Atlassian Corp (TEAM) 2026 Q4 法說會逐字稿

Good afternoon, and thank you for joining Atlassian's earnings conference call for the fourth quarter of fiscal year 2026. As a reminder, this conference call is being recorded and will be available for replay on the Investor Relations section of Atlassian's website following this call.

下午好，感謝各位參加 Atlassian 2026 會計年度第四季財報電話會議。提醒各位，本次電話會議將被錄音，會後可於 Atlassian 官網「投資人關係」專區收聽重播。

I will now hand the call over to Martin Lam, Atlassian's Head of Investor Relations.

接下來我將把電話交給 Atlassian 投資人關係主管 Martin Lam。

Welcome to Atlassian's fourth quarter fiscal year 2026 earnings call. Thank you for joining us today. On the call with me today, we have Atlassian's CEO and Co-Founder, Mike Cannon-Brookes; and Chief Financial Officer, James Chuong.

歡迎參加 Atlassian 2026 會計年度第四季財報電話會議。感謝各位今天與會。今天與我一同出席的有 Atlassian 執行長暨共同創辦人 Mike Cannon-Brookes，以及財務長 James Chuong。

Earlier today, we published a shareholder letter and press release with our financial results and commentary for our fourth quarter of fiscal year 2026. The shareholder letter is available on the Investor Relations section of our website, where you will also find other earnings-related materials, including the earnings press release and supplemental investor data sheet. As always, our shareholder letter contains management's insight and commentary for the quarter. So during the call today, we'll have a brief opening remarks and then focus our time on Q&A.

今天稍早，我們已發布股東信與新聞稿，說明 2026 會計年度第四季的財務結果與評論。股東信可於我們官網的「投資人關係」專區取得，您也可在該處找到其他與財報相關的資料，包括財報新聞稿與補充投資人資料表。一如往常，我們的股東信包含管理團隊對本季的洞見與評論。因此，今天的電話會議我們將先做簡短開場，接著把時間重點放在問答。

This call will include forward-looking statements. Forward-looking statements involve known and unknown risks, uncertainties and assumptions. If any such risks or uncertainties materialize or if any of the assumptions prove incorrect, our results could differ materially from the results expressed or implied by the forward-looking statements we make.

本次電話會議將包含前瞻性陳述。前瞻性陳述涉及已知與未知的風險、不確定性與假設。若任何此類風險或不確定性發生，或任何假設被證明不正確，我們的結果可能與我們所作前瞻性陳述中明示或暗示的結果存在重大差異。

You should not rely upon forward-looking statements as predictions of future events. Forward-looking statements represent our management's beliefs and assumptions only as of the date such statements are made, and we undertake no obligation to update or revise such statements should they change or cease to be current.

您不應將前瞻性陳述視為對未來事件的預測。前瞻性陳述僅代表管理團隊在陳述作出當日的信念與假設；若其後有所變動或不再適用，我們不承擔更新或修訂該等陳述的義務。

Further information on these and other factors that could affect our business performance and financial results is included in filings we make with the Securities and Exchange Commission from time to time, including the section titled Risk Factors and our most recently filed annual and quarterly reports.

關於可能影響我們業務表現與財務結果的上述及其他因素之進一步資訊，載於我們不時向美國證券交易委員會（SEC）提交的文件中，包括「風險因素」章節，以及我們最近提交的年度與季度報告。

During today's call, we will also discuss non-GAAP financial metrics. These non-GAAP financial measures are in addition to and are not a substitute for or superior to measures of financial performance prepared in accordance with GAAP. A reconciliation between GAAP and non-GAAP financial measures is available in our shareholder letter, earnings release and investor data sheet on the Investor Relations section of our website.

在今天的電話會議中，我們也將討論非 GAAP 財務指標。這些非 GAAP 財務衡量指標是 GAAP 編製之財務績效衡量指標的補充，並非其替代，亦不優於 GAAP 指標。GAAP 與非 GAAP 財務衡量指標之間的調節表，已載於我們的股東信、財報新聞稿與投資人資料表，皆可於官網「投資人關係」專區取得。

We'd like to allow as many of you to participate in Q&A as possible. Out of respect for others on the call, we'll take one question at a time.

我們希望讓盡可能多的各位參與問答。為尊重其他與會者，我們將一次回答一個問題。

With that, I'll turn the call over to Mike for opening remarks.

接下來我把電話交給 Mike 進行開場致詞。

Thank you all for joining us today. As you've already read in our shareholder letter, Q4 was another outstanding quarter, capping off an incredible fiscal year for Atlassian. Total revenue grew 28% to $1.8 billion. Cloud revenue surged to $1.2 billion as growth accelerated again, up 31% year-over-year. RPO grew 44% year-over-year to $4.8 billion, and we delivered a GAAP operating margin of 12%.

感謝各位今天加入我們。如各位已在股東信中讀到的，第四季再度交出亮眼成績，為 Atlassian 非凡的會計年度畫下句點。總營收成長 28%，達 18 億美元。雲端營收躍升至 12 億美元，成長再度加速，年增 31%。RPO 年增 44% 至 48 億美元，我們也達成 12% 的 GAAP 營業利益率。

We are firing across all of our strategic priorities, enterprise the system of work and AI, all while driving durable profitable growth. In the enterprise, we had an all-time record quarter at the $1 million level, the $3 million level and the $5 million level of deals. With our $3 million ARR customers growing over 50% year-on-year and our $5 million ARR customers growing over 70%.

我們在所有策略優先事項上全面推進——企業市場、工作系統（system of work）與 AI——同時推動可持續的獲利成長。在企業市場方面，我們在 100 萬美元、300 萬美元與 500 萬美元等級的交易上，都創下歷史單季新高。其中，我們 300 萬美元 ARR 客戶年增超過 50%，500 萬美元 ARR 客戶年增超過 70%。

In the Atlassian system of work, leading enterprises like Google Cloud are using our platform to power the way their teams collaborate. And in AI, Rovo is now used by over 80% of the Fortune 500. Rovo assisted actions grew 50% over the prior quarter and Rovo adopters that is customers using Rovo continue to grow their ARR at more than twice the rate of non-adopters.

在 Atlassian 的工作系統方面，像 Google Cloud 這樣的領先企業正使用我們的平台，驅動其團隊協作方式。在 AI 方面，Rovo 目前已被超過 80% 的《財富》500 強企業使用。Rovo 協助完成的動作較前一季成長 50%；而採用 Rovo 的用戶（也就是使用 Rovo 的客戶）其 ARR 持續以非採用者兩倍以上的速度成長。

When we think about what AI is doing for Atlassian, the results speak for themselves. Customers are getting real value from the AI embedded directly into their workflows, and it's translating into direct growth in our business. Customers are committing more deeply to the Atlassian platform because of the value delivered. We're seeing longer customer commitments, larger deals, stronger seat expansion, great cross-sell momentum and higher NRR.

當我們思考 AI 對 Atlassian 帶來的影響時，結果不言自明。客戶正從直接嵌入其工作流程的 AI 中獲得實質價值，並轉化為我們業務的直接成長。由於所交付的價值，客戶對 Atlassian 平台的投入更深。我們看到更長的客戶承諾期、更大的交易規模、更強的席位擴張、良好的交叉銷售動能，以及更高的 NRR。

Last quarter, we recognized customers like NVIDIA and ResMed at Team '26 for how they're putting the Atlassian platform to work in the AI era to drive their businesses using the Teamwork Graph. The reason customers like these and many more are getting so much value from Atlassian comes down to a key structural advantage and that advantage is context. We believe models will keep improving and organizations will hire that intelligence by the token.

上季在 Team '26 上，我們表揚了 NVIDIA 與 ResMed 等客戶，肯定他們如何在 AI 時代運用 Teamwork Graph，讓 Atlassian 平台發揮效用以推動其業務。像這些客戶以及更多客戶之所以能從 Atlassian 獲得如此大的價值，歸因於一項關鍵的結構性優勢——也就是「情境（context）」。我們相信模型會持續進步，而組織將能以 token 計價「雇用」那種智慧。

A context, their internal knowledge, experience and memory is much harder for organizations to build, and it cannot be hired. With the Teamwork Graph, we have 25 years of deep data about work. That's enabled us to build one of the best context graphs that exist for enterprise knowledge, now spanning over 200 billion objects and connections. For agents grounded in the Teamwork Graph, organizations can see up to 44% more accurate answers while consuming 48% fewer tokens.

但情境——也就是其內部知識、經驗與記憶——對組織而言更難建立，而且無法用雇用的方式取得。透過 Teamwork Graph，我們累積了 25 年關於工作的深度資料。這使我們得以打造企業知識領域中最優秀的情境圖譜之一，目前已涵蓋超過 2,000 億個物件與連結。對於以 Teamwork Graph 為基礎的代理（agents），組織可在消耗少 48% token 的同時，獲得最高多 44% 的答案準確度。

In short, it enables to get customers' results from their AI that are cheaper better and faster. Context is a clear compounding differentiator. I shared more on this in my investor [letter].

簡而言之，它讓客戶從其 AI 中取得更便宜、更好、更快的成果。情境是一項明確且會複利累積的差異化優勢。我在給投資人的[信] 中分享了更多內容。

As you read in our shareholder letter, we're executing well on our long-term strategy from a position of strength and we're seeing this play out each quarter in our strong results. And the best is truly yet to come for Atlassian.

如各位在股東信中所讀到的，我們正以強勢的基礎良好執行長期策略，而我們也在每一季的強勁成果中看到這點持續展現。而 Atlassian 真正最精彩的部分仍在後頭。

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

接下來我把電話交給接線員進行問答。

(Operator Instructions)

(接線員指示)

Adam Wood, Morgan Stanley.

Adam Wood，摩根士丹利。

Congratulations on an amazing -- amazingly strong end to the year. I just wanted to ask about the subscription ARR framework. Obviously, it's kind of newer framework to us. And you again had another way down to year 23% growth for the fiscal year '27 guide is down to 18%.

恭喜你們在年底交出非常——非常強勁的表現。我想請教訂閱 ARR 的框架。顯然，這對我們來說算是比較新的框架。而你們再次出現下滑：2027 會計年度的指引成長率從 23% 下調至 18%。

Could you maybe just help us out around the difference between par and data center within that and how much of this is kind of some moderation around new seed expansion, et cetera? And how much of it is around prudent surrounding macro and continued scaling of the enterprise go-to-market engine?

你能否幫我們釐清一下，在這裡「par」與資料中心（data center）之間的差異？以及這當中有多少是與新種子（new seed）擴張等相關的某種程度調整？又有多少是出於在宏觀環境下的審慎考量，以及持續擴大企業端 go-to-market 引擎的規模？

Thanks, Adam, for the question, and welcome to the coverage. First, I would highlight that we're really pleased to see the strength that we saw in the quarter, printing ARR growth, sub-ARR growth of 23% year-over-year. When it comes to the guide, we'll continue to take a thoughtful and prudent approach here, really balancing the momentum that we're seeing in the business, the demand signals that we're getting from our customers with the uncertainty as you look out in the macro environment, fiscal policy and geopolitical landscape.

謝謝你，Adam，謝謝你的提問，也歡迎你開始覆蓋我們。首先我想強調的是，我們對本季所展現的強勁表現感到非常滿意，ARR 成長、以及次級 ARR（sub-ARR）年增 23%。至於指引（guide），我們會持續採取周全且審慎的做法，在我們看到的業務動能、從客戶端收到的需求訊號，與宏觀環境、財政政策及地緣政治局勢所帶來的不確定性之間取得平衡。

And as you mentioned, this is the first time we'll be issuing ARR guidance for the full fiscal year. We're also still early in the fiscal year here. So taking a bit more of a prudent approach in that second half of FY27. It's also worth noting that in FY26, the ARR year-over-year growth rate benefited by approximately one point from the acquisition of DX at the close of November in 2025. That's our fiscal Q2.

如你所提到，這也是我們首次針對整個會計年度提供 ARR 指引。而且我們目前仍處於會計年度的早期階段。因此在 FY27 的下半年，我們採取更偏審慎一些的做法。另外也值得注意的是，在 FY26，ARR 的年增率約有 1 個百分點受惠於我們在 2025 年 11 月底完成的 DX 收購。那是我們的會計年度第 2 季（fiscal Q2）。

And so we'll be lapping that DX acquisition and the healthy acceleration that we saw in the second half of FY26. And of course, we're going to still continue to scale our enterprise go-to-market sales motion as we unlock the opportunities that we see in front of us. So all I feel really good about the state of the business here. And coming off a quarter where RPO landed at 44% year-on-year NRR was north of 120%. So you're just taking a balanced and prudent approach to the guide.

因此我們接下來會進入與 DX 收購的比較基期（lapping），以及 FY26 下半年我們看到的健康加速成長的比較基期。當然，我們也會持續擴大企業端 go-to-market 的銷售動作（sales motion），以把握我們眼前看到的機會。所以整體而言，我對目前業務狀態感到非常有信心。而且在剛結束的這一季，RPO 年增達 44%，NRR 也超過 120%。因此我們在指引上採取的是平衡且審慎的做法。

Allan Verkhovski, US Bancorp.

Allan Verkhovski，美國合眾銀行（US Bancorp）。

Hey guys, thanks for taking the question and congrats from me as well on another blowout quarter. You've been highlighting that the Teamwork Graph is driving a 44% improvement in task efficiency and a 48% reduction in token usage. Can you elaborate on how these gains are showing up in your customers' deployments and how they're influencing customer conversations, buying behavior and overall sales cycle dynamics?

各位好，謝謝讓我提問，也同樣恭喜你們再交出一個亮眼的季度。你們一直強調 Teamwork Graph 帶來任務效率提升 44%，以及 token 使用量降低 48%。能否進一步說明，這些改善如何在客戶的部署中呈現？以及它們如何影響客戶對話、採購行為與整體銷售週期的動態？

I'm sure I can take that one. Look, I can tell you I've met more than 60 customers over the last 6 weeks and 5 different countries around the world and the Teamwork Graph and AI comp in every single conversation I've had. You can see from our results, those customers are continuing to expand their use of the Teamwork Collection, either upgrading to it if they're buying single applications or adding to the seats as they move into more of their knowledge workers and across the organization.

我想我可以回答這題。我可以告訴你，過去 6 週我在全球 5 個不同國家見了超過 60 位客戶，而 Teamwork Graph 與 AI comp 在我每一場對話中都會被提到。從我們的業績也看得出來，這些客戶持續擴大對 Teamwork Collection 的使用：如果原本只買單一應用，就升級到 Teamwork Collection；或是在他們把使用範圍擴展到更多知識工作者、並推廣到整個組織時，增加席次（seats）。

And AI and the Teamwork Graph is one of the biggest reasons -- it's one of the top two reasons quoted by our customers for their reason to upgrade to the cloud and to upgrade the Teamwork Collection.

而 AI 與 Teamwork Graph 是其中最大的原因之一——它是客戶引用的前兩大原因之一，促使他們升級到雲端，以及升級到 Teamwork Collection。

Now the question is why and where those results come from. Again, our applications have a lot of connections to other applications around the world, right, whether that's their CRM system, whether that's their documentation knowledge systems, we have, for about eight years, been building the Teamwork Graph as the best way to get answers across all of those applications for their users or for their agents, which increasingly ask a lot of questions in a very short period of time.

那問題是為什麼，以及這些成果從何而來。再說一次，我們的應用程式與全球許多其他應用有大量連結，對吧？不論是他們的 CRM 系統，或是他們的文件與知識管理系統。

The quality of those answers is directly related to the [clearability] by graph. The density of that graph or the amount of information and the connections we have between it the quality of the social and collaboration group, the people information that we have. And fundamentally, the breadth of information, a number of contexts you have. We outlined a little bit of this in the shareholder letter, and there's a number of videos that show direct examples of where that gain comes from.

這些答案的品質，與圖譜的「[clearability]」直接相關。圖譜的密度——也就是我們掌握的資訊量與彼此之間的連結程度——以及我們擁有的社交與協作群組品質、人物資訊品質。而從根本上來說，還包括資訊的廣度、以及你所擁有的情境（contexts）數量。我們在股東信中稍微概述了其中一部分，也有一些影片展示這些效益具體是從哪裡來的。

Fundamentally, the better answers, the quality of information across all those applications we can give to your agents, the less token they consume and the less time it takes them to get to an equivalent quality enter. Customers are seeing that. It's based on all of the structured and unstructured workflows we have from those customers over the last 25 years.

從根本上說，我們能在所有這些應用之間提供給你的代理更好的答案、以及更高品質的資訊，他們就會消耗更少的 token，並且用更短的時間達到同等品質的輸入（enter）。客戶正在看到這一點。這是基於我們在過去 25 年中，從這些客戶累積的所有結構化與非結構化工作流程。

And for all those customers that are adopting the Teamwork Collection and the Teamwork Graph coming inside that. I frequently get asked to ask to buy the Teamwork Graph, and the answer is no, you already have it. I can often go and say you have 150 million objects and connections in your Teamwork Graph today and they're like, I do?

而對於所有正在採用 Teamwork Collection、以及其中所包含的 Teamwork Graph 的客戶。我經常被問到能不能單獨購買 Teamwork Graph，而答案是：不需要，你已經擁有它了。我常常可以直接告訴他們：你今天的 Teamwork Graph 裡已經有 1.5 億個物件與連結，他們會說：真的嗎？

And then we can do live demos. We can show them in real time their own graph, how it works using the CLI. It's an amazingly powerful story, especially in an age where all of the customers are asking about token costs and AI spend. It's an incredibly powerful story about how we can actually connect their organizational knowledge, their history, directly to those AI agents and give them a cost and performance, better quality answers and cheaper answers. So it's a very powerful story, 1 we've been building for a long time and very customer isn't in all those conversations.

接著我們可以做現場示範。我們可以用 CLI 即時向他們展示他們自己的圖譜，以及它如何運作。這是一個非常有力的故事，尤其在這個所有客戶都在詢問 token 成本與 AI 支出的時代。這也非常有力地說明：我們如何把他們的組織知識與歷史直接連結到那些 AI 代理，並以更低成本與更佳效能，提供更高品質、也更便宜的答案。所以這是一個非常有力的故事——我們已經打造很久了——而且幾乎每位客戶在所有這些對話中都會提到。

Ryan MacWilliams, Wells Fargo.

Ryan MacWilliams，富國銀行（Wells Fargo）。

Hey, thanks for taking the question. For James, I'd love to hear about the components of what drove cloud strength in the fourth quarter beyond the guidance? Like was it more customer seat additions or more Teamwork Collection? And have you thought about approaching this full year cloud guide in this uncertain environment? Was there any changes to guidance philosophy overall?

嗨，謝謝讓我提問。想請教 James：第四季雲端表現強於指引，背後有哪些構成因素？例如，是更多客戶席次增加，還是更多 Teamwork Collection？另外，在這個不確定的環境下，你們如何思考今年全年的雲端指引？整體的指引哲學有任何改變嗎？

Thanks, Ryan. No, the philosophy remains -- really remains the same here. And on the cloud revenue drivers here on the $49 million beat, maybe I'll start with the fact that and highlight the fact that the DC migrations to cloud were really in line with expectations and not a contributor to the $49 million beat. It's progressing well and in line with our expectations for FY26 contributing approximately mid- to high single digits to the cloud growth revenue.

謝謝你，Ryan。沒有，哲學——在這裡我們的哲學確實維持不變。至於雲端營收的驅動因素，針對這次超出 4,900 萬美元（$49 million）的部分，我先從一點開始並強調：資料中心（DC）遷移到雲端的進度其實符合預期，並不是這 4,900 萬美元超預期的貢獻來源。遷移進展良好，且符合我們對 FY26 的預期，預計將對雲端營收成長貢獻約中到高個位數（mid- to high single digits）。

So the two primary drivers here of the outperformance were really the cross-sell and seat expansion that we were seeing similar to what we saw in the cross sell side, we continue to see strong uptake in our collections across Service Collection and Teamwork Collection. Service Collection revenue growth accelerated in Q4, just a strong signal of the AI adoption that we're seeing and that Mike mentioned a little bit earlier here. In fact, we're seeing over 3 times increase in Agentic automation in Service collection as customers are continuing to deploy more agents beyond IT teams in HR and legal and finance and marketing teams.

因此，這裡表現優於預期的兩個主要驅動因素，確實是我們看到的交叉銷售與席位擴張，與我們在交叉銷售端所看到的情況相似；我們持續看到客戶在 Service Collection 與 Teamwork Collection 各項套組中都有強勁採用。Service Collection 的營收成長在第四季加速，這是我們所看到的 AI 採用（Mike 先前也稍微提到）的一個強烈訊號。事實上，我們看到 Service Collection 中的 Agentic 自動化增加超過 3 倍，因為客戶持續在 IT 團隊之外，於人資、法務、財務與行銷團隊部署更多代理。

So again, also great to see the recognition here being named a leader in Gartner's Magic Quadrant for conversational AI platforms for employee services. We also continue to see a lot of strong performance on the Teamwork Collection side. A good reminder that Teamwork collection is still the best vehicle for customers to adopt and unleash AI on the Atlassian platform.

另外，也很高興看到我們獲得肯定，被評為 Gartner《魔力象限》員工服務對話式 AI 平台的領導者。我們也持續看到 Teamwork Collection 端有許多強勁表現。再次提醒，Teamwork Collection 仍然是客戶在 Atlassian 平台上採用並釋放 AI 能力的最佳載體。

This is where customers are upgrading to increased Rovo credits, it's giving customers what they want in terms of the predictability as well as the ability to really load balance those tokens across their organizations, and we're seeing that drive our strong ARPU uplift here as well.

這也是客戶升級到更多 Rovo 點數的地方；它在可預測性方面，以及讓客戶能夠在其組織內真正地對這些 token 進行負載平衡方面，提供了客戶想要的能力，而我們也看到這同樣推動了我們強勁的 ARPU 提升。

So really, I think what this shows is that the customers are consolidating their collaboration stack onto the Atlassian platform, recognizing the value that the Teamwork Graph really brings here. And we do see that in the results with Teamwork customers having 4 times to 5 times more paid seats per instance than stand-alone Jira or Confluence customers.

所以我認為，這真正顯示的是客戶正在把他們的協作技術堆疊整合到 Atlassian 平台上，並認可 Teamwork Graph 在此帶來的價值。我們也在結果中看到這點：Teamwork 客戶每個實例的付費席位數，是單獨使用 Jira 或 Confluence 的客戶的 4 到 5 倍。

Karl Keirstead, UBS.

Karl Keirstead，瑞銀（UBS）。

A number of software stocks this week have been clocked on weaker-than-expected margin performance as their AI use goes up and their AI mix goes up and yet I'm looking at the Atlassian results and your margins despite you guys also experiencing an uptick in AI usage holding in there very well. And I'm wondering if you can explain that difference what you guys might be doing? Cost structure might be different. That might be helpful.

本週有不少軟體股因為毛利率表現低於預期而受到衝擊，原因是它們的 AI 使用量上升、AI 組合占比提高；但我在看 Atlassian 的結果時，你們的利潤率在同樣經歷 AI 使用量上升的情況下仍維持得很好。我想請你們解釋一下這個差異，你們可能做了哪些不同的事情？成本結構可能不同。如果能說明會很有幫助。

I'll Start with that. Again, on the margin guide here, as Mike and I stated back in the Q3 shareholder letter, we're committed to driving durable profitable growth here and a positive GAAP operating margin target for FY27 really sets us on that path to accelerate and expand GAAP profitability over time. And in terms of drivers, look, the big part of that is disciplined hiring approach, right?

我先回答這個問題。再次就這裡的利潤率指引而言，正如 Mike 和我在第三季股東信中所說，我們致力於推動可持續的獲利性成長，而 FY27 的正向 GAAP 營業利潤率目標，確實讓我們走在這條路徑上，隨時間推進加速並擴大 GAAP 獲利能力。至於驅動因素，坦白說，其中很大一部分是有紀律的招募策略，對吧？

We talked about that as well as reallocating resources against some of the biggest opportunities and ROI that we see right now as well as moderating the stock-based compensation here.

我們也談到這點：以及把資源重新配置到我們目前看到的最大機會與投資報酬（ROI）上，同時也放緩股票基礎薪酬（stock-based compensation）。

At the same time, we're going to continue to judiciously invest in AI and scale our enterprise sales motion where we do see a lot of customer demand right now and our ability to really kind of capture that sustained opportunity in front of us. So that goes to really help inform the 6.5% GAAP operating margin guide for Q1 and a 4.5% GAAP operating margin for FY27 as we continue to focus on driving that profitability and growing that over time.

同時，我們也會持續審慎投資 AI，並擴大我們的企業銷售推進（sales motion），因為我們目前確實看到大量客戶需求，也看到我們有能力去掌握眼前這個可持續的機會。因此，這也有助於形成第一季 6.5% 的 GAAP 營業利潤率指引，以及 FY27 4.5% 的 GAAP 營業利潤率指引；我們將持續專注於推動獲利能力並隨時間提升。

Yes. Look, Karl, I can add on from the gross margin in the COGS side. I would say it's a classic Atlassian story of being thoughtful and long-term thinking -- long-term orientation. We've talked for 3 years now around our AI gateway and how we think about AI. We've been blending multiple models for a multiyear period, and we're getting really, really good at it. It's a nontrivial challenge. I'll give a huge shout out to all of our R&D teams who have been working on that cost to serve question for a long time in terms of getting the best quality at the best price and the best speed.

是的。Karl，我可以從毛利率與銷貨成本（COGS）端補充。我會說這是典型的 Atlassian 故事：深思熟慮、以長期為導向——長期取向。我們已經談了 3 年我們的 AI gateway，以及我們如何思考 AI。我們在多年期間一直在混合多種模型，而且我們做得越來越好。這是一個非常不容易的挑戰。我要特別向所有投入這個「服務成本（cost to serve）」議題很久的研發團隊致敬：如何以最佳品質、最佳價格與最佳速度來達成。

People want front fast answers from their AI across Teamwork Graph, across our enterprise search architecture, which is now the best in the world in terms of giving you answers across your knowledge estate, whether that's to a person or human or whether that's to an agent or something clearing in that manner from an AI perspective.

人們希望他們的 AI 能夠快速給出第一時間的答案——無論是在 Teamwork Graph 上、在我們的企業搜尋架構上；而這套架構現在在「跨整個知識資產（knowledge estate）提供答案」方面可說是世界最佳，不論是提供給人（使用者）或是提供給代理（agent），或是以 AI 的方式在那樣的流程中運作。

Our ability to blend models, our ability to manage traffic costs over time is truly phenomenal. We get better at it every quarter. And you can see that we had a huge continued growth in AI usage. Again, Rovo assisted actions are up 50% quarter-on-quarter. And we've continued to do that and focus on the durable profitable growth that we've talked about for a while now.

我們混合模型的能力、以及隨時間管理流量成本的能力，確實非常出色。我們每一季都在變得更好。你也可以看到，我們的 AI 使用量持續大幅成長。再次強調，Rovo 輔助動作（assisted actions）季對季成長 50%。我們一直在這麼做，並專注於我們談了好一段時間的可持續獲利性成長。

So it's a huge credit to everyone and the Atlassian tend to be able to do that. It's a lot of hard work. It's a lot of long-term thinking about the architecture that we built underneath our AI to be able to give those answers in a -- in an incredibly competitive value way to our customers as well. So not only do we give them the best answers, we give them the best answers at an incredible price, which is part of the reason you're seeing the growth in the service question, the growth in the Teamwork Collection, again, as James mentioned service question, again accelerating AI being a huge driver of that service space.

所以這要歸功於每一位同仁，而 Atlassian 往往能做到這點。這需要大量的努力。也需要對我們在 AI 底層所建構的架構進行長期思考，才能以——以極具競爭力的價值方式為客戶提供那些答案。因此，我們不只給客戶最好的答案，也以非常有吸引力的價格提供最好的答案；這也是你看到 Service（這裡所提到的服務領域）成長、Teamwork Collection 成長的部分原因。再次呼應 James 的說法，Service（領域）成長加速，AI 是該服務領域的一個巨大驅動因素。

Again, hearing record numbers for customer conversions from legacy service providers, et cetera. So -- and you see that with our MCP calls as well as an example, that's also nontrivial to serve cheaply to manage that over time. MCP calls were up 400% in the quarter with daily blip and people actually using the applications through the label through a mobile client. So really happy where we sit from a technological perspective of being able to deliver this and more improvements to come.

再次，我們也聽到來自傳統服務供應商等的客戶轉換數量創下新高。所以——你也可以從我們的 MCP 呼叫量看到這點，這同樣不是件容易的事：要能以低成本提供服務並隨時間管理。本季 MCP 呼叫量成長 400%，每日都有明顯跳升，而且人們確實透過行動端用戶端來使用這些應用程式。因此，從技術角度來看，我們對目前能交付這些能力的位置感到非常滿意，後續也會有更多改進。

Gregg Moskowitz, Mizuho.

Gregg Moskowitz，瑞穗（Mizuho）。

All right. Great. Congratulations on another tremendous result. It does seem that the Teamwork Graph, the functional improvements to Rovo, some of the other innovations to Jira, JSM, et cetera, are really resonating with your customers. In terms of my question, we've seen some market data that have indicated somewhat of a resurgence in developer hiring. Does that align with what you're seeing and hearing across your clients?

好的。很棒。恭喜再次取得非常亮眼的成果。看起來 Teamwork Graph、Rovo 的功能改進，以及對 Jira、JSM 等其他創新，確實正在引起客戶共鳴。就我的問題而言，我們看到一些市場數據顯示開發者招募有些回溫。這是否與你們在客戶端所看到、所聽到的情況一致？

And secondly, how does the rate of total paid seat expansion this quarter spanning technical users and knowledge workers? How does that compare to what you've been seeing in recent periods?

其次，本季涵蓋技術使用者與知識工作者的「付費席位（paid seat）」總擴張率表現如何？與你們近期幾個季度所看到的情況相比如何？

Thanks, Gregg. I can start with that and maybe throw it over to Mike. To Ryan's question a little bit earlier about some of the cloud outperformance drivers that we saw. I talked about the cross-sell side of it, and we are seeing customers continue to go wall to wall with Atlassian here and continue to adopt higher additions and move towards Teamwork Collection, Service Collection.

謝謝你，Gregg。我可以先回答這題，然後也許再交給 Mike 補充。回到 Ryan 稍早問到的、我們看到的一些雲端表現優於預期的驅動因素。我談過交叉銷售（cross-sell）那一塊；我們看到客戶持續在 Atlassian 這裡「全產品導入（wall to wall）」並持續採用更高階版本，且朝向 Teamwork Collection、Service Collection 推進。

But I think importantly, we're seeing that growth in Teamwork Collection while also seeing continued strong seat expansion in our core Jira and Confluence offerings here. And this expansion is broad-based. Recall that roughly two thirds of users across Jira and Confluence or knowledge workers, right? So these are non-software development teams, really sitting in HR, legal, finance and really speaks to that diversified user base that we're seeing here.

但我認為更重要的是，我們在看到 Teamwork Collection 成長的同時，也看到核心 Jira 與 Confluence 產品持續強勁的席位擴張。而且這種擴張是廣泛分布的。請記得，Jira 與 Confluence 的使用者中約有三分之二是知識工作者，對吧？也就是非軟體開發團隊，主要在 HR、法務、財務等部門，這也反映出我們正在看到的多元化使用者基礎。

And I think this growth does demonstrate that in AI-driven world, the need for collaboration at the enterprise scale, tracking, planning, managing work, it's actually increasing. And Atlassian is a mission-critical platform here, helping our customers orchestrate their teams, their agents, their workflows to really unlock that value in AI and driving real ROI and real outcomes here. And that really was the other primary driver of the outperformance that we saw in cloud, including the cross-sell that I talked about a little bit earlier.

我認為這樣的成長也顯示，在 AI 驅動的世界裡，企業規模的協作需求——追蹤、規劃、管理工作——其實正在增加。Atlassian 在這裡是一個關鍵任務（mission-critical）的平台，協助客戶協調其團隊、其代理（agents）、其工作流程，以真正釋放 AI 的價值，並帶來實質的 ROI 與成果。而這也確實是我們在雲端看到表現優於預期的另一個主要驅動因素，包含我稍早提到的交叉銷售。

Look, I can talk about it from a developer or engineering front in a number of different angles, Gregg. I know we've talked about this before. We've been long on record. You can look at the receipts I think there's getting more developers in the world in five years' time than there are today. The cost of building technology is going down. The amount of technology we're going to build is going to go up. I believe that will continue to be a good trend. We've said this for a number of years. Those are the signals we see in our customers, the ability to grow that is huge.

Gregg，我可以從開發者或工程的角度，用幾個不同面向來談。我知道我們以前談過這個。我們長期以來一直公開表達這個觀點。你可以看各種佐證；我認為五年後世界上的開發者會比現在更多。打造技術的成本正在下降。我們將要打造的技術量會上升。我相信這會持續是一個正向趨勢。我們已經講了好幾年。這些都是我們從客戶身上看到的訊號，而能夠推動這樣的成長非常重要。

I think secondly, the number of people building technology that we don't necessarily call developers is also growing. So we're seeing people in finance and in HR and in marketing, using a lot of these tools to build technology of various different forms. Again, we see that in our MCP server usage and our CLI usage coming across the board in lots of different functional areas, not just inside the engineering fraternity. And I think that's really important.

第二點，我認為在打造技術的人數中，有一群我們不一定稱之為開發者的人也在增加。因此我們看到財務、HR、行銷等部門的人，也在使用許多工具來打造各種形式的技術。同樣地，我們從 MCP 伺服器使用量與 CLI 使用量中看到這種趨勢，遍及許多不同職能領域，而不僅僅是在工程圈內。我認為這點非常重要。

From our perspective, again, the system of work is about helping companies to connect their technology teams with their business teams to have a single platform that allows both sides of the organization in a technology-driven organization, an organization where software is their key competitive advantage. It doesn't make them a technology company.

從我們的角度來看，「工作系統（system of work）」是要幫助企業把技術團隊與業務團隊連結起來，提供一個單一平台，讓組織兩端都能在一個以技術驅動的組織中運作——也就是軟體是其關鍵競爭優勢的組織。但這不代表它就是一家科技公司。

It can be a bank, an insurance company, a health care company, a university. There are lots of companies for whom. Many organizations in the world, technology is a key competitive advantage. That is -- I believe that is a big growth cycle that we are continuing to participate in. Our strength is in connecting the technology team through the business team and I do believe that the developer hiring will continue to grow as we see in our customer signals.

它可以是一家銀行、保險公司、醫療保健公司或大學。有很多公司都是如此。對世界上許多組織而言，技術是一項關鍵競爭優勢。我相信這是一個我們持續參與其中的大型成長循環。我們的強項在於把技術團隊與業務團隊連結起來；而且我確實相信，隨著我們從客戶訊號所看到的趨勢，開發者招募將會持續成長。

DJ Hynes, Canaccord.

Canaccord 的 DJ Hynes。

Hey, thank you, guys. Congrats on a nice quarter. Mike, you gave us some data points around Rovo and the customers who have adopted complete more of Jira work, they create more Confluence content, they grow ARR twice as fast. How do you separate kind of the incremental impact of Rovo from the possibility that your most engaged customers are simply the earliest adopters? I'm trying to think about the durability of those trends as we get further along into the Rovo penetration story.

嗨，謝謝各位。恭喜這是一個很不錯的季度。Mike，你提供了一些關於 Rovo 的數據點：採用 Rovo 的客戶會完成更多 Jira 工作、建立更多 Confluence 內容，而且 ARR 成長速度是兩倍。你們如何區分 Rovo 的增量影響（incremental impact），以及「最投入的客戶只是最早採用者」這種可能性？我在思考，隨著 Rovo 滲透率故事往後推進，這些趨勢的持久性（durability）會如何。

Hi, DJ, think about how to answer that one. I would say we see signals that is only going to be a durable trend. I don't think it's anything to do with early adopters, to be honest. This is a technology that's been adopted incredibly fast. Again, over 80% of the Fortune 500 are using Rovo. All of the customer meetings that I walk into and it's a lot, they have Rovo usage, AI usage, they're looking to expand their usage. They're looking to understand. This is a massive sort of super cycle, right?

嗨，DJ，我想想要怎麼回答這題。我會說，我們看到的訊號顯示這只會是一個可持續的趨勢。老實說，我不認為這和早期採用者有關。這項技術的採用速度非常快。同樣地，《財富》500 大企業中有超過 80% 正在使用 Rovo。我走進的每一場客戶會議——而且很多——他們都有 Rovo 使用、AI 使用，並且希望擴大使用。他們想要理解。這是一個巨大的「超級循環（super cycle）」，對吧？

They are customers inevitably trying to work out how can I apply AI and smart and the Teamwork Graphs and agents to this workflow or that workflow, structured workflows like in Jira or a service cushion or unstructured workflows like Balloon and Confluence and Trello, and other areas where they have unstructured workflow that they believe they can make more efficient or importantly, higher quality, right? The results of that workflow can be better than they were before. I don't see that as being an early adopter problem.

客戶不可避免地在嘗試釐清：我該如何把 AI、智慧能力、Teamwork Graphs 與 agents 應用到這個或那個工作流程上——像 Jira 或 Service Collection 這類結構化流程，或像 Balloon、Confluence、Trello 這類非結構化流程，以及其他他們認為可以變得更有效率、或更重要的是品質更高的非結構化流程領域。這些流程的產出結果可以比以前更好。我不把這視為早期採用者的問題。

And every single customer I walk into, they're all learning, adopting, moving and trying. What they are also seeing is the quality of Rovo is incredibly high, right? It outperforms almost every chat tool on the market across their data, the more they connect into the Teamwork Graph, the better those results get because of our context, because of a lot of great work in R&D.

而且我走進的每一位客戶，他們都在學習、採用、推進並嘗試。他們同時也看到 Rovo 的品質非常高。在他們的資料範圍內，Rovo 的表現優於市面上幾乎所有聊天工具；而且他們越是把資料連結到 Teamwork Graph，因為我們的情境（context）以及研發上的大量優秀工作，結果就會越好。

So I don't think these are sort of early adopter challenges. I think it's a signal of the more people lean in, the better results they get, the better value they feel like they're getting, they're then expanding across more of their business teams, more areas of their organization or understanding that when they add Strategy Collection or they add DX or they add the Service Collection the platform, the Teamwork Graph they've built just compounds in value for them.

所以我不認為這是所謂的早期採用者挑戰。我認為這是一個訊號：越多人投入使用（lean in），他們得到的結果就越好，感受到的價值也越高，接著就會擴展到更多業務團隊、組織內更多領域；或是理解到當他們加入 Strategy Collection、加入 DX、或加入 Service Collection 時，他們所建立的平台與 Teamwork Graph 的價值會持續複利式累積。

One of the reasons Service Collection continues to grow so strongly is because of the compounding effect between the Teamwork Collection and the Service Collection. And that's a really important thing for them to understand. And I think there's as we've shown in our addressable market figures of $14 billion within the existing customer base without changing any products or pricing, et cetera, that there is a huge amount of upside opportunity still for us to go out and capture. And that's our challenge to do on a weekly basis.

Service Collection 之所以能持續如此強勁成長，其中一個原因是 Teamwork Collection 與 Service Collection 之間的複利效應。而這是他們需要理解的一個非常重要的點。而且我認為，正如我們在可服務市場（addressable market）數據中所展示的，在不改變任何產品或定價等前提下，僅在既有客戶基礎內就有 140 億美元的機會，對我們而言仍有非常大的上行空間可以去捕捉。而這就是我們每週都要面對的挑戰。

Koji Ikeda, Bank of America.

Koji Ikeda，美國銀行。

Yeah, hey, guys. Thanks so much for taking my question. I wanted to ask about kind of the long-term strategy here. And so clearly, the long-term Atlassian share gain thesis is really hinged on owning the enterprise context. And so what evidence are you hearing or seeing from your customers, especially the largest ones that you have that even if they increasingly centralized context inside other AI data layers, the Atlassian Teamwork Graph is still the moat versus potentially becoming a feature.

是的，嗨，各位。非常感謝讓我提問。我想問一下這裡的長期策略。很明顯，Atlassian 長期市占提升的論點，關鍵在於掌握企業情境（context）。那麼，從你們的客戶（尤其是你們最大的那些客戶）那裡，你們聽到或看到哪些證據，能支持即便他們愈來愈把情境集中在其他 AI 資料層之中，Atlassian Teamwork Graph 仍然是護城河，而不至於可能淪為一個功能（feature）？

Koji, I can talk to that one. I think context is obviously very hard to build. This is a nontrivial problem. It's a unique differentiator that we have, and we continue to improve on. We made massive improvements in the last quarter in the Teamwork Graph. The quality of connectors and the information that we have, how we continue to correlate cross-reference and infer across that data, for example skills we can now infer in the talent app inside the Strategy Collection across all the information in the Teamwork Graph on which are inferring the skills and AI inactivity of your employees, for example.

Koji，我可以回答這題。我認為情境顯然非常難建立。這不是一個簡單的問題。這是我們獨特的差異化優勢，我們也持續在改進。上個季度我們在 Teamwork Graph 做了大幅改進。連接器的品質、我們掌握的資訊，以及我們如何持續在這些資料之間做關聯、交叉引用與推論；例如，我們現在可以在 Strategy Collection 裡的 talent app 中，基於 Teamwork Graph 的所有資訊推論技能，並推論員工的技能與 AI 活動情況等。

The quality of our net results is up markedly in the last quarter and was already at all-time highs. We've seen it in a number of true fans of chat growing up and up and up, right? The quality of our search, the semantic indexing of code that we showed, there is a lot of hard work to continue to do to make that context to make the Teamwork Graph better and better every quarter. We have more coming Team EU is in October in Amsterdam. No doubt there will be further improvements shown there.

我們的淨結果品質在上個季度顯著提升，而且原本就已經在歷史高點。我們也在一些真正的 chat 重度使用者身上看到，使用量一路成長、持續上升，對吧？我們搜尋的品質、我們展示的程式碼語意索引（semantic indexing），都還有很多艱難的工作要做，才能讓這個情境、讓 Teamwork Graph 每一季都變得更好。我們還會有更多內容，Team EU 會在 10 月於阿姆斯特丹舉行。毫無疑問，屆時會展示更多進一步的改進。

I don't think it's as simple as every organization will have one graph and we're done. I think most organizations will have three to five large-scale knowledge graphs, and we intend to be one of those three. We believe we have a significant advantage at the moment and continue to invest, as James said, in our AI and R&D teams to continue to make that accessible and to make that value and quality continue to improve.

我不認為事情會簡化成每個組織只會有一個 graph，然後就結束了。我認為大多數組織會有 3 到 5 個大型知識圖譜，而我們打算成為其中三個之一。我們相信目前我們有顯著優勢，並且會持續投資；正如 James 所說，在我們的 AI 與研發團隊上持續投入，讓它更容易被使用，並讓其價值與品質持續提升。

The Teamwork Graph, importantly, we opened up at Team '26 US, which is a really important step with our MCP server, with our command line [unified] so CLI, the ability to put the Teamwork Graph to access its information in any Agentic harness that you're using either chat applications, agents running on other platforms is really important.

Teamwork Graph 很重要的一點是，我們在 Team '26 US 將其開放出來；這是非常重要的一步，透過我們的 MCP server、以及我們的命令列（[unified]，也就是 CLI），讓 Teamwork Graph 的資訊可以在你使用的任何 Agentic 框架中被存取，不論是聊天應用程式，或是跑在其他平台上的代理（agents），這都非常重要。

This is about being an interconnected part of a customer's technology estate. It's about the Atlassian platform being a strategic piece in the puzzle of how it is that they run their business. And I think that integration is a really important point.

這關乎成為客戶技術版圖中彼此互聯的一部分。這關乎 Atlassian 平台成為他們如何營運業務這個拼圖中的策略性一塊。我認為整合性是非常重要的一點。

We're seeing that the reason we talk about the MCP server, MCP and CLI users in the quarter, passing 1 million now. We don't know where that puts us in a world scale, but I think it's 1 of the biggest MCP servers that exist, more than doubling in the quarter is because of people getting to the Teamwork Graph getting to that context from their agents wherever they're deployed. Again, we also signed and increased our partnerships with all of the Foundation Labs.

我們看到之所以會談 MCP server、MCP 與 CLI 使用者在本季已突破 100 萬，我們不知道這在全球尺度上代表什麼，但我認為它是現存最大的 MCP servers 之一，而且本季成長超過一倍，是因為人們能從他們部署在各處的 agents 取得 Teamwork Graph、取得那個情境。同樣地，我們也與所有 Foundation Labs 簽署並擴大了合作夥伴關係。

During the quarter, we released Cursor and Claude and GitHub agents within Jira to again bring our interfaces, those agents into our interfaces, so you can use them as natively as possible so that we are part of the network of tools that help an integral part, the strategic part of the network of tools that helps those companies to do what they need to do. I think the Teamwork Graph has incredible long for promise, and we continue to invest heavily to ensure that it remains best source of context and enterprise commit.

本季我們在 Jira 內發布了 Cursor、Claude 與 GitHub agents，再次把那些 agents 帶進我們的介面中，讓你能盡可能以原生方式使用它們，讓我們成為工具網路的一部分——而且是其中不可或缺、具策略性的那一部分，協助這些公司完成他們需要做的事。我認為 Teamwork Graph 有非常長期的潛力，我們也持續大力投資，以確保它仍然是最佳的情境來源與企業承諾。

Alex Zukin, Wolfe Search.

Alex Zukin，Wolfe Search。

Just a few. I guess what -- like if you had to think about what really happened this quarter that seemingly unlocked and unblocked the enterprise motion because it did seem like there was a meaningful inflection in large deal activity and consequently, our RPO. So just great to get maybe an extra lens on like what was the unlock or the unlock that you saw?

我有幾個問題。我想，如果你們要回顧本季到底發生了什麼，似乎解鎖並打通了企業端的推進動能，因為看起來大型交易活動確實出現了有意義的拐點，進而帶動我們的 RPO。所以很希望能多一個角度，談談你們看到的解鎖點是什麼？

Thanks, Alex. I would say the most important thing is there was no unblock. So I would say this is a continued delivery of the activity we've seen over the last two quarters, three quarters, four quarters. We've talked for three or four years about how we continue to improve our enterprise motion that is on the technology side. Compliance, we should hit a compliance, for example, for Rovo during the quarter.

謝謝，Alex。我會說最重要的是，並沒有什麼被「打通」的單一事件。所以我會說，這是我們在過去兩季、三季、四季所看到活動的持續交付。我們談了三、四年，說我們如何持續改善企業端的推進動能，這包含技術面。例如合規性——我們在本季讓 Rovo 達成了一項合規里程碑。

So the enterprise group is doing a fantastic job on the technology side about making these tools technologies, applications, collections available to companies at scale at incredible scale now in the cloud in different geographies with different compliance requirements.

因此，企業團隊在技術面做得非常出色，讓這些工具、技術、應用程式與 collections 能夠以規模提供給企業——現在在雲端已經能以驚人的規模，在不同地理區域、符合不同合規要求地提供。

So technology-wise, we're doing a great job. We've continued to improve over the last few years, the human side of our enterprise connectivity as we become more of a strategic partner from sales to customer success to all of the groups that have to interact with our customers at the enterprise level. We've continued to do a better and better job every week for the last multiple years, and you see those results coming through.

所以在技術面，我們做得很好。過去幾年，我們也持續改善企業連結的「人」的面向：從銷售到客戶成功，再到所有必須在企業層級與客戶互動的團隊，隨著我們成為更具策略性的合作夥伴。過去多年來，我們每週都在做得更好，而你也看到這些成果逐步反映出來。

The customer adoption is a continued trend, right? The RPO acceleration that we have seen again is about customers simply understanding our platform seeing more value across multiple collections in the strategic platform sense of Atlassian and then signing temps, longer deals, larger deals and then expanding their footprint. That's why you see the strong NRR numbers and the strong RPO numbers.

客戶採用是一個持續的趨勢，對吧？我們看到的 RPO 加速，仍然是因為客戶更理解我們的平台，在 Atlassian 作為策略性平台的意義下，看到跨多個 collections 的更多價值，然後簽更長期、更大金額的合約，並擴大他們的使用版圖。這也是為什麼你會看到強勁的 NRR 數字與強勁的 RPO 數字。

So the most important thing I would say is it's not a singular unlock. It is many years of hard work, continual work across lots of areas of the business behind a deep belief in the strength we have in the enterprise segment. James?

所以我會說最重要的是，這不是單一的解鎖點。這是多年辛勤努力、持續投入，在業務的許多領域同步推進，背後是我們深信自己在企業市場區隔的優勢。James？

And just maybe to add to that, I do want to reiterate what Mike said that it's not an overnight thing. We've got sort of similar questions last quarter about Service Collection when we had announced the $1 billion ARR growing north of 30% year-on-year. Those are quarters and years of investment that we're making here.

我再補充一下，我確實想重申 Mike 說的：這不是一夕之間發生的。上季我們宣布 Service Collection 的 ARR 達到 10 億美元、年增超過 30% 時，也收到類似的問題。這些都是我們在這裡投入了好幾季、好幾年的投資成果。

And if you take a look at the enterprise go-to-market motion that we're continuing to evolve here, we really are still early stages when you think about the fact that we're in 85% of the Fortune 500, but really that only represents roughly 10% or so of our total revenue. There's still a lot of headroom there. And sure some of these stats a little bit earlier, but our $3 million cohort, that's 164 customers growing 50% year-over-year. Our $5 million and greater cohort of customers, that's 69, and that's on 70% year-over-year.

而如果你看看我們在此持續演進的企業級 go-to-market 推進方式，考量到我們已進入《財富》500 強中的 85%，但實際上這僅約占我們總營收的 10% 左右，你會發現我們仍處於相當早期的階段。仍有很大的成長空間。當然，前面也提到一些數據，但我們 300 萬美元級別的客群共有 164 家客戶，年增 50%。而 500 萬美元以上的客群共有 69 家客戶，年增 70%。

But again, on a base of $6.6 billion of ARR, it really just highlights how early we are here in terms of the opportunity.

但再說一次，在 66 億美元 ARR 的基礎上，這更凸顯了就機會而言，我們目前仍非常早期。

Overnight success, three years in the making. I guess maybe just on the initial ARR guide, James, I think it implies roughly flat organic net new ARR in fiscal '27. And you said you're reflecting the same level of prudence that you have historic -- last historically in this outlook. Is there anything different exiting fiscal '26 that increases the confidence in those numbers given some of those dynamics that you guys just went over? Or just maybe just help level set that a little bit?

一夕成名，其實是三年磨一劍。我想就最初的 ARR 指引來說，James，我認為它暗示在 27 財年有機淨新增 ARR 大致持平。你說你在這份展望中反映了與過去一貫相同程度的審慎——歷史上一直如此。鑑於你們剛剛提到的那些動態，在 26 財年結束時，有沒有任何不同之處能提高你對這些數字的信心？或者能否幫我們把這點再校準一下？

Yes. No. Again, I'll just say again, you're really pleased with the momentum that we're carrying into FY27 against subs ARR growth of 23%, RPO growth of 44% year-over-year. So we do really feel good about the state of the business and the healthy broad-based customer demand signals that we're seeing.

是的。沒有。再說一次，我只想再強調，我們對帶入 27 財年的動能感到非常滿意：訂閱 ARR 成長 23%，RPO 年增 44%。因此，我們確實對業務狀況以及我們看到的健康、廣泛的客戶需求訊號感到很有信心。

And as I mentioned a little bit earlier, worth noting the FY26 ARR year-over-year growth rate did benefit from approximately 1 point from the DX acquisition. So when it comes to that initial guide of 18% for subs growth for FY27, again taking a balanced and thoughtful approach here, and it's still early in the year. So just adding a bit more prudence in that second half of FY27. So hopefully that gives you some color there.

另外，如我稍早提到的，值得注意的是，26 財年 ARR 的年增率約受益於 DX 併購帶來的約 1 個百分點。因此，談到 27 財年訂閱成長 18% 的初始指引，我們再次採取平衡且審慎的做法，而且目前仍在年度初期。所以在 27 財年下半年我們加入了更多審慎假設。希望這能提供你一些補充說明。

Rob Owens, Piper Sandler.

Rob Owens，Piper Sandler。

Hello, can you hear me? Yes, my apologies. So the question was around increased Rovo credits with customers and how that's in very strong ARPU uplift. I'm just curious, at what point might we start to contemplate a more consumptive elements to the model? And within the Rovo installed base of customers using it, can you give us any increased color on how often they're coming back? Have you seen multiple increases in just what that's doing to ARPU overall?

你好，你聽得到我嗎？是的，不好意思。所以問題是關於客戶的 Rovo 點數增加，以及這如何帶來非常強勁的 ARPU 提升。我只是好奇，我們可能在什麼時點開始考慮把模型加入更多按用量計費（consumptive）的元素？另外，在已安裝並使用 Rovo 的客戶基礎中，你能否提供更多資訊，說明他們回來使用的頻率？你們是否看到多次提升，進而對整體 ARPU 造成的影響？

Hi, Rob, look, as we said, we continue to see a great increase in Rovo assisted actions, which is probably a rough proxy for credit usage. It's not directly linear, but the 50% quarter-on-quarter growth gives you a sense of the customer adoption of Rovo across the areas.

嗨，Rob，你看，正如我們所說，我們持續看到 Rovo 輔助動作大幅增加，這大概可以作為點數使用量的一個粗略代理指標。它不是完全線性對應，但季度對季度 50% 的成長，能讓你感受到客戶在各個領域對 Rovo 的採用情況。

Secondly, as we've said that the Teamwork Collection with approximately 10x as many Rovo credits included is a great motivator inducement for upgrade in terms of hitting those limits and looking at upgrading the Teamwork Collection, both in terms of moving to the Teamwork Collection, but also expanding the number of seats, the reps of deployment.

第二，如我們所說，Teamwork Collection 內含約 10 倍的 Rovo 點數，是促使升級的一個很強的動機與誘因，因為客戶會觸及那些上限，進而考慮升級到 Teamwork Collection——不僅是轉換到 Teamwork Collection，也包括擴大席位數、擴大部署範圍。

Again, as we talked about, as I see the Teamwork Graph working in most different types of teams, you see seat increases, right, which is, again, driving that ARPU number, if you're looking at it at a really macro level, that is happening there.

再者，如我們討論過的，當我看到 Teamwork Graph 在各種不同類型的團隊中運作時，你會看到席位增加，對吧？這同樣在推動那個 ARPU 數字；如果你從非常宏觀的角度來看，這確實正在發生。

We have a series of consumptive elements today. We continue to work on how we judiciously and prudently roll out consumption-based offerings. We talked about both the direct consumption-based offering for overages.

我們目前已經有一系列按用量計費的元素。我們也持續在研究如何審慎且穩健地推出以用量為基礎的產品方案。我們談到過兩者：其一是針對超額使用的直接按用量計費方案。

And secondly, the Flex model, which we continue to work on to allow customers to opt in to the entire Atlassian platform with a sort of dollar-based commitment that they can use across various different collections and consumptive elements in a singular package, both of which are early in their cycle holds great promise for the future.

其二是 Flex 模式，我們持續推進，讓客戶可以選擇以某種以美元計價的承諾金額來採用整個 Atlassian 平台，並可在各種不同的 collections 與按用量計費元素之間使用，打包成單一方案；這兩者都仍處於早期週期，但對未來都很有潛力。

But as always, we maintain our sort of long-term thinking position and working through there. Right now, the simplest thing I would say is that the Teamwork Collection and the Service Collection are the vehicles for AI monetization, right? Customers adopting those, driving the usage within those. That is as we've said, that's the vehicle to measure AI monetization and adoption and both are going very, very strongly.

但一如既往，我們會維持長期思維的立場並持續推進。就目前而言，我最簡單的說法是：Teamwork Collection 和 Service Collection 是 AI 變現的載體，對吧？客戶採用這些方案，並在其中推動使用量。正如我們所說，這就是衡量 AI 變現與採用的方式，而兩者進展都非常、非常強勁。

Thank you, and that concludes our question-and-answer session. I will now turn the call over to Mike for closing remarks.

謝謝，問答環節到此結束。我現在把電話交給 Mike 做結語。

Look, thank you, everyone, for joining the call today. As always, we appreciate the thoughtful questions and continued support. A huge shout out to everyone across the Atlassian team, another great quarter into the books and continued long-term thinking and hard work to deliver for our customers. I hope to see as many of those customers and as many of you on the investment community at Team '26 Europe, which is coming up in October. Other than that, have a kickass day.

各位，感謝今天加入這通電話會議。一如既往，我們感謝各位深思熟慮的提問與持續支持。也要向 Atlassian 團隊的每一位致上最誠摯的感謝，又一個精彩的季度告一段落；我們持續以長期思維與努力工作，為客戶交付成果。我希望在 10 月即將到來的 Team '26 Europe，能見到更多客戶，以及投資社群的各位。除此之外，祝各位今天過得很痛快。