行率將達到 19% 到 20%，超過 10 億美元的運行率。但我們不認為今年下半年會出現任何形式的加速。是的，我認為這種設定讓我們在進入我所說的不確定的時期時感覺良好。

Derrick Wood, TD Cowen, followed by Barclays.

德里克伍德 (Derrick Wood)、TD Cowen，其次是巴克萊銀行。

I'll start with Jay. Anything to share around what kind of demand reception you've seen from Tableflow since it went GA? And just curious how to think about the revenue benefits. Is there a separate charge for Tableflow? Or is it more about just bringing new workloads onto the core platform?

我先從傑伊開始。自 Tableflow 正式發布以來，您能分享一下它的需求嗎？我只是好奇如何看待收入收益。Tableflow 是否需要另外收費？或者只是將新的工作負載引入核心平台？

Yeah. It's a great question. So the reception has been extremely strong. Now it's difficult, of course, for new products to figure out exactly how fast it will ramp and exactly to what size. So we're always cautious when we try and forecast anything around it.

是的。這是一個很好的問題。因此反應非常熱烈。當然，現在新產品的產量很難準確地確定成長速度和規模。因此，當我們嘗試預測任何事物時，我們總是很謹慎。

But I'd say, overall, it's about as strong as we've ever seen for any new product, and we're very excited about it. It will take us some time to get it across all the clouds and across all the different formats to fully open up and capitalize on the opportunity. That will happen over the course of this year. It is priced separately. So there's additional charges directly for the Tableflow usage.

但我想說，總的來說，它是我們見過的最強大的新產品，我們對此感到非常興奮。我們需要一些時間才能讓它跨越所有雲端和所有不同格式，從而完全開放並利用這個機會。這將在今年內實現。其價格另行計算。因此，使用 Tableflow 會產生額外費用。

But actually, the bigger opportunity in what we've seen with customers is, yes, they actually start to bring new data sets on entirely, right? So they'll spend some on Tableflow, but they'll spend some on Connect and Kafka and Flink along with that. And so the opportunity is really on both sides. And I think one of the reasons we're particularly excited about it is it's a very easy product to enable. For existing data, you can kind of just turn it on.

但實際上，我們在客戶身上看到的更大機會是，是的，他們實際上開始完全帶來新的數據集，對嗎？因此，他們會在 Tableflow 上花一些錢，但他們也會在 Connect、Kafka 和 Flink 上花一些錢。因此，雙方確實都有機會。我認為我們對它特別興奮的原因之一是它是一款非常容易啟用的產品。對於現有數據，您可以直接開啟它。

It doesn't require a ton of development work. And so we feel like, hey, as we really crack this and learn how to take it to market, it will be something that we can ramp relatively quickly. But of course, we have to first get it to GA and all the clouds and make it fully available over the course of this year to be able to do that.

它不需要大量的開發工作。因此我們覺得，嘿，當我們真正解決這個問題並學會如何將其推向市場時，它將是我們能夠相對快速地發展的東西。但當然，我們必須先將其推向 GA 和所有雲端，並在今年內使其完全可用，才能做到這一點。

Great. And then, Rohan, I mean, thanks for the color on the change in guidance assumptions. I wanted to ask about the government side of things. First, could you give us a sense of kind of how big of a business public sector or US Fed is? And then maybe talk about pipelines and how you may have changed your assumptions in the second half of this year given all the new developments on the DOGE side of things.

偉大的。然後，羅漢，我的意思是，感謝你對指導假設變化的闡述。我想詢問一下政府方面的事情。首先，您能否讓我們了解公共部門或聯準會的業務規模有多大？然後也許可以談談管道，以及考慮到 DOGE 方面的所有新發展，您可能如何在今年下半年改變您的假設。

Yeah. Derrick, thanks for your question. When we think about the government opportunity, I always put the government opportunity in the category of the opportunity is ahead of us because it's fairly underpenetrated. For context, our exposure to the federal government is in the low single digits, so immaterial. And the shape of the business is consistent with what you see in other companies.

是的。德里克，謝謝你的提問。當我們考慮政府機會時，我總是把政府機會放在我們面前的機會類別中，因為它的滲透程度相當低。就背景而言，我們對聯邦政府的風險敞口僅為個位數，因此無關緊要。該公司的業務形態與其他公司的情況一致。

Q3 tends to be a stronger Fed quarter. And some of these data points are baked into the guidance that we've set. So short answer, it's obviously fairly underpenetrated, big opportunity. And we spoke about getting FedRAMP status from a technology perspective. Now we are working with our government partner to get formal FedRAMP status. That will obviously open up the cloud business for us with the government is. So that's how we're thinking about it, small but --

第三季往往是聯準會表現較強勁的一個季度。其中一些數據點已納入我們設定的指導中。簡而言之，這顯然是一個尚未完全滲透的領域，是一個很大的機會。我們從技術角度討論如何獲得 FedRAMP 地位。現在我們正在與政府合作夥伴合作以獲得正式的 FedRAMP 地位。這顯然將為我們與政府打開雲端運算業務的大門。這就是我們的想法，雖然小，但--

And I think fair to say that in the current environment, very modest expectations for that area this year, right?

我認為可以公平地說，在當前環境下，今年對該領域的期望非常低，對嗎？

That's right. Very modest expectations.

這是正確的。期望非常低。

Raimo Lenschow, Barclays, followed by Stifel.

雷莫‧倫紹 (Raimo Lenschow)、巴克萊 (Barclays)，其次是 Stifel。

Two quick questions. One, Rohan, as you thought about the guidance, you mentioned more of the larger accounts have the issues. How do I think about that? Like in theory, this could play out two ways. So the larger account just do a little bit and then it's all good again or it kind of goes down deeper into large accounts are first and then SMB has a problem.

兩個簡單的問題。第一，羅漢，當你考慮指導時，你提到更多的大帳戶存在問題。我怎麼看待這個？理論上，這可能產生兩種結果。因此，較大的帳戶只需做一點，然後一切又會好起來，或者它會深入到大帳戶首先出現問題，然後 SMB 才會出現問題。

So when you thought about guidance, how would you kind of thought about that kind of reflecting that? And then one for Jay, if you think about -- obviously, there were kind of movements from other players in the broader ecosystem to think more about acquisitions in your space. Did you see anything in the customer conversations that they realized, okay, real time is a lot more important, people are trying to be it out? Are you seeing the benefit on that conversation that they realize, okay, you are actually in the real-time company, so I should talking more with you?

那麼當您考慮指導時，您會如何考慮如何反映這一點？然後我想問傑伊一個問題，如果你想想——顯然，更廣泛的生態系統中的其他參與者正在採取某種行動，更多地考慮在你的領域進行收購。您是否在與客戶的對話中看到他們意識到了什麼，好吧，即時性更為重要，人們正在努力實現這一點？您是否看到了他們意識到的對話的好處，好吧，您實際上是在實時公司，所以我應該與您多交談？

Yeah. To the second part of the question, I do think that there's definitely been a change in the perception of streaming. I wouldn't say it's a step change. I mean the -- in our space, I feel like people always expect there to be a single event that triggers it, but what we've actually seen is a real build, where this went for something that was viewed as kind of niche and on the side, like what could you really do with it? Just something that, of course, every company is doing but isn't necessarily part of the top line strategy, just something that's really like a major part of data strategy.

是的。對於問題的第二部分，我確實認為人們對串流媒體的看法確實發生了變化。我不會說這是一個重大的改變。我的意思是——在我們的領域，我覺得人們總是期望有一個單一事件來觸發它，但我們實際上看到的是一個真正的構建，它用於被視為一種小眾的東西，就像你能用它做什麼？當然，每家公司都在做這件事，但它不一定是頂線策略的一部分，而只是資料策略的重要組成部分。

And I think these kind of things always help that along, but there's no one event that kind of goes from 0 to 100%. And so yes, certainly that comes up in conversations with the customers. I don't think it's a huge catalyst on that in terms of how they're thinking about things. But there's no question that just -- if I were to just kind of go back on two-year increments and describe what does the conversation with the average CIO seem like? It kind of went from nothing what is any of the stuff to like, okay, Apache Kafka, yes, but like why does it matter?

我認為這些事情總是會有所幫助，但沒有一個事件可以從 0 上升到 100%。是的，這在與顧客的對話中肯定會出現。就他們思考問題的方式而言，我不認為這會是一個巨大的催化劑。但毫無疑問的是——如果我回顧一下兩年的增量並描述與普通 CIO 的對話是什麼樣的？它從無到有，有什麼東西可以喜歡，好的，Apache Kafka，是的，但為什麼它重要？

And what is Confluent to like really a broader appreciation of what's possible with streaming, what Confluent is, what could potentially be doable in the stream processing world as they think about kind of the real-time use of data. And that's not universal. It's not everybody has that understanding, but just more broadly across the industry that's changing. And I think it's a really powerful thing. When you think about -- there's a certain inevitability to the move to streaming and real-time data. It's sort of just aspect of how companies are changing. And I do think that, that benefits us.

Confluent 真正讓人們更廣泛地了解串流媒體的可能性，Confluent 是什麼，當他們考慮即時使用資料時，在串流處理領域可能實現什麼。但這並不是普遍現象。並不是每個人都有這樣的理解，但整個產業都在發生更廣泛的變化。我認為這是一件非常強大的事情。當你思考時－轉向串流媒體和即時數據具有一定的必然性。這只是公司變革的一個面向。我確實認為這對我們有好處。

Yeah. And Raimoo, to your first part of the question, for our smaller customers or not larger customers, our consumption patterns have been fairly steady. And that's a good news. And also, when you look at the net new customer adds that we have, a vast majority of our top-of-the-funnel net new ads come on the cloud side. So that is a steady stream of consumption customers flowing in. So the variation in that middle segment is a lot less. So that's why we feel good with the outlook that we have for rest of the year.

是的。Raimoo，關於您問題的第一部分，對於我們的小客戶或大客戶，我們的消費模式一直相當穩定。這是個好消息。而且，當您查看我們新增的淨客戶數量時，您會發現，我們漏斗頂端的絕大多數淨新廣告都來自雲端。因此，消費客戶就會源源不絕地湧入。因此中間部分的變化就小得多。這就是為什麼我們對今年剩餘時間的前景感到樂觀。

Yeah. And kind of adding to that a little color, it's like when you look behind the curtains, kind of customer by customer and the large customers, it's like, okay, is there some pattern or some big event that's taking place? Is there any reason to believe that would kind of spread? There's really not.

是的。並且添加一點色彩，就像當你看窗簾後面時，逐一查看客戶和大客戶，就好像，好吧，是否存在某種模式或正在發生的重大事件？有什麼理由相信這種現象會蔓延開來嗎？確實沒有。

It's just kind of the normal like customer activity you would see around infrastructure and optimization. And of course, the smaller customers do just have a certain law of large numbers in terms of how they operate. So yes, we don't see any kind of emerging trend there. When we put the pattern together, we certainly saw very strong pipe gen in the quarter. So kind of that top of funnel of what's coming in is really good. So we didn't see some kind of overall systematic softening of the demand environment or anything like that.

這只是您在基礎設施和優化方面看到的正常客戶活動。當然，小客戶的營運方式確實遵循一定的大數定律。所以是的，我們沒有看到任何新興趨勢。當我們把這些模式放在一起時，我們確實看到了本季非常強勁的管道生成。因此，漏斗頂部出現的情況確實很好。因此，我們並沒有看到需求環境出現某種整體系統性的疲軟或類似情況。

Brad Reback, Stifel with a final question coming from Raymond James.

Stifel 的 Brad Reback 提出了最後一個問題，來自 Raymond James。

As you kind of look at the business going forward in the dynamic macro, have you made any changes to your OpEx plans internally? Or are you cutting back on travel? Are you spending a little less? If you haven't done that yet, what potentially would cause you to do that?

當您從動態宏觀角度看待未來業務時，您是否對內部營運支出計劃做過任何更改？或者您正在減少旅行嗎？你的花費是不是少了一點？如果您還沒有這樣做，那麼有什麼原因可能導致您這樣做？

Yeah. We haven't done anything particularly aggressive, but we always manage expenses along with the top line. So we would obviously watch what happens throughout the year and make adjustments. We've certainly operated over the life of the company through very dynamic times and done all kinds of things around that. So yes, there's not been any big adjustment at this point.

是的。我們並沒有採取什麼特別激進的措施，但我們始終在控制收入的同時管理支出。因此，我們顯然會觀察全年發生的情況並做出調整。在公司發展過程中，我們確實經歷了非常動盪的時期，並圍繞著這段時期做了各種各樣的事情。是的，目前還沒有任何重大的調整。

Mark Cash, Raymond James.

馬克卡許、雷蒙詹姆斯。

For Jay, you mentioned how last year a lot more go-to-market adjustments with the quota alignment and now thinking about this year and having more products and employment options to sell. So how would you characterize sales productivity? And I think in particular, can most reps and partners go and incrementally sell DSP use cases or sell WarpStream and Freight versus sun cloud and platform previously?

對於傑伊，您提到了去年如何透過配額調整進行更多的市場調整，現在考慮今年如何銷售更多的產品和就業選擇。那麼您如何描述銷售生產力？我特別認為，大多數銷售代表和合作夥伴是否可以逐步銷售 DSP 用例或銷售 WarpStream 和 Freight，而不是先前的 Sun Cloud 和平台？

Yeah. It's certainly the case that we feel that the addition to the product portfolio to the other Kafka SKUs that kind of filling gaps in DSP that over time, that's a tailwind because we can land and convert more of the existing Kafka. We can turn these into bigger use cases, bigger platforms, a bigger spend customer by customer. The new product things are still smaller.

是的。我們確實認為，將產品組合添加到其他 Kafka SKU 中可以填補 DSP 中的空白，隨著時間的推移，這將是一個順風，因為我們可以登陸和轉換更多現有的 Kafka。我們可以將這些轉化為更大的用例、更大的平台、更大的客戶支出。新產品的東西還是比較小的。

So they move the overall numbers a little bit and grow fast on a percentage basis. As they get to scale, they move the overall numbers a lot. And we've been through a few of those curves before with the ramp of our cloud product and other things. And so yes, that's kind of the way we look at it. We watch these new things very closely to make sure that they're on the ramp that we want. And we would expect to see that kind of have broad impact as they get to scale.

因此，它們會稍微改變總體數字，並在百分比基礎上快速增長。隨著規模的擴大，整體數字會發生很大變化。隨著我們的雲端產品和其他產品的發展，我們也經歷過一些這樣的曲線。是的，這就是我們看待這個問題的方式。我們密切關注這些新事物，以確保它們能夠以我們想要的方式發展。我們期望看到，隨著規模擴大，這種影響將十分廣泛。

Great. Thanks, Mark. And this concludes our earnings call. Thanks again for joining us. Have a nice evening. We'll see you soon.

偉大的。謝謝，馬克。我們的收益電話會議到此結束。再次感謝您的加入。祝你今晚愉快。我們很快就會見到你。

Thanks, everyone.

謝謝大家。

Thank you, everyone

謝謝大家