GitLab Inc (GTLB) 2025 Q2 法說會逐字稿

(audio in progress)

（音訊正在播放）

(Operator Instructions) Please note today's call is being recorded. I will be standing by should you need assistance.

（操作員指示）請注意，今天的通話正在錄音。如果您需要幫助，我將隨時待命。

And now, it is my pleasure to turn the conference over to Kelsey Turcotte. Kelsey, over to you.

現在，我很高興將會議交給凱爾西·特科特 (Kelsey Turcotte)。凱爾西，交給你了。

Good afternoon. We appreciate you joining us for GitLab's second-quarter fiscal year 2025 financial results conference call. GitLab's Co-Founder and CEO, Sid Sijbrandij; and GitLab's Chief Financial Officer, Brian Robins, will provide commentary on the quarter and guidance for the fiscal year.

午安.感謝您參加 GitLab 2025 財年第二季財務業績電話會議。GitLab 聯合創辦人兼執行長 Sid Sijbrandij； GitLab 財務長 Brian Robins 將對本季發表評論並對本財年提供指導。

Before we begin, I'll cover the Safe Harbor statement. I would like to direct you to the cautionary statement regarding forward-looking statements on page 2 of our presentation and in our earnings release issued earlier today, which are both available under the Investor Relations section of our website. The presentation and earnings release include a discussion of certain risks, uncertainties, assumptions, and other factors that could cause our results to differ from those expressed in any forward-looking statements within the meaning of the Private Securities Litigation Reform Act. As is customary, the content of today's call and presentation will be governed by this language.

在我們開始之前，我將介紹安全港聲明。我想請您參閱我們簡報第 2 頁和今天稍早發布的收益報告中有關前瞻性陳述的警告聲明，這些聲明均可在我們網站的「投資者關係」部分下找到。本示範和收益報告討論了某些風險、不確定性、假設和其他因素，這些因素可能導致我們的結果與《私人證券訴訟改革法案》所定義的任何前瞻性陳述中表達的結果不同。按照慣例，今天的電話會議和演講的內容將遵循這一語言。

In addition, during today's call, we will be discussing certain non-GAAP financial measures. These non-GAAP financial measures exclude certain unusual or non-recurring items that management believes impact the comparability of the periods referenced. Please refer to our earnings release and presentation materials for additional information regarding these non-GAAP financial measures and the reconciliations to the most directly comparable GAAP measure.

此外，在今天的電話會議中，我們將討論某些非公認會計準則財務指標。這些非公認會計準則財務指標不包括管理階層認為會影響參考期間可比較性的某些不尋常或非經常性項目。有關這些非 GAAP 財務指標以及與最直接可比較的 GAAP 指標的調節的更多信息，請參閱我們的收益報告和演示材料。

I will now turn the call over to GitLab's Co-Founder and Chief Executive Officer, Sid Sijbrandij.

現在我將把電話轉給 GitLab 的聯合創始人兼執行長 Sid Sijbrandij。

Thank you for joining us today. I'm excited to share our second-quarter results with you and talk about our market-leading DevSecOps platform and its AI capabilities. We continue to deliver strong results that reflect our team's focus on customers, helping them realize faster time-to-value and customer-specific business outcomes.

感謝您今天加入我們。我很高興與您分享我們的第二季業績，並討論我們市場領先的 DevSecOps 平台及其 AI 功能。我們繼續提供強勁的業績，這反映了我們團隊對客戶的關注，幫助他們更快地實現價值和特定於客戶的業務成果。

Second-quarter revenue increased 31% year over year to $183 million, driven by new logos like Delaware North and Guild Mortgage, as well as expansion by existing customers. Our non-GAAP operating margin also meaningfully exceeded our expectations in the quarter, increasing over 1,300 basis points year over year to 10%. This underscores our continued commitment to responsible growth.

第二季營收年增 31% 至 1.83 億美元，這得益於 Delaware North 和 Guild Mortgage 等新品牌以及現有客戶的擴張。本季度，我們的非公認會計準則營業利潤率也大大超出了我們的預期，年成長超過 1,300 個基點，達到 10%。這強調了我們對負責任成長的持續承諾。

Now, more than ever, organizations need to deliver software faster to respond to intense competition and accelerate performance. This is what our DevSecOps platform is purpose-built to do. We bring together developers, security experts, and operations teams to better collaborate, improve quality, and prioritize security, all while decreasing software delivery cycle times. And with AI integrated throughout the software development life cycle, GitLab customers can take those gains even further.

現在，企業比以往任何時候都更需要更快地交付軟體，以應對激烈的競爭並提高績效。這就是我們的 DevSecOps 平台的專門建置目的。我們將開發人員、安全專家和營運團隊聚集在一起，以便更好地協作、提高品質並優先考慮安全性，同時縮短軟體交付週期。透過在整個軟體開發生命週期中整合 AI，GitLab 客戶可以進一步獲得這些收益。

Enterprises are focusing on real results and real use cases for AI. They're looking beyond just code generation. They are looking to integrate AI into all aspects of software development to deliver tangible results. This requires a strategic approach that aligns AI solutions with business goals, provides measurable benefits, and improves security. And this is where GitLab excels.

企業關注的是人工智慧的實際成果和實際用例。他們關注的不僅僅是程式碼生成。他們希望將人工智慧融入軟體開發的各個方面，以取得實際的成果。這需要一種策略方法，將人工智慧解決方案與業務目標結合，提供可衡量的效益並提高安全性。這正是 GitLab 的優勢所在。

Our customers are excited about the meaningful productivity and security benefits of GitLab Duo, which has demonstrated up to 90% reduction in time spent on toolchain operations, 50% faster lead time, and 50% faster vulnerability detection. They are also excited about our ability to help them drive real business outcomes. For example, we recently heard from the State of Washington Public Disclosure Commission about GitLab Duo. They mentioned that GitLab Duo is improving their developer productivity and effectiveness, which is helping their teams focus more on the substance of their work.

我們的客戶對 GitLab Duo 顯著的生產力和安全性優勢感到非常興奮，它已證明可將工具鏈操作所花費的時間減少 90%，交貨時間縮短 50%，漏洞檢測速度縮短 50%。他們也對我們幫助他們實現實際業務成果的能力感到興奮。例如，我們最近從華盛頓州公共揭露委員會聽說了 GitLab Duo。他們提到 GitLab Duo 正在提高他們的開發人員的生產力和效率，這有助於他們的團隊更專注於實質工作。

AI is also resulting in larger deal sizes. Barclays purchased GitLab Duo seats this quarter, coupled with additional Ultimate licenses. They are rolling out GitLab Duo to thousands of developers so they can take advantage of AI-powered capabilities in the same platform where they're building and deploying their code. It's exciting to see large enterprises like Barclays adopt AI as a natural step for simplifying toolchains and improving their developer experience.

人工智慧也使得交易規模變得更大。巴克萊銀行本季購買了 GitLab Duo 席位，以及額外的 Ultimate 許可證。他們正在向數千名開發人員推出 GitLab Duo，以便他們可以在建置和部署程式碼的相同平台上利用人工智慧功能。令人興奮的是，看到像巴克萊這樣的大型企業採用人工智慧作為簡化工具鏈和改善開發人員體驗的自然步驟。

F5, a multi-cloud application security and delivery company, is yet another customer who's adopting GitLab Duo after seeing value from Ultimate in driving improved developer experience and productivity. Developers who participated in F5's pilot of GitLab Duo shared that our AI capabilities are easy to use and help them to be more productive in their work. Now, F5 is rolling out GitLab Duo to all of the company's 2,000 developers.

F5 是一家多雲應用安全和交付公司，也是另一個採用 GitLab Duo 的客戶，因為它看到了 Ultimate 在提升開發人員體驗和生產力方面的價值。參與 F5 GitLab Duo 試點的開發人員表示，我們的 AI 功能易於使用，並能幫助他們提高工作效率。現在，F5 正在向該公司全部 2,000 名開發人員推出 GitLab Duo。

The KeyBank team also wanted to improve developer productivity, so they adopted GitLab Duo. Combined with our DevSecOps platform, GitLab Duo is helping KeyBank developers resolve pipeline issues six times faster. GitLab Duo is KeyBank's first approved AI technology due to our focus on transparency and privacy first. It's the combination of both our end-to-end platform and AI that's driving results for our customers.

KeyBank 團隊也希望提高開發人員的工作效率，因此他們採用了 GitLab Duo。結合我們的 DevSecOps 平台，GitLab Duo 正在幫助 KeyBank 開發人員以六倍的速度解決管道問題。GitLab Duo 是 KeyBank 首個獲準的人工智慧技術，因為我們首先專注於透明度和隱私。我們的端到端平台和人工智慧的結合為我們的客戶帶來了豐厚的成果。

According to Gartner, by 2027, the number of platform engineering teams using AI to augment every phase of the SDLC will have increased from 5% to [14%]. This is why we're very pleased with the outcomes of two of Gartner's recent Magic Quadrants. First, GitLab was recently recognized as a leader in the first-ever 2024 Gartner Magic Quadrant for AI code assistance. We believe this recognition highlights our commitment to delivering AI-powered capabilities that accelerate software delivery, enhance security, and drive innovation for our customers.

據 Gartner 稱，到 2027 年，使用 AI 來增強 SDLC 每個階段的平台工程團隊數量將從 5% 增加到[14%]。這就是我們對 Gartner 最近的兩個魔力像限的結果感到非常滿意的原因。首先，GitLab 最近被評為 2024 年首個 Gartner AI 程式碼輔助魔力像限的領導者。我們相信，這項認可彰顯了我們致力於提供人工智慧功能的承諾，這些功能可加速軟體交付、增強安全性並推動客戶創新。

And for the second year in a row, GitLab was recognized as a leader in Gartner's 2024 Magic Quadrant for DevOps platforms. We were positioned highest in both our ability to execute and our completeness of vision. Our market leadership comes from our integrated security and compliance capabilities, deployment flexibility, and our unified data store. This provides end-to-end context across an organization's entire software development and deployment workflow.

GitLab 連續第二年被 Gartner 評為 2024 年 DevOps 平台魔力像限的領導者。我們在執行能力和願景完整性方面均名列前茅。我們的市場領導地位來自於我們的綜合安全和合規能力、部署靈活性以及統一的資料儲存。這為組織的整個軟體開發和部署工作流程提供了端到端的背景資訊。

Customers realize significant return on investment from adoption of our DevSecOps platform. Our new Forrester study on the total economic impact of GitLab Ultimate found that organizations can achieve 482% return on investment over three years. That's a nearly 60-percentage-point increase over the last time we conducted this study two years ago. These results are based on our continued focus on increasing developer productivity, improving developer experience, and accelerating feature delivery, improving security, and toolchain consolidation.

客戶透過採用我們的 DevSecOps 平台獲得了可觀的投資回報。我們對 GitLab Ultimate 的總體經濟影響進行的一項新的 Forrester 研究發現，組織可以在三年內實現 482% 的投資回報。這比我們兩年前進行這項研究時增加了近 60 個百分點。這些結果是基於我們持續專注於提高開發人員生產力、改善開發人員體驗、加速功能交付、提高安全性和工具鏈整合。

Toolchain consolidation and related benefits are mentioned frequently by our customers. In fact, in our annual survey of more than 5,000 DevSecOps professionals, 62% reported that their teams use more than five tools and 64% wanted to consolidate their toolchain to drive efficiencies. This presents a tremendous opportunity for us as a platform to allow customers to consolidate vendors and reduce total costs of ownership.

我們的客戶經常提到工具鏈整合和相關好處。事實上，在我們對 5,000 多名 DevSecOps 專業人士進行的年度調查中，62% 的人表示他們的團隊使用超過五種工具，64% 的人希望整合他們的工具鏈以提高效率。這為我們提供了一個平台，讓我們能夠讓客戶整合供應商並降低整體擁有成本。

One of GitLab's strengths is our ability to help customers replace legacy point solutions. For example, by consolidating on GitLab, Lockheed Martin managed to run CI pipeline builds 80 times faster, retired thousands of legacy CI servers, and reduced the time spent on system maintenance by 90%. That shift resulted in a significant increase in efficiency and productivity for Lockheed Martin.

GitLab 的優點之一是我們能夠幫助客戶替換傳統的點解決方案。例如，透過整合 GitLab，洛克希德馬丁公司成功將 CI 管道建置速度提高了 80 倍，淘汰了數千台舊式 CI 伺服器，並將系統維護所花費的時間減少了 90%。這一轉變顯著提高了洛克希德馬丁公司的效率和生產力。

Another example is one of the world's leading innovators in materials science. They are using our Enterprise Agile Planning add-on, in combination with GitLab Ultimate, to centralize planning tools into a single platform and improve visibility across the business.

另一個例子是世界材料科學領域的領先創新者之一。他們正在使用我們的企業敏捷規劃外掛程式與 GitLab Ultimate 結合，將規劃工具集中到單一平台並提高整個業務的可見度。

Our customers also see security as a mission-critical need. Recent news cycles continue to increase awareness and urgency in the executive suites about the need to embed security at the earliest stages of software development. GitLab Ultimate helps solve this by shifting security left in the process, seamlessly instrumenting security checks and guardrails into the software development pipeline.

我們的客戶也將安全視為一項關鍵任務需求。最近的新聞週期不斷提高高階主管對於在軟體開發的早期階段嵌入安全性的必要性的認識和緊迫性。GitLab Ultimate 透過將安全性轉移到流程的左側，將安全檢查和護欄無縫地插入軟體開發流程來幫助解決此問題。

This comprehensive approach not only earns the confidence of security leaders, but also dramatically enhances the developer experience. By eliminating context switching and post-deployment firefighting, developers maintain their crucial state of flow. This results in faster, more secure code delivery. GitLab transforms security from a bottleneck into a strategic advantage in innovation and reliability.

這種綜合方法不僅贏得了安全領導者的信任，而且大大增強了開發人員的體驗。透過消除上下文切換和部署後的救火工作，開發人員可以保持其關鍵的流動狀態。這使得程式碼交付更快、更安全。GitLab 將安全性從瓶頸轉變為創新和可靠性的戰略優勢。

Security and compliance capabilities are at the heart of Ultimate and set us apart from the competition. Ultimate is a particularly good fit for customers who require the enterprise-grade capabilities of our platform to meet constant demands to move faster and produce more software. In the second quarter, 7 of our 10 largest deals were Ultimate purchases, and 7 of our top 10 first order customers landed with Ultimate. At the end of Q2, Ultimate is now 47% of total ARR. For example, the National Oceanic and Atmospheric Administration upgraded from Premium to Ultimate this quarter for its enhanced security and compliance. They also purchased GitLab Duo to improve their developer productivity.

安全性和合規性能力是 Ultimate 的核心，並使我們在競爭中脫穎而出。Ultimate 特別適合那些需要我們平台的企業級功能來滿足更快發展和生產更多軟體的持續需求的客戶。在第二季度，我們 10 筆最大交易中有 7 筆是透過 Ultimate 採購的，我們 10 大首筆訂單客戶中有 7 筆是透過 Ultimate 採購的。截至第二季末，Ultimate 佔總 ARR 的 47%。例如，美國國家海洋暨大氣總署本季從高級版升級到終極版，以增強其安全性和合規性。他們還購買了 GitLab Duo 來提高開發人員的工作效率。

Security is an important factor when customers adopt GitLab Dedicated, our single-tenant SaaS offering that is completely managed by GitLab. This offering is unique in the market and is especially valuable to companies with highly complex security and compliance requirements and in regulated industries such as the public sector and financial services.

當客戶採用 GitLab Dedicated（我們的完全由 GitLab 管理的單一租戶 SaaS 產品）時，安全性是一個重要因素。該產品在市場上是獨一無二的，對於具有高度複雜的安全性和合規性要求的公司以及公共部門和金融服務等受監管行業的公司尤其有價值。

Snowflake recently migrated to GitLab Dedicated for source code management, CI, and security for their corporate environment. With GitLab Dedicated, Snowflake has the security of a single-tenant environment plus all the benefits of an end-to-end DevSecOps platform.

Snowflake 最近遷移到 GitLab Dedicated 來為其企業環境提供原始碼管理、CI 和安全保障。透過 GitLab Dedicated，Snowflake 擁有單一租戶環境的安全性以及端對端 DevSecOps 平台的所有優點。

We're also excited to share that we have achieved the in-process designation for FedRAMP Moderate. GitLab Dedicated for Government helps public sector agencies and customers in highly regulated industries meet stringent security and compliance requirements from the US government. We expect this designation to build upon the significant momentum we already have in the public sector.

我們也很高興地宣布，我們已經獲得了 FedRAMP Moderate 的進程內認證。GitLab 專門為政府服務，幫助公共部門機構和高度監管行業的客戶滿足美國政府嚴格的安全和合規要求。我們期望這一頭銜能進一步鞏固我們在公共部門已擁有的巨大發展勢頭。

In summary, Q2 was a good quarter, and I'm proud of what we accomplished. I also want to thank the GitLab team for everything you contributed to our ongoing success. Looking at the second half of fiscal 25, I'm really energized by our ability to continue to drive customer success and the opportunity we have with AI to accelerate business outcomes.

總而言之，第二季表現不錯，我為我們所取得的成就感到自豪。我還要感謝 GitLab 團隊為我們持續成功所做的一切貢獻。展望 25 財年下半年，我為我們繼續推動客戶成功的能力以及利用人工智慧加速業務成果的機會感到非常振奮。

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

說完這些，我會把話題交給 Brian。

Thank you, Sid, and thank you again for everyone joining us today. This quarter's results validates the value that our customers get from our integrated platform. In today's cautious macroeconomic environment, technology needs to deliver quick time to value while solving complex, impactful problems. That is what our AI-powered DevSecOps platform does.

謝謝你，Sid，再次感謝今天與我們一起的各位朋友。本季度的業績驗證了我們的客戶從我們的整合平台獲得的價值。在當今謹慎的宏觀經濟環境中，科技需要在解決複雜、有影響力的問題的同時快速實現價值。這就是我們的人工智慧 DevSecOps 平台所做的事情。

A great example is Intuitive Machines, which became the first US venture in 50 years to land a spacecraft on the moon. Integral to the success of the project was GitLab. Our end-to-end platform enabled dozens of developers to write code, gain visibility, and collaborate on shared projects. The result was a 10x increase in release cadence, 99% reduction in downtime, and 20x decrease in pipeline execution time. Quoting one of the software leads on the project, we absolutely could not have built a spacecraft in five years without GitLab. It helped us make history.

一個很好的例子就是直覺機器公司 (Intuitive Machines)，它成為 50 年來第一個將太空船登陸月球的美國企業。GitLab 是此專案成功的關鍵。我們的端到端平台使數十名開發人員能夠編寫程式碼、獲得可見性並協作進行共享專案。結果是發布節奏增加了 10 倍，停機時間減少了 99%，管道執行時間減少了 20 倍。引用該專案一位軟體負責人的話來說，如果沒有 GitLab，我們絕對不可能在五年內建造出一艘太空船。它幫助我們創造歷史。

Turning to Q2 FY25, results exceeded our expectations as we delivered another quarter of greater than 30% top-line growth and significant year-over-year operating margin expansion. Second-quarter revenue reached $182.6 million, an increase of 31% from Q2 of the prior year.

展望 2025 財年第二季度，業績超出了我們的預期，我們又一個季度實現了超過 30% 的營收成長，且營業利潤率同比大幅擴大。第二季營收達1.826億美元，較去年同期成長31%。

We ended the quarter with dollar-based net retention rate, or DBNRR, of 126%. Q2 DBNRR was driven by a combination of seat expansion at approximately 40%, increased customer yield at approximately 50%, and tier upgrades at approximately 10%. In addition, all of our historical cohorts continue to steadily expand.

本季結束時，我們的美元淨留存率（DBNRR）為 126%。Q2 DBNRR 的成長動力包括席位擴大約 40%、客戶收益增加約 50% 以及層級升級約 10%。此外，我們所有的歷史群體都在繼續穩步擴大。

We now have 9,314 customers with ARR of at least $5,000, an increase of approximately 19% year over year, and contributed over 95% of total ARR in Q2. In particular, we monitored performance of our larger customer cohort of $100,000 plus in ARR, which reached 1,076 this quarter, an increase of 33% year over year. In fact, more than 65% of new dollars invested by this cohort was in Ultimate this quarter.

我們目前擁有 9,314 名 ARR 至少為 5,000 美元的客戶，年增約 19%，並在第二季度貢獻了總 ARR 的 95% 以上。特別是，我們監控了 ARR 為 100,000 美元以上的較大客戶群的表現，本季度該客戶群達到了 1,076 個，同比增長 33%。事實上，本季該群體投資的新資金中有 65% 以上都投入了 Ultimate。

A great example of customer success with these large customers is Bol, one of the biggest online retailers in the Netherlands. As Bol's revenues grew, they needed to keep up with the strict and constantly changing compliance regulations. With GitLab, Bol can set up policies that automate compliance configurations and checks, saving thousands of developer hours per month.

荷蘭最大的線上零售商之一 Bol 就是這些大客戶成功的一個很好的例子。隨著 Bol 的收入不斷增長，他們需要跟上嚴格且不斷變化的合規法規。借助 GitLab，Bol 可以設定自動化合規性配置和檢查的策略，從而每月節省數千個開發人員小時。

This quarter, total RPO grew 51% year over year to $747.9 million, while CRPO grew 42% year over year to $475 million.

本季度，RPO總額年增51%至7.479億美元，而CRPO較去年同期成長42%至4.75億美元。

Non-GAAP gross margins were 91% for the quarter. SaaS now represents 28% of total revenue, in part a reflection of considerable traction we were getting with GitLab Dedicated. Year-over-year SaaS revenue grew 46%. Given the continued high growth in SaaS, I am very happy with the team's attention to operating efficiencies, which continues to result in best-in-class non-GAAP gross margins.

非公認會計準則本季毛利率為91%。SaaS 現在佔總收入的 28%，一定程度上反映了我們在 GitLab Dedicated 上獲得的巨大吸引力。SaaS營收年增46%。鑑於 SaaS 的持續高速成長，我很高興看到團隊對營運效率的關注，這繼續帶來一流的非 GAAP 毛利率。

Once again, we saw a significant year-over-year improvement in operating leverage. Q2 non-GAAP operating income was $18.2 million, compared to a loss of $4.3 million in the second quarter of last year. This quarter, we dropped all of our revenue outperformance to the bottom line which, in combination with the team's continued focus on smart resource allocation, translated to non-GAAP operating margin of 10%, compared to negative 3.1% in the Q2 of last year. This once again demonstrates our commitment to responsible growth.

我們再次看到經營槓桿較去年同期顯著改善。第二季非公認會計準則營業收入為 1,820 萬美元，而去年第二季虧損 430 萬美元。本季度，我們將所有收入超額部分降至底線，再加上團隊繼續專注於智慧資源配置，非公認會計準則下的營業利潤率為 10%，而去年第二季度為負 3.1%。這再次證明了我們對負責任成長的承諾。

Cash from operating activities was $11.7 million in the second quarter, compared to $27.1 million in the prior year period. Adjusted free cash flow was $10.8 million in the second quarter of FY25, compared to $26.8 million in the prior year period. Q2 FY25 cash flow from operations and adjusted free cash flow reflect the timing of payments for a Q1 global employee gathering made in Q2.

第二季經營活動現金為 1,170 萬美元，去年同期為 2,710 萬美元。25 財年第二季調整後自由現金流為 1,080 萬美元，去年同期為 2,680 萬美元。2025 財年第二季的經營現金流和調整後的自由現金流反映了第二季第一季全球員工聚會的付款時間。

Now, turning to guidance, for the third quarter of FY25, we expect total revenue of $187 million to $188 million, representing a growth rate of 25% to 26% year over year. We expect a non-GAAP operating income of $19 million to $20 million, and we expect a non-GAAP net income per share of $0.15 to $0.16, assuming $168 million weighted average diluted share is outstanding.

現在，談到指引，對於 25 財年第三季度，我們預計總營收為 1.87 億美元至 1.88 億美元，年增 25% 至 26%。我們預期非 GAAP 營業收入為 1,900 萬美元至 2,000 萬美元，假設流通在外的加權平均攤薄股份為 1.68 億美元，非 GAAP 每股淨收入為 0.15 美元至 0.16 美元。

For the full year FY25, we expect total revenue of $742 million to $744 million, representing a growth rate of approximately 28% year over year. We expect a non-GAAP operating income of $55 million to $58 million, and we expect a non-GAAP net income per share of $0.45 to $0.47, assuming $168 million weighted average diluted share is outstanding.

對於 25 財年全年，我們預計總營收為 7.42 億美元至 7.44 億美元，年增約 28%。我們預期非 GAAP 營業收入為 5,500 萬美元至 5,800 萬美元，假設流通在外的加權平均攤薄股份為 1.68 億美元，非 GAAP 每股淨收入為 0.45 美元至 0.47 美元。

Separately, I'd like to provide an update on JiHu, our China joint venture. In Q2 FY25, non-GAAP expenses related to JiHu were $3.3 million, compared to $4.8 million in Q2 of last year. Our goal remains to deconsolidate JiHu. However, we cannot predict the likelihood or timing when this may potentially occur. Thus, for FY25 modeling purposes, we forecast approximately $14 million of expenses related to JiHu, compared with $18 million of last year.

另外，我想介紹一下我們在中國的合資企業奇虎的最新情況。2025財年第二季度，與奇虎相關的非GAAP費用為330萬美元，去年第二季為480萬美元。我們的目標仍然是拆分奇虎。然而，我們無法預測這種情況發生的可能性或時間。因此，出於 FY25 建模目的，我們預測與奇虎相關的費用約為 1400 萬美元，而去年為 1800 萬美元。

Thank you all for joining this afternoon. We delivered a strong Q2, and I'm really pleased with how we are positioned as we head into the back half of FY25. We appreciate your support and look forward to speaking with many of you during the quarter.

感謝大家今天下午的參與。我們第二季的業績表現強勁，對於我們進入 25 財年下半年的定位，我感到非常滿意。我們感謝您的支持，並期待在本季度與你們中的許多人進行交談。

With that, I will turn it over to Kelsey, who will moderate the Q&A.

說完這些，我會把話題交給 Kelsey，他將主持問答環節。

(Event Instructions) Joel Fishbein, Truist.

（活動說明）Joel Fishbein，Truist。

Thanks, Kelsey, for the question, and congrats on a strong quarter. I'd love to get an update from you on how you're thinking about the go-to-market going forward. Obviously, you had a changeover in leadership there, and despite that, you had a very solid quarter. Just curious what you're looking for in a new leader. And just a quick question for you, Brian, on a follow-up.

謝謝 Kelsey 提出的問題，並祝賀您本季業績表現強勁。我很想了解您對未來行銷的最新想法。顯然，你們的領導層發生了變化，儘管如此，你們本季的業績仍然非常穩健。只是好奇您對新領導者的要求是怎樣的。布萊恩，我只想問你一個簡單的後續問題。

Thanks for that, Joel. Yeah, Chris was doing a great job. I would love to continue to have him here, but he found another opportunity. We've launched a search, and we've been really impressed with the quality and the quantity of candidates.

謝謝你，喬爾。是的，克里斯做得很好。我很想讓他繼續留在這裡，但他找到了另一個機會。我們已經啟動搜索，候選人的品質和數量給我們留下了深刻的印象。

And in the interim, we're really pleased with how smoothly the team has transitioned to Ashley's leadership. In her role as Chief Marketing and Strategy Officer, she was already very in touch with both our customers and the sales leadership. But that enabled a smooth transition, which supported our raising guidance this quarter.

在此期間，我們對團隊順利過渡到阿什利的領導感到非常高興。在擔任首席行銷和策略長期間，她已經與我們的客戶和銷售領導層保持密切聯繫。但這實現了平穩過渡，並支持我們本季提高預期。

Great. And Brian, just quick one for you. I mean, you outperformed very significantly in the operating margin line. I believe you're still going to prioritize growth over margins, but any color you could give us there would be really helpful. Thank you.

偉大的。Brian，我只想問你一個快速問題。我的意思是，你們的營業利潤率表現非常出色。我相信您仍然會優先考慮成長而不是利潤，但您能給我們的任何細節都會非常有幫助。謝謝。

Thanks, Joel. I appreciate that. Yeah, nothing's changed on that pre-IPO every quarter. Sid and I have said growth is the number one thing, but we'll do that responsibly. And we continue to get increased operating leverage in the model, even growing greater than 30% year over year. So happy with the performance this quarter and happy with the beat and raise for the year.

謝謝，喬爾。我很感激。是的，每季的 IPO 前情況都沒有任何變化。西德和我都說過，成長是首要的事情，但我們會以負責任的方式去做。而我們在該模型中持續提高經營槓桿，甚至較去年同期成長超過30%。我對本季的表現非常滿意，對今年的業績和成長也感到滿意。

Great. Thank you.

偉大的。謝謝。

Ryan MacWilliams, Barclays.

巴克萊銀行的瑞安‧麥克威廉斯 (Ryan MacWilliams)。

Hey. Thanks for the question. Just in terms of what you're seeing in the macro right now, do you notice any differences between the second quarter and the first quarter? And how are you seeing developer hiring at this point? Thanks.

嘿。謝謝你的提問。就您現在看到的宏觀情況而言，您是否注意到第二季和第一季之間有什麼差異？您目前對開發人員的招募情況如何？謝謝。

Yeah, thanks for the question. There's really been no difference between the two quarters. It still remains a cautious spending environment out there. And so we're enabling our teams to go out there, and the pitches are more financially related.

是的，謝謝你的提問。這兩個季度之間確實沒有什麼區別。目前仍處於謹慎的消費環境。因此，我們讓我們的團隊能夠出去，並且投球也更多地與財務相關。

And we haven't seen really any changes in trends on developer hiring either. And so first order continues to remain really strong. And that's really the power of the platform in the payback period that Sid talked about with the Forrester report and the ROI that our customers are receiving.

而且我們也沒有看到開發人員招募趨勢有任何變化。因此，第一批訂單繼續保持強勁。這確實就是 Sid 在 Forrester 報告中談到的投資回收期平台的威力以及我們的客戶所獲得的投資報酬率。

Second question was around developer hiring.

第二個問題是關於開發人員招募。

Yeah, developer hiring has been the same. No changes.

是的，開發人員的招募一直都是一樣的。沒有變化。

Excellent. And then for Sid, there's been a lot of focus on coding assistance around generative AI. But as you're seeing the velocity of software development pick up, are you seeing more interest from customers around non-coding tools, but as they utilize AI within their coding process, such as securing binaries or more security around their software development life cycle? Would appreciate more color here. Thanks, Sid.

出色的。對 Sid 來說，重點關注的是生成式人工智慧的編碼輔助。但是，隨著您看到軟體開發速度的加快，您是否看到客戶對非編碼工具的興趣越來越大，但他們在編碼過程中使用了人工智慧，例如保護二進位檔案或圍繞軟體開發生命週期提供更多安全性？希望這裡能有更多的色彩。謝謝，Sid。

Yeah, thanks. And that's certainly how we see the market. The first market was AI code creation. We're really glad that the last Gartner AI Assistant Magic Quadrant rates us as the only non-hypercloud that's a leader in this space. The second phase is AI throughout the whole software life cycle. And I think for that, we're in a great spot with Duo Enterprise. And in the Gartner Magic Quadrant for DevOps that was released this morning, we're a leader with -- and we have the highest score for both execution and for vision. And I think that having the best and broadest platform enables us to win in this category.

是的，謝謝。這確實是我們看待市場的方式。第一個市場是AI程式碼創建。我們非常高興，最新的 Gartner AI 助理魔力像限將我們評為該領域唯一一家非超雲領導者。第二階段是貫穿軟體整個生命週期的AI。我認為，就這一點而言，我們與 Duo Enterprise 合作處於有利地位。在今天早上發布的 Gartner DevOps 魔力像限中，我們處於領先地位——並且在執行和願景方面都獲得了最高分。我認為擁有最好、最廣泛的平台使我們能夠在這一類別中獲勝。

And beyond that, you'll have more autonomous AI, AI going from reactive to proactive. And I'm really excited about what we're doing here. If you attended our GitLab 17 event, we showed GitLab Duo Workflow, an autonomous agent that can take more initiative of its own. That's where the puck is going, and we're skating towards it.

除此之外，你還將擁有更自主的人工智慧，從被動變成主動的人工智慧。我對我們在這裡所做的事情感到非常興奮。如果您參加了我們的 GitLab 17 活動，我們展示了 GitLab Duo Workflow，這是一個可以自行採取更多主動行動的自主代理。這就是冰球要去的地方，我們正向它滑去。

Thanks, guys.

謝謝大家。

Jason Ader, William Blair.

傑森·阿德，威廉·布萊爾。

All right. I guess there's a sense out there, Sid, that because of GitHub's faster growth rate recently that they're growing faster than you and taking share. I mean, I guess they are growing faster than you, but how do you think about the kind of share shift, if at all, in the market? Are users both doing better than everybody else, and that's the explanation? Or do you think GitHub could be taking some market share?

好的。我想大家有一種感覺，Sid，因為 GitHub 最近的成長速度很快，所以他們的成長速度比你快，而且佔據了更大的份額。我的意思是，我猜他們的成長速度比你快，但你如何看待市場上的份額轉移（如果有的話）？是不是兩個用戶的表現都比其他人好，這就是解釋？或者您認為 GitHub 可以佔據一些市場份額？

I think that customers are migrating to platforms and it's benefiting both of us. And I think we're early. You look at our revenue together, it's a small part of the $40 billion market. At the same time, there's -- the other thing is that they had a head start in AI co-creation. They purchased OpenAI and they had a head start.

我認為客戶正在遷移到平台，這對我們雙方都有好處。我覺得我們還早。如果你看一下我們的總收入，你會發現這只是 400 億美元市場的一小部分。同時，另一件事是他們在人工智慧共同創造方面已佔據先機。他們收購了 OpenAI，並佔據了先機。

Today, we're the only non-hypercloud that's a leader, according to Gartner. And that's because of two things, because you need a great model and you need a great context. We're vendor agnostic. Today, we use the best model on the market for Code Generation, Anthropic Claude 3.5.

根據 Gartner 的說法，今天我們是唯一一家非超級雲端領域的領導者。這是因為兩個原因，因為你需要一個偉大的模型和一個偉大的背景。我們與供應商無關。今天，我們使用市場上最好的程式碼產生模型，Anthropic Claude 3.5。

And context-wise, we know more of what a user is working on and what they've worked on in the past. Because we got the broadest platform, we have the most -- more context and better context leads to better AI answers. So together with that, we feel comfortable in competing.

從上下文來看，我們更了解用戶目前正在做什麼以及過去做過什麼。因為我們擁有最廣泛的平台，所以我們擁有最多的背景，更好的背景可以帶來更好的人工智慧答案。因此，我們在競爭中感到很安心。

Okay, great. And then, Brian, just a follow-up for you on net retention rate, NDRR (sic - NRR). It was 126%. It's continuing to come down. Do you think we bottomed? Where do you see NRR maybe exiting the year?

好的，太好了。然後，布萊恩，我想跟進一下淨留存率，NDRR（原文如此 - NRR）。達到了 126%。它還在繼續下降。您認為我們已經觸底了嗎？您認為 NRR 今年的業績會如何？

Thanks for the question. We're happy with where we landed in the quarter. I don't see anything with the dollar-based net retention rate that's a concern for me. Although our historical cohorts continue to steadily expand, the composition of our net dollar retention rate includes seats, which is an output for us.

謝謝你的提問。我們對本季的業績感到滿意。我不認為基於美元的淨留存率有什麼值得我擔心的。儘管我們的歷史隊列繼續穩步擴大，但我們的淨美元留存率的組成包括席位，這對我們來說是一項產出。

Just as a quick reminder, last year, we signed the largest deal in company history, which has been a tailwind for us for seats over the year. And we don't view a change in the ratio as a reflection of any recent developer hiring trends. As Sid said, this is a big market, very low penetration, and we have lots of room in front of us for growth.

簡單提醒一下，去年我們簽署了公司歷史上最大的一筆交易，這對我們全年的席位成長起到了推動作用。我們不認為該比例的變化反映了最近的開發人員招募趨勢。正如 Sid 所說，這是一個很大的市場，但滲透率很低，我們還有很大的成長空間。

Kash Rangan, Goldman Sachs.

高盛的卡什·蘭根 (Kash Rangan)。

Great. Thank you so much. Congrats on the beaten race quarter. I hope you enjoyed your summer. One for Sid, one for Brian. Sid, you talked about GitLab as being the unique company, that it's not a hyperscaler. You have these AI capabilities. Is that strength a key selling point in your end markets? I think we can all safely agree that your work just got going some seven to eight months ago. It's still too early in the AI race. So how do you think that, given that we're early and you're off to a good start, that this hyperscaler neutral positioning actually does give you an advantage in the long run?

偉大的。太感謝了。恭喜您贏得本季的比賽。我希望你度過了一個愉快的夏天。一個給 Sid，一個給 Brian。Sid，您說過 GitLab 是一家獨特的公司，它不是一家超大規模企業。你擁有這些人工智慧能力。這種優勢是否是你們在終端市場的關鍵賣點？我想我們都可以肯定地同意，你的工作大約在七到八個月前才開始。人工智慧競賽還為時過早。那麼，考慮到我們處於早期階段並且您已經有了良好的開端，您認為這種超大規模中立定位實際上是否會為您帶來長期優勢？

And then one for you, Brian. I think you talked about a smaller percentage of the growth rate, NER, coming from price increases. Can you help us understand what is ahead for the company, especially as you go through the motions of renewals from the other pricing tiers and the subsequent price increases that could help your contribution to the growth rate get even better? Thank you so much.

接下來還有一篇給你，布萊恩。我認為您談到的成長率（NER）中較小比例的部分來自於價格上漲。您能否幫助我們了解公司未來的發展方向，尤其是當您經歷從其他定價層級的續約以及隨後的價格上漲時，這可能會幫助您對成長率的貢獻變得更好？太感謝了。

Yeah. Kash, thank you for that question. Being hyperscaler, independent, being vendor agnostic, it's good, but it's not enough. The key reasons we win against Microsoft GitHub are that we have, first of all, the most comprehensive platform. When the platform can do more, our customers can get a 10 times faster cycle time. The second reason is we have the best security. We allow our customers to shift security left, always do security, and prove that they've done it. And the third reason is that we really listen to our customers. We're the only vendor with a single tenant SaaS solution, GitLab Dedicated, that's growing really, really fast.

是的。Kash，謝謝你的提問。成為超大規模、獨立、與供應商無關的企業，這很好，但還不夠。我們擊敗微軟 GitHub 的關鍵原因首先是我們擁有最全面的平台。當平台能夠完成更多功能時，我們的客戶可以獲得快 10 倍的週期時間。第二個原因是我們擁有最好的安全保障。我們允許我們的客戶將安全左移，始終進行安全保護，並證明他們已經做到了這一點。第三個原因是我們真正傾聽客戶的意見。我們是唯一擁有單一租戶 SaaS 解決方案的供應商，GitLab Dedicated 的成長速度非常非常快。

And we're getting the acknowledgement that we're executing well. Today, leader in the MQ for DevOps. We're all the way to the top, all the way to the right. And that's enabling our customers to get the biggest return, 482% ROI, according to Forrester. And the interesting thing is at 60% more than the previous studies, so the benefits of being on the broadest platform are increasing. Our customers are better and better off over time being on GitLab.

我們的表現也得到了認可。如今，DevOps MQ 中的領導者。我們一直到頂部，一直到右邊。根據 Forrester 的調查，這使我們的客戶獲得了最大的回報，即 482% 的投資回報率。有趣的是，這一數字比先前的研究高出了 60%，因此，在最廣泛的平台上獲益正在增加。隨著時間的推移，我們的客戶使用 GitLab 的狀況越來越好。

Thanks, Sid. I'll touch on the price increase real quickly, Kash. And so there's a couple of things to touch on there. One is, we do break out quarterly and the dollar-based net retention, what impact is related to seats, price, which has also increased customer yield, as well as tier upgrades. And what I really care about when I look at that is the positive unit economics are growing really nicely in the business. I talked about this in the Q1 call, and I'm happy with how we're doing there.

謝謝，Sid。卡什，我很快就會談到價格上漲的問題。這裡有幾件事要談一下。一是，我們確實公佈了季度和基於美元的淨留存數據，這與座位數和價格有什麼影響，這也增加了客戶收益以及層級升級。當我看到這一點時，我真正關心的是積極的單位經濟學在業務中確實成長良好。我在第一季電話會議上談到了這一點，我對我們的表現感到滿意。

And then going forward, we expect the price increase to continue to layer in over time as we cycle through the renewal portfolio. And so we're partially there, but we'll see impacts throughout this year and next year related to that. And so overall, I'm really pleased with the returns that we're seeing from the price increase and the consistently improving unit economics, which shows the value that the customers are getting from the platform and is part of the total economic study increasing as well.

展望未來，我們預計隨著我們不斷更新產品組合，價格將持續上漲。因此，我們部分已經達到了這一水平，但今年和明年我們將看到與此相關的影響。因此，總體而言，我對我們從價格上漲和持續改善的單位經濟中看到的回報感到非常滿意，這表明客戶從平台獲得了價值，也是總體經濟研究成長的一部分。

Karl Keirstead, UBS.

瑞銀的卡爾·科爾斯特德（Karl Keirstead）

Okay, great. Maybe on the AI side, Sid, those anecdotes about Barclays, F5, and KeyBank are pretty powerful. Maybe a two-parter for you. Are any of those customers, or maybe some of your other early Duo wins, customers that were using GitHub Copilot and now that Duo has closed a lot of the functionality gaps, are moving off of Microsoft?

好的，太好了。也許在人工智慧方面，Sid，關於巴克萊、F5 和 KeyBank 的軼事非常有力。對你來說也許可以分成兩個部分。這些客戶，或者可能是您早期贏得的 Duo 的客戶中是否有人正在脫離微軟？

And then I guess the second question, Sid, I could be wrong on this, but my understanding was that Duo, at least initially, was quite rooted in Google's LLMs. And I'm wondering if this change, assuming it is, to Anthropic's Claude model had a marked improvement in the functionality of that Code Gen tool. Thank you.

然後我想問第二個問題，Sid，我可能錯了，但我的理解是，Duo 至少在最初階段，深深植根於谷歌的法學碩士學位。我想知道，如果對 Anthropic 的 Claude 模型進行這樣的改變，是否會顯著改善該 Code Gen 工具的功能。謝謝。

Yeah, thanks for that, Karl. If you look at customers evaluating Duo Pro and Duo Enterprise, I think there's nearly zero customers that don't know that Copilot exists. And for most of these customers, they've had pockets of use and they do commonly a head-to-head comparison. So we're winning against Copilot.

是的，謝謝你，卡爾。如果您觀察評價 Duo Pro 和 Duo Enterprise 的客戶，我認為幾乎沒有客戶不知道 Copilot 的存在。對於大多數客戶來說，他們都有使用過的產品，並且通常會進行面對面的比較。因此我們戰勝了 Copilot。

Regarding the best model for the task, we're still using some of Google's models, but for Code Generation, the most kind of vivid application. Right now, the best model on the planet is Anthropic Claude 3.5. You look on the internet, that's consensus. We found the same in our testing, and we can switch to that. We're not beholden to using any hypercloud's product. We haven't bought an AI company or an AI foundational modeling company. And being able to use the latest and greatest is a great advantage because it gives our customers better code.

關於最適合該任務的模型，我們仍在使用一些 Google 的模型，但對於程式碼生成，這是最生動的應用。目前，地球上最好的模型是 Anthropic Claude 3.5。你在網路上看看，這就是共識。我們在測試中發現了同樣的情況，我們可以切換到那一點。我們並沒有強制使用任何超級雲端的產品。我們還沒有收購一家人工智慧公司或一家人工智慧基礎建模公司。能夠使用最新和最好的程式碼是一個巨大的優勢，因為它為我們的客戶提供了更好的程式碼。

Okay, terrific. Thanks, and congrats.

好的，太棒了。謝謝，恭喜。

Rob Owens, Piper Sandler.

羅伯歐文斯、派珀桑德勒。

Great. Thanks for taking my question. Brian, realizing there's a lot of puts and takes around the revenue growth number, just what changes to your SSP, I guess I'll have two questions kind of masked in one. Number one, did that come in relative to your expectations, and is there no change to the year if we think about the headwind that you called out on last quarter?

偉大的。感謝您回答我的問題。布萊恩，我意識到收入成長數字存在著許多不確定性，您的 SSP 究竟會發生哪些變化，我想我會把兩個問題概括起來問。第一，這是否符合您的預期？

And I guess secondarily, looking at a lot of the leading indicators, RPO up quarter over quarter, CRPO up quarter over quarter, I think one of the best growth results you've seen in the last year. Can you speak to large deal activity right now, the upsell versus new customer commitments and just what you guys are seeing on that front? Thanks.

其次，我想，從許多領先指標來看，RPO 環比成長，CRPO 環比成長，我認為這是去年看到的最好的成長結果之一。您能談談目前的大型交易活動、追加銷售與新客戶承諾的情況以及您在這方面看到的情況嗎？謝謝。

Yeah, absolutely. Thanks for the question, Rob. When you look at sort of the SSP -- and for everybody, that's just accounting on how it gets allocated; it doesn't impact the overall deal itself -- we actually took a little bit of a hit this quarter related to SSP that we talked about on the last call. And so that was assumed in the numbers.

是的，絕對是如此。謝謝你的提問，羅布。當你查看 SSP 時——對於所有人來說，這只是關於其如何分配的核算；它不會影響整個交易本身——事實上，我們本季受到了上次電話會議上討論的 SSP 相關的一點打擊。所以這是在數字中假設的。

We're doing well across the business. First order was doing well. Expansion was doing well. We had the best quarter in churning contraction in the last eight quarters. And so on prior calls, I said that I thought that second quarter would have run through most of the contracts. And churning contraction was better than in eight quarters.

我們的整個業務表現都很好。第一批訂單進展順利。擴張進展順利。我們經歷了過去八個季度中客戶流失收縮最好的一個季度。所以在之前的電話會議中，我說我認為第二季將會完成大部分合約。且流失率收縮情況比前八個季度有所改善。

The other thing that we saw is we actually saw a real strength in the Enterprise, greater than 100k customers growing over 30% year over year, and we saw a lot of expansions and lands in that area as well. And I think it just goes to some of the things that Sid said previously around time to value, positive business outcomes, ROI, consolidating a toolchain onto a single platform has just created a lot of benefit. And we're seeing the positives of that sort of show up in the model.

我們看到的另一件事是，我們實際上看到了企業的真正實力，超過 10 萬名客戶同比增長超過 30％，而且我們也看到該領域有大量擴張和土地。我認為這涉及到 Sid 之前所說的一些內容，例如價值實現時間、積極的業務成果、投資回報率，將工具鏈整合到單一平台上已經創造了許多好處。我們在模型中看到了這種表現的正面意義。

Thank you.

謝謝。

Michael Turrin, Wells Fargo.

富國銀行的邁克爾·圖林。

Hey. Great. Thanks, Kelsey. Appreciate you taking the question to your team. I guess just that I'll ask both parts up front, but to some of the prior question, the leading indicators here all look pretty good. And it's not something we're seeing a lot of across software, CRPO billings, bookings up more than 40%.

嘿。偉大的。謝謝，凱爾西。感謝您向您的團隊提出這個問題。我想我會先問這兩個部分，但對於之前的一些問題，這裡的領先指標看起來都相當不錯。我們在軟體、CRPO 帳單、預訂量成長超過 40% 的情況下，這種情況並不常見。

Just hoping you can unpack that strength a bit more if there's anything more one time in nature for us to be mindful of in those metrics. And then, Brian, maybe walk us through the assumptions you're embedding in forecasts for the rest of the year on the back of that Q2 strength across macro, seats, sales execution, or anything else worth mentioning for us. Thank you.

只是希望您能夠進一步解釋這種力量，看看自然界中是否存在更多需要我們在這些指標中註意的事情。然後，布萊恩，也許您可以向我們介紹一下，基於第二季度宏觀、席位、銷售執行或其他任何值得一提的因素，您對今年剩餘時間的預測中嵌入了哪些假設。謝謝。

Appreciate the two questions. On the forecast and how we're running the business and the transition of actually taking over the interim CRO role, there's really been no changes whatsoever. And Ashley has been really plugged into the entire sales force and visiting customers and worked really closely with Chris. So we're happy about the minimal disruption there with Chris's departure.

感謝這兩個問題。關於預測以及我們如何經營業務以及實際接管臨時 CRO 角色的過渡，實際上沒有任何變化。阿什利 (Ashley) 一直深入整個銷售團隊並拜訪客戶，與克里斯 (Chris) 密切合作。因此，我們很高興看到克里斯的離開只給這裡帶來了很小的干擾。

All the metrics that you talked about -- CRPO grown 42% year over year, short-term calculated billings from 40% year over year, Ultimate ARR, now 47% of total ARR and was greater than 50% of bookings within the quarter -- and that really coming from the Enterprise base that we have. And so our deals are getting larger, Ultimate adoptions increasing. Ultimate is good for us and our customers, and I think you see that in the metrics itself. And then we also saw Dedicated. Dedicated grew roughly 150% year over year. SaaS was great. So we're just seeing strength across the board from our customers who are adopting the platform.

您談到的所有指標 - CRPO 同比增長 42%、短期計算賬單同比增長 40%、最終 ARR 現在佔總 ARR 的 47% 並且佔本季度預訂量的 50% 以上 - 而這些實際上都來自我們擁有的企業基礎。因此，我們的交易規模越來越大，最終採用的數量也不斷增加。Ultimate 對我們和我們的客戶都有好處，我認為您可以從指標本身中看到這一點。然後我們還看到了 Dedicated。專用量年增約 150%。SaaS 很棒。我們看到採用該平台的客戶全面展現強勁實力。

Great job. Thank you.

做得好。謝謝。

Pinjalim Bora, JPMorgan.

摩根大通的 Pinjalim Bora。

Great. Hey. Congrats on the quarter. Thanks for taking the questions. I want to ask you, Sid, obviously it seems like we are seeing some of the AI adoption, but any way to further quantify it in terms of maybe the portion of the customer base or users that are touching the product today and what portion of the codes that is being written by the AI is being committed? Any further kind of quantitative metrics?

偉大的。嘿。恭喜本季取得佳績。感謝您回答這些問題。我想問你，Sid，顯然我們似乎看到了一些人工智慧的採用，但有沒有辦法進一步量化它，例如今天接觸產品的客戶群或用戶的比例，以及人工智慧編寫的程式碼中有多少部分被提交了？還有其它種類的量化指標嗎？

Yeah, thanks for that. It's still early for us. You heard from Barclays, F5, Noah, but if you look at the quotes going on, not all of them have AI in it. And if they have AI, they typically don't have AI for all the users of our customer yet. So customers are early in this adoption cycle. We cater to the enterprise. They are relatively slower to adopt this and more considerate.

是的，謝謝。對我們來說還很早。您聽說過巴克萊、F5、Noah 的消息，但如果您看看這些公司的報價，您會發現並不是所有公司都涉及人工智慧。即使他們有人工智慧，他們通常還沒有為我們客戶的所有用戶提供人工智慧。因此，客戶仍處於此採用週期的早期階段。我們服務企業。他們接受這一點的速度相對較慢，而且更加體貼。

It's really important that we listen well to these customers. For example, we are accelerating our work on offline models because our customers are requesting that. We see a giant opportunity for AI throughout the life cycle. So our GitLab Ultimate customers, for them, we think that Duo Enterprise is a really, really good value proposition, and that's a big focus of ours.

認真傾聽這些客戶的意見確實非常重要。例如，我們正在加快離線模型的開發工作，因為我們的客戶有這樣的要求。我們在整個生命週期中看到了人工智慧的巨大機會。因此，對於我們的 GitLab Ultimate 客戶來說，我們認為 Duo Enterprise 是一個非常非常好的價值主張，這也是我們關注的重點。

Yeah, understood. One follow-up for Brian. Brian, what portion of the contracts from existing customers at this point are completely through kind of the first phase of your pricing change of existing customers, the 19 to 24? Is it fair to assume that it's largely complete by this point? And the second phase of the rollout, which I believe started in end of April, if I'm not wrong, did it have any impact in billings RPO performance this quarter?

嗯，明白了。對 Brian 的一次後續行動。布萊恩，目前現有客戶的合約中有多少部分已經完全完成了現有客戶價格調整的第一階段，即 19 至 24 日？到目前為止，可以合理地認為它基本上已經完成了嗎？我認為第二階段的推廣是在 4 月底開始的，如果我沒有記錯的話，它對本季的 Billings RPO 績效有什麼影響嗎？

Yeah. So on the price increase impact overall, remember, we weren't allowing people to early renew. And so it really comes up when their renewal comes up and they would go 19 to 24, then 24 to 29. And then new customers would go straight to 29.

是的。因此，就整體價格上漲的影響而言，請記住，我們不允許人們提前續約。所以當他們的續約到來時，這個問題真的出現了，他們會從 19 到 24，然後從 24 到 29。然後新客戶就會直接轉到 29。

And so I think it's really important that the price increase will continue to layer in over time as we cycle through the renewal portfolio. And so I expect next year to continue to see the benefit of that. And it really points to the unit economics improving based on what we're delivering to our customers.

因此，我認為，隨著我們不斷更新產品組合，價格上漲將隨著時間的推移而持續遞增，這一點非常重要。因此我預計明年將繼續看到其好處。這確實表明，根據我們向客戶提供的產品，單位經濟效益有所改善。

Got it. Thank you.

知道了。謝謝。

(Event Instructions) Gregg Moskowitz, Mizuho.

（活動說明） 瑞穗的 Gregg Moskowitz。

Okay, great. Thank you, Kelsey, and congrats on a very good performance. I wanted to follow up on Ultimate because the percentage of ARR continues to expand nicely. And Sid, you called out some very good success among your largest customers. Despite the much higher price point that exists for Ultimate, as compared with Premium, what I'm curious about is if you're finding that you're landing more frequently with Ultimate versus what you were seeing 6 to 12 months ago. Thanks.

好的，太好了。謝謝你，凱爾西，恭喜你出色的表現。我想跟進 Ultimate，因為 ARR 的百分比繼續順利擴大。而且 Sid，您指出您的最大客戶取得了非常好的成功。儘管與 Premium 相比，Ultimate 的價格要高得多，但我很好奇的是，您是否發現與 6 到 12 個月前相比，您使用 Ultimate 的頻率更高了。謝謝。

Thanks for that. We're certainly changing our approach to leading with Ultimate when we approach a new customer. Because the more comprehensiveness is driving so much value because this integrated security is driving so much value, we've started to lead with Ultimate. Just talked about the return, like the 482% return that is for GitLab Ultimate.

謝謝。當我們接觸新客戶時，我們肯定會改變以 Ultimate 為主導的方法。因為更全面性帶來瞭如此大的價值，因為這種整合的安全性帶來瞭如此大的價值，所以我們已經開始以 Ultimate 為主導。剛剛談到了回報，例如 GitLab Ultimate 的回報為 482%。

So maybe counterintuitively, our most expensive product is the one you get the biggest return on, because the return is not coming so much from paying us less, but it's about deprecating all those existing point solutions. We talked about Lockheed Martin being able to deprecate our legacy CI vendor. It's either not just on software, not just on cycle time, not just on efficiency, even on hardware costs. So being able to consolidate is the big winner, and Ultimate can replace the most point solutions, and that's why we're leading with it.

因此，也許有悖常理的是，我們最昂貴的產品是您獲得最大回報的產品，因為回報並非來自於支付給我們更少的費用，而是貶低所有現有的點解決方案。我們討論了洛克希德馬丁公司是否能夠棄用我們傳統的 CI 供應商。這不僅涉及軟體，不僅涉及週期時間，不僅涉及效率，甚至還涉及硬體成本。因此，能夠整合才是最大的贏家，而 Ultimate 可以取代大多數點解決方案，這就是我們以此為主導的原因。

Great. Thank you.

偉大的。謝謝。

Matt Hedberg, RBC.

馬特·赫德伯格（Matt Hedberg），RBC。

Great. Thanks, Kelsey. I'll offer my congrats as well. I guess for either of you, regarding GenAI, it seems like you guys are increasingly fitting into that work stream for a customer when they're thinking through their own GenAI adoption. I'm just sort of curious, when you're having conversations with customers, how important is GenAI to them? I think we've all been sort of like, it's the first and last question I think everybody asks, but when you're talking to the customers, how important is that in their software development life cycle right now? And I guess, how do they see GitLab fitting into that ecosystem that they're all developing?

偉大的。謝謝，凱爾西。我也會表示祝賀。我想，對於你們任何一個人而言，關於 GenAI，當客戶考慮他們自己是否採用 GenAI 時，你們似乎越來越適應客戶的工作流程。我只是有點好奇，當您與客戶交談時，GenAI 對他們有多重要？我想我們都曾有過類似的感覺，這是每個人都會問的第一個也是最後一個問題，但是當你與客戶交談時，這在他們目前的軟體開發生命週期中有多重要？我想，他們如何看待 GitLab 融入他們正在開發的生態系統？

Yeah, it's really important to our customers. We had a bunch of caveats why all of them are rushing to implement it to 100% of their users. First of all, they want it secure. Second of all, they want it to work for the people who are working on existing applications. A lot of the demos you see out there, they're for new applications. Most developers in enterprises are working on giant existing applications.

是的，這對我們的客戶來說確實很重要。我們對為什麼他們都急於將其實施給 100％ 的用戶提出了一系列警告。首先，他們希望它是安全的。其次，他們希望它能夠為那些從事現有應用程式開發的人所服務。您所看到的許多演示都是針對新應用程式的。企業中的大多數開發人員都在開發大型現有應用程式。

So making that work well is really important. We've got a project internally called Davinci to make AI work even better for those existing applications. That's super, super important. It has to meet all the security requirements that the customer has, and they want to see an actual return. And most of the time, that return comes not when it's just for the coding, but when it's throughout.

因此，做好這項工作確實非常重要。我們內部有一個名為 Davinci 的項目，旨在讓 AI 更好地為現有應用程式服務。這非常非常重要。它必須滿足客戶的所有安全要求，並且他們希望看到實際的回報。而且大多數時候，這種回報並不是只出現在編碼過程中，而是貫穿整個過程。

So those are the considerations we run into. They want vendors who can deliver on that, who have a great vision going forward, but they're not jumping in to 100% of the people for just a coding solution today.

這些就是我們遇到的考慮因素。他們想要能夠實現這一目標、擁有偉大未來願景的供應商，但他們不會僅僅為了今天的編碼解決方案而向 100% 的人提供解決方案。

Got it. Thanks, Sid.

知道了。謝謝，Sid。

Okay. So we'll go on to Mike from Needham.

好的。接下來我們繼續採訪來自尼德姆的麥克。

Hey, team. Thank you for getting me on. Just wanted to circle up, and I appreciate you guys continuing to give us the composition of the DBNRR. If I could just focus on the seats contributing about 40% this quarter, wanted to get a sense, first, how that compared versus your internal expectations.

嘿，團隊。感謝您邀請我加入。只是想繞一圈，我感謝你們繼續向我們提供 DBNRR 的組成。如果我只能關注本季度貢獻率約 40% 的席位，那麼首先我想了解一下，這與您的內部預期相比如何。

And then, secondly, can you help us think about -- is the sales force indexing more potentially toward pricing, given the changes to the different packages you have out there in the market? Or any other color there as well to help us get a better sense of how the DBNRR flows from one quarter to the next. Thank you.

其次，您能否幫助我們思考一下—考慮到市場上不同套餐的變化，銷售人員是否會更有可能根據定價進行調整？或任何其他顏色，以幫助我們更好地了解 DBNRR 從一個季度流向下一個季度的方式。謝謝。

Yeah, absolutely. Thanks for the question. The dollar-based net retention, with the seat fluctuation based on the largest deal that we had last year, we knew that was going to be different this quarter than previous quarters. It's also an output. And so as we go in and solution sell, we're trying to figure out what the best solution is for the customer, and then we'll land there.

是的，絕對是如此。謝謝你的提問。以美元計算的淨保留率，以及基於去年最大交易的席位波動，我們知道本季的情況將與前幾季有所不同。這也是一種輸出。因此，當我們進入並銷售解決方案時，我們會嘗試找出對客戶最好的解決方案，然後我們就會找到解決方案。

The fact that we're landing larger and landing more on Ultimate, that's good news, and we'll have an impact on dollar-based net retention rate. But there's nothing that I saw within the quarter that caused any concerns as it relates specifically to seats as that component of the dollar-based net retention rate.

事實上，我們的目標客戶群越來越大，而且在 Ultimate 上的客戶越來越多，這是個好消息，我們將對基於美元的淨留存率產生影響。但我在本季度沒有看到任何引起擔憂的事情，因為它與作為美元淨保留率組成部分的席位具體相關。

Thank you very much.

非常感謝。

Peter Weed, Bernstein.

彼得威德、伯恩斯坦。

Thank you very much, and congrats on the continued momentum. I think if I'm doing my back of the envelope properly, it looks like you are anticipating a nice acceleration in quarter four implied in the numbers. How should we think about that acceleration relative to what appears to be a little bit of a deceleration in your guidance for quarter three?

非常感謝，並祝賀您繼續保持良好的發展勢頭。我認為，如果我正確地做了統計，那麼看起來你正在預期第四季度的數字暗示將會出現良好的加速。相對於您對第三季的預測中出現的一點點減速，我們該如何看待這種加速？

Thanks, Peter. As always, I appreciate the question. When I look back at sort of last year and look at when we reported Q2 and what we guided for Q3 and Q4, they're somewhat similar. And so I didn't see any sort of sequential or year-over-year changes that jumped out to me to be surprising. Fourth quarter, historically, has always been the strongest quarter in the company. Q2 and Q3 have relatively been the same, and Q1 seasonally has been a little weak.

謝謝，彼得。一如既往，我很感謝你提出這個問題。當我回顧去年並看看我們報告的第二季業績以及我們對第三季和第四季的預期時，我發現它們有些相似。因此，我並沒有看到任何令我感到驚訝的連續變化或同比變化。從歷史上看，第四季一直是該公司業績最強勁的季度。Q2 和 Q3 相對持平，而 Q1 季節性表現略顯疲軟。

We continue to guide to strong top-line growth rates and improving non-GAAP operating margins. The thing I personally like about the business model is we have a lot of visibility heading into any given quarter, given the ratable nature of the business. And as we continue to scale, we're seeing efficiencies at scale in the business, which is great as well. And so I'm pleased with the guidance we've provided this afternoon for the third quarter and for the full year.

我們將繼續引領強勁的營收成長率和不斷提高的非公認會計準則營業利潤率。我個人喜歡這種商業模式的原因是，考慮到業務的可評估性質，我們對任何特定季度都具有很高的可預見性。隨著我們不斷擴大規模，我們看到業務規模效率的提高，這也很好。因此，我對我們今天下午提供的第三季和全年業績指引感到滿意。

Thank you.

謝謝。

George McGreehan, Bank of America.

美國銀行的喬治·麥格里漢。

Hey. All right. Hey, Sid, Bryan, appreciate you guys taking the question. It's George McGreehan on for Koji Ikeda. I wanted to say congrats on Duo Enterprise going GA. The list of features highlighted on the press release were very impressive, and I kind of wanted to dig in more specifically about how you're selling the product to the install base. Can you talk a bit about the strategy there and in terms of any changes to sales incentives? This could kind of get a sense of how to think about potential adoption rates over the next several quarters.

嘿。好的。嘿，Sid、Bryan，感謝你們回答這個問題。喬治‧麥格里漢 (George McGreehan) 取代池田浩二 (Koji Ikeda)。我想對 Duo Enterprise 正式上市表示祝賀。新聞稿中重點介紹的功能清單非常令人印象深刻，我想更具體地了解您如何向安裝基礎銷售產品。您能否談談那裡的策略以及銷售激勵措施的任何變化？這或許能讓我們了解如何考慮未來幾季的潛在採用率。

Yeah, thanks for that. The main audience for Duo Enterprise is existing GitLab Ultimate customers. And Ultimate is typically our larger companies, and they are served by our sales force, so it's a direct motion. Ultimate is also our fastest-growing SKU. If you think about it, if you look at the cohort of a $100,000 plus, two-thirds of the net new ARR went into Ultimate.

是的，謝謝。Duo Enterprise 的主要受眾是現有的 GitLab Ultimate 客戶。Ultimate 通常是我們的較大公司，由我們的銷售團隊為他們提供服務，因此這是一種直接動議。Ultimate 也是我們成長最快的 SKU。如果你仔細想想，如果你看一下10萬美元以上的群體，你會發現淨新ARR的三分之二都進入了Ultimate。

So that's Enterprise growth on top of that. And we're talking to the customer because there's lots of considerations implementing it. That's, for example, why we're working on an offline version. That's why we have tons of features to control who's using it, when they are using it. And we're selling it because of the improvement in productivity, not just for coding, not just for devs, but also for their security people and their operations people.

這就是企業在此基礎上的成長。我們正在與客戶溝通，因為實施方案需要考慮許多因素。例如，這就是我們開發離線版本的原因。這就是為什麼我們擁有大量的功能來控制誰在使用它以及何時使用它。我們銷售它是因為它提高了生產率，不僅適用於編碼，不僅適用於開發人員，還適用於他們的安全人員和營運人員。

Nick Altmann, Scotiabank.

加拿大豐業銀行的尼克‧阿爾特曼 (Nick Altmann)。

Thanks, guys. Just a quick one for me. Last quarter, you guys talked about Duo going to be sort of a bigger needle mover next year. You guys did rattle off a handful of marquee wins. And so I guess, when you think about the second-half pipeline, as you sort of get more customers on Duo, have more referenceable logos, what are you anticipating from the second half of the year in terms of Duo contribution versus, say, a quarter ago? Thanks.

謝謝大家。對我來說這只是一個快速步驟。上個季度，你們談到 Duo 明年將成為更大的推動者。你們確實取得了一些重大勝利。所以我想，當您考慮下半年的計劃時，隨著您在 Duo 上獲得更多客戶，擁有更多可參考的標識，與上個季度相比，您對下半年 Duo 的貢獻有何預期？謝謝。

Thanks for the question. I'll answer directly, and I'll give some more context on how I think about it. So we really expect AI to start contributing to the model in FY 2026 and beyond. First, from just a practical perspective, there was sort of a big media hype cycle, and people are just partially adopting it now. So I think right now, we're starting to see customers trying to figure out how to implement AI safely and compliantly within their organizations. And we're starting to see that with what we gave you from a reference perspective.

謝謝你的提問。我會直接回答，並提供一些我對此的看法。因此，我們確實希望人工智慧能夠在 2026 財年及以後開始為該模型做出貢獻。首先，從實際角度來看，媒體曾經有過一次大炒作，而現在人們只是部分地接受它。所以我認為現在，我們開始看到客戶試圖弄清楚如何在他們的組織內安全、合規地實施人工智慧。我們從參考角度向您提供的內容開始看到這一點。

And also, just from a mechanics perspective, and I know you know this, but it makes sense to say it is, from a new product perspective, for it to have an impact at a company that's $700 million plus in run rate revenue growing 30%, it's just going to take a little while to sort of build that. From the AI contribution in 2Q, I will say that we're three times our planned number, so we did a lot better than what we expected internally. And so from a model perspective, long term, I just view this as good news because this is going to be a long-term growth driver for the business.

而且，僅從機制的角度來看，我知道您知道這一點，但從新產品的角度來看，要對一家營業額超過 7 億美元、營業收入增長 30% 的公司產生影響，需要一點時間來建立這種影響。從第二季的人工智慧貢獻來看，我可以說我們是計畫數字的三倍，所以我們的表現比內部預期的要好得多。因此，從模型角度來看，從長遠來看，我認為這是個好消息，因為這將成為業務的長期成長動力。

Thanks, Brian.

謝謝，布萊恩。

Kingsley Crane, Canaccord.

金斯利‧克蘭 (Kingsley Crane)，Canaccord。

Great. It sounds like you've had some nice wall-to-wall Duo deployments with a couple banks in F5. Should we continue to think about high user penetration, but a smaller total number of customers? Or are you seeing some green shoots in terms of some bottoms-up adoption with Duo? Thanks.

偉大的。聽起來您已經與 F5 中的幾家銀行進行了一些不錯的 Duo 部署。我們是否應該繼續考慮高用戶滲透率，但減少客戶總數？或者您看到了 Duo 自下而上採用方面的一些積極跡象？謝謝。

We're seeing some bottoms-up adoption as well. I think if you look across our entire customer base, the more common scenario is that they buy Duo for part of their users. For various reasons, they're not yet ready to put all the users in it.

我們也看到了一些自下而上的採用。我認為，如果你縱觀我們的整個客戶群，更常見的情況是他們為部分用戶購買 Duo。由於種種原因，他們還沒有準備好讓所有用戶都參與其中。

We think that will come over time, the returns are there, but that's a more common scenario. Of course, we also have these great customers that do it wall to wall immediately. And if we meet all the requirements, we can do that, and that has our preference, but the partial scenario is more common.

我們認為隨著時間的推移，這種情況會發生，回報就會存在，但這是一種更常見的情況。當然，我們也有一些很棒的客戶，他們立即就這麼做了。如果我們滿足所有要求，我們就可以這樣做，這是我們的偏好，但部分情況更為常見。

Great. Thanks, Sid and Brian.

偉大的。謝謝，Sid 和 Brian。

So this concludes our second-quarter call. Thank you very much for joining us, and we look forward to seeing many of you over the coming quarter. Have a great evening.

我們的第二季電話會議到此結束。非常感謝您的加入我們，我們期待在接下來的一個季度見到你們。祝您有個愉快的夜晚。