Dropbox Inc (DBX) 2026 Q3 法說會逐字稿

Thank you for standing by, and welcome to the Dropbox third-quarter 2026 earnings conference call. (Operator Instructions)

感謝各位耐心等候，歡迎參加 Dropbox 2026 年第三季財報電話會議。(接線員指示)

And now I'd like to introduce your host for today's program, Sarah Schubach, Chief Accounting Officer and Head of Investor Relations. Please go ahead.

現在我想介紹今天節目的主持人，Sarah Schubach，首席會計長兼投資人關係主管。請開始。

Good afternoon, and welcome to Dropbox's third-quarter 2026 earnings call. As a reminder, we will discuss non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings release and our earnings presentation posted on our IR website at investors.dropbox.com.

各位下午好，歡迎參加 Dropbox 2026 年第三季財報電話會議。提醒各位，本次電話會議將討論非 GAAP 財務衡量指標。我們 GAAP 與非 GAAP 結果之間的定義與調節表，可在我們的財報新聞稿與財報簡報中找到，相關資料已發布於我們的投資人關係網站 investors.dropbox.com。

We will also make forward-looking statements on this call, including statements about our future outlook for our fourth-quarter and fiscal year 2026, as well as our expectations regarding fiscal year 2027, our business, assets, strategies, and the macroeconomic environment, and the planned acquisition of Tigris data, which remains subject to the satisfaction or waiver of certain customary closing conditions. Such statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those described.

我們也將在本次電話會議中作出前瞻性陳述，包括關於我們 2026 年第四季與 2026 會計年度的未來展望，以及我們對 2027 會計年度、我們的業務、資產、策略與總體經濟環境的預期，並包括對 Tigris data 的擬議收購；該交易仍須滿足或豁免若干慣常交割條件。此類陳述受已知與未知的風險與不確定性影響，可能導致實際結果與所述內容存在重大差異。

Many of those risks and uncertainties are described in our SEC filings, including our most recent report on Form 10-Q and forthcoming report on Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We disclaim any obligation to update any forward-looking statements except as required by law.

其中許多風險與不確定性已在我們向 SEC 提交的文件中說明，包括我們最近一期的 Form 10-Q 報告以及即將提交的 Form 10-Q 報告。前瞻性陳述僅代表我們在作出該等陳述當日的信念與假設。除法律要求外，我們不承擔更新任何前瞻性陳述的義務。

I will now turn the call over to Dropbox's co-CEO, Ashraf Alkarmi.

接下來我將把電話交給 Dropbox 共同執行長 Ashraf Alkarmi。

Thanks, Sarah, and good afternoon everyone. I’m proud of what this team delivered in Q3. Five months into this role, I’m more convinced than ever that Dropbox is on the right path and this quarter showed why. Q3 was another beat and raise quarter.

謝謝你，Sarah，各位下午好。我為團隊在第三季的交付成果感到自豪。上任五個月以來，我比以往任何時候都更確信 Dropbox 走在正確的道路上，而本季也證明了原因。第三季再次是一個「超預期並上調指引」的季度。

We delivered revenue of $637 million and operating margin of over 42%, both ahead of our guidance, as well as our fifth straight quarter of year-over-year revenue growth, excluding FormSwift. We added approximately 20,000 net new paying users, our fourth consecutive quarter of paying user growth. In our upmarket managed sales motion, North America had its strongest quarter in two years.

我們實現營收 6.37 億美元、營業利益率超過 42%，兩者皆優於我們的指引；同時，在排除 FormSwift 的影響後，營收已連續第五個季度年增。我們新增約 2 萬名淨新增付費用戶，為付費用戶連續第四個季度成長。在我們面向中大型客戶的受管銷售推進上，北美創下兩年來最強勁的一季。

Today we also took two major steps forward. We rolled out AI-powered Dropbox, which now takes action directly on the content customers already keep with us, and we announced DBX Agent Cloud, our new infrastructure business for companies and developers building AI applications. Together with our planned acquisition of Tigris, an already proven player in object storage, both move us forward toward our vision, Dropbox as the trusted independent content foundation for the AI era.

今天我們也向前邁出兩大步。我們推出了 AI 驅動的 Dropbox，現在可直接對客戶已存放在我們這裡的內容採取行動；同時我們宣布 DBX Agent Cloud——我們面向打造 AI 應用的企業與開發者的新基礎設施業務。結合我們擬議收購的 Tigris（已在物件儲存領域證明實力的業者），這些舉措都推動我們朝願景前進：讓 Dropbox 成為 AI 時代值得信賴、獨立的內容基礎。

Last quarter I laid out a three-pillar strategy to get there. First, restore durable growth in our core business through foundational improvements, including bringing AI in context directly into Dropbox. Second, use that stronger foundation to build intelligent workflows that automate critical work for our customers. And third, put the platform we built to work across the broader AI ecosystem.

上季我提出了三大支柱策略以達成該目標。第一，透過基礎性改善恢復核心業務的可持續成長，包括將具情境的 AI 直接帶入 Dropbox。第二，利用更強的基礎打造智慧工作流程，為客戶自動化關鍵工作。第三，讓我們打造的平台在更廣泛的 AI 生態系中發揮作用。

This quarter we made progress on all three. Let me walk you through each.

本季我們在三個方面都取得進展。我將逐一說明。

First, Dropbox now brings AI directly to the content customers already keep with us, and we’ve continued to expand the rollout. This changes what Dropbox is for our customers. It’s no longer just where their work is stored, but it’s where their work gets done.

第一，Dropbox 現在將 AI 直接帶到客戶已存放在我們這裡的內容上，我們也持續擴大推廣範圍。這改變了 Dropbox 對客戶而言的用途。它不再只是存放工作的地方，而是完成工作的地方。

And it does that with the controls that make customers comfortable handing AI a shared folder. Dropbox works within existing permissions, shows its plan before anything changes, and customers can roll back anything with one click. With a new desktop experience, we’re also bringing these capabilities to the many customers who work primarily in their Dropbox folders.

而且它透過各種控制機制，讓客戶放心把共享資料夾交給 AI。Dropbox 會在既有權限範圍內運作，在任何變更發生前先展示其計畫，且客戶可一鍵回復任何內容。透過全新的桌面體驗，我們也把這些能力帶給許多主要在 Dropbox 資料夾中工作的客戶。

Here’s one example. A customer had a catch-all folder of more than 1,200 images with file names that made no sense, and it had simply never gotten organized. Dropbox sorted them into folders in about 10 minutes, and that’s just one use case.

這裡舉一個例子。有位客戶有一個「大雜燴」資料夾，裡面超過 1,200 張圖片，檔名毫無意義，因此一直都沒有整理。Dropbox 在大約 10 分鐘內就把它們分類到各個資料夾中，而這只是其中一種使用情境。

Customers are using Dropbox to find old files and images just by describing them, pull a project’s drawings and documents together into a single PDF, transcribe recordings, create and refine video clips, all from the content they already keep with us. That’s what we mean by one secure home for content and context. Dropbox does the work, but the customer always stays in control.

客戶正在使用 Dropbox，只要描述就能找到舊檔案與圖片；把某個專案的圖面與文件彙整成單一 PDF；將錄音轉寫；建立並精修影片片段——全部都基於他們已存放在我們這裡的內容。這就是我們所說的：內容與情境的一個安全之家。Dropbox 來完成工作，但客戶始終保有控制權。

Early signals are encouraging. Almost 90% of users who received the experience preferred it, and of the teams that have received the new experience, over 20% have started using the AI credits included in their plans.

早期訊號令人鼓舞。將近 90% 獲得此體驗的使用者更偏好它；而在已獲得新體驗的團隊中，超過 20% 已開始使用其方案內含的 AI 點數。

As an independent neutral platform, Dropbox also lets customers choose how they want to bring AI into their work. They can switch models within Dropbox or work on their Dropbox content in ChatGPT, Claude, Gemini, Muse, and other tools.

作為獨立且中立的平台，Dropbox 也讓客戶能選擇他們希望如何把 AI 帶入工作流程。他們可以在 Dropbox 內切換模型，或在 ChatGPT、Claude、Gemini、Muse 等工具中處理其 Dropbox 內容。

Last quarter, 150,000 users had connected Dropbox to ChatGPT and Claude. Today, more than 1.4 million have connected Dropbox to our AI integrations. And that work comes back to Dropbox. Since September, weekly uploads from these AI tools into Dropbox have grown from over 600,000 to more than 2 million. More agents. Better Dropbox.

上季，有 15 萬名使用者已將 Dropbox 連接到 ChatGPT 與 Claude。今天，已有超過 140 萬人將 Dropbox 連接到我們的 AI 整合。而這些工作成果也會回到 Dropbox。自 9 月以來，從這些 AI 工具上傳到 Dropbox 的每週上傳量，已從超過 60 萬成長到超過 200 萬。更多代理。更好的 Dropbox。

Second, we’re bringing intelligent workflows that automate critical work for our customers. Dropbox Replay is a prime example. It’s a video review and approval tool for creative teams, which we rebuilt from the ground up this year as an AI-native application. Teams can review, approve, and deliver work to clients in one place, built on the content they already keep in Dropbox. Next week we’ll begin putting the new Replay in the hands of select customers.

第二，我們正在帶來智慧工作流程，為客戶自動化關鍵工作。Dropbox Replay 就是一個典型例子。它是一款供創意團隊使用的影片審閱與核准工具；我們今年從零開始重建，將其打造為 AI 原生應用。團隊可在同一處完成審閱、核准並交付給客戶，並建立在他們已存放於 Dropbox 的內容之上。下週我們將開始把全新的 Replay 交到部分精選客戶手中。

Third, DBX Agent Cloud, which opens an entirely new growth opportunity for Dropbox. Every company’s need for storage keeps growing, and AI applications and agents are accelerating that demand. The datasets, model artifacts, images, and video they depend on, need a reliable, portable home that works across clouds, works across models and compute environments. And agents need persistent access to their files across tasks and sessions.

第三，DBX Agent Cloud 為 Dropbox 開啟了全新的成長機會。每家公司的儲存需求都在持續成長，而 AI 應用與代理正在加速這股需求。它們所依賴的資料集、模型產物、影像與影片，需要一個可靠、可攜的家，能跨雲運作、跨模型與運算環境運作。而代理也需要在不同任務與工作階段之間，持續存取其檔案。

Keeping content reliable, protected, and accessible at scale is what Dropbox has done for nearly two decades. DBX Agent Cloud brings that expertise to companies and developers, starting with a standalone S3-compatible object storage they can buy independently and use with the tools they already know.

讓內容在大規模下保持可靠、受保護且可存取，是 Dropbox 近二十年來一直在做的事。DBX Agent Cloud 將這項專業帶給企業與開發者，首先推出可獨立購買的獨立式、相容 S3 的物件儲存，並可搭配他們已熟悉的工具使用。

Our planned acquisition of Tigris gets us there faster. Tigris has thousands of customers who depend on it to keep their applications running, and a developer business that would take us substantially longer to build ourselves.

我們擬議收購 Tigris 將讓我們更快達成目標。Tigris 擁有數千名客戶仰賴其維持應用程式運行，並且具備一個開發者業務，若由我們自行打造，將需要顯著更長的時間。

We see four clear ways to make the business stronger. First, capability. When the acquisition is complete, Tigris will bring features built for AI workloads, and we can add Dropbox’s strengths in organizing and governing content. Second, reach. Tigris’ rapid growth has come from developers, and we expect our enterprise sales motion and relationships with larger customers will open up use cases it can’t easily reach on its own today, like backup, disaster recovery, and media archives.

我們看到四個明確的方式可讓這項業務更強。第一，能力。收購完成後，Tigris 將帶來為 AI 工作負載打造的功能，而我們也能加入 Dropbox 在內容組織與治理方面的優勢。第二，觸及範圍。Tigris 的快速成長來自開發者，我們預期我們的企業銷售推進方式以及與大型客戶的關係，將開啟它目前難以單獨觸及的使用情境，例如備份、災難復原與媒體封存。

Third, scale. Demand for Tigris storage is growing quickly, and we plan to bring efficient, exabyte-scale infrastructure to support it. And fourth, the team. The combination would bring together a proven leadership team with a deep developer focus and experience building for AI workloads. Together we can support larger workloads while keeping Tigris’ focus on reliability and developer experience. We expect the acquisition to close in Q4 2026.

第三，規模。對 Tigris 儲存的需求正在快速成長，我們計劃導入高效率、百艾位元組（exabyte）等級的基礎設施來支援這項需求。第四，團隊。這項結合將匯聚一支經驗證的領導團隊，具備深厚的開發者導向，並擁有為 AI 工作負載打造產品的經驗。我們將能在維持 Tigris 對可靠性與開發者體驗的專注下，共同支援更大型的工作負載。我們預期此收購將於 2026 年第四季完成。

And object storage is the entry point, not the destination. Agents need what people need: a place to keep their work, the right permissions, history, and a way to recover when something goes wrong. We plan to add file system and content processing capabilities next, giving applications and agents a persistent place to manage files and continue their work across sessions. Over time we see DBX Agent Cloud becoming a portable content and data foundation for AI applications and agents across models and compute environments.

而物件儲存是入口，而非終點。代理（Agents）需要的與人相同：一個保存工作成果的地方、適當的權限、歷史紀錄，以及在出錯時能復原的方法。我們計劃下一步加入檔案系統與內容處理能力，讓應用程式與代理擁有一個可持續保存的地方來管理檔案，並能跨工作階段延續其工作。隨著時間推進，我們看見 DBX Agent Cloud 將成為跨模型與運算環境、可攜式的內容與資料基礎，支撐 AI 應用程式與代理。

Executing this strategy requires focus. This year, we brought our DAS and Dropbox teams together, rebalanced our go-to-market organization, and stepped back from work outside our priorities. And most recently, as we tightened our focus and shifted resources to align with our new strategy, we made the difficult decision to reduce our workforce by approximately 9%.

執行這項策略需要聚焦。今年，我們將 DAS 與 Dropbox 團隊整合，重新平衡我們的上市推廣（go-to-market）組織，並從優先事項以外的工作中抽身。而最近，隨著我們進一步聚焦並調整資源以配合新策略，我們做出艱難決定，將人力規模縮減約 9%。

I want to thank the colleagues who are leaving us for everything they’ve contributed. Together, these changes put our resources behind the three pillars where we have the strongest right to win and let us invest in our strategy while expanding margins.

我要感謝即將離開的同事，感謝他們所做的一切貢獻。綜合而言，這些變革讓我們把資源投入在三大支柱上——這些是我們最有勝算的領域——並使我們能在擴大利潤率的同時，投資於我們的策略。

Everything we’re doing points in one direction. Dropbox is becoming the trusted independent content foundation for the AI era. For people and teams, that’s AI-powered Dropbox. For companies and developers, it’s DBX Agent Cloud. In both, customers keep control of their content and it stays portable across the AI models, tools, and compute they choose as AI evolves.

我們所做的一切都指向同一個方向。Dropbox 正在成為 AI 時代值得信賴的獨立內容基礎。對個人與團隊而言，那就是 AI 驅動的 Dropbox。對企業與開發者而言，那就是 DBX Agent Cloud。在這兩者中，客戶都能掌控其內容，並且隨著 AI 演進，內容可在其選擇的 AI 模型、工具與運算環境之間保持可攜性。

We’re still early and the measure of this strategy is whether it shows up in growth. That’s the standard I’m holding us to, but Q3 gives me confidence, another beat-and-raise quarter, AI-powered Dropbox in customers’ hands, and an entirely new avenue of growth with DBX Agent Cloud and Tigris. I’m excited about what’s ahead.

我們仍處於早期階段，而衡量這項策略的標準是它是否反映在成長上。這就是我對我們要求的標準；而第三季讓我更有信心——又是一個超預期並上調展望的季度，AI 驅動的 Dropbox 已交到客戶手中，並且透過 DBX Agent Cloud 與 Tigris 開啟了一條全新的成長途徑。我對未來感到振奮。

With that, I’ll turn it over to Ross.

接下來，我把時間交給 Ross。

Thank you, Ashraf. Ashraf laid out our vision for Dropbox to be the trusted, independent content foundation for the AI era. We’re building on our durable storage foundation with AI-powered capabilities that bring intelligence to the content our customers already keep with us, creating a smarter Dropbox experience. And DBX Agent Cloud, accelerated by our planned acquisition of Tigris data, builds on our storage heritage to provide object storage and the content infrastructure, governance, and developer capabilities companies need to build AI applications.

謝謝你，Ashraf。Ashraf 闡述了我們的願景：讓 Dropbox 成為 AI 時代值得信賴、獨立的內容基礎。我們在穩健的儲存基礎上，打造 AI 驅動的能力，為客戶已存放在我們這裡的內容帶來智慧，創造更聰明的 Dropbox 體驗。而 DBX Agent Cloud 在我們計劃收購 Tigris data 的推動下加速發展，延續我們的儲存傳承，提供物件儲存以及企業打造 AI 應用程式所需的內容基礎設施、治理與開發者能力。

Q3 was another beat-and-raise quarter. We delivered our fifth straight quarter of revenue growth, excluding FormSwift, with another quarter of net new paying user growth, and operating margin nearly 4 points ahead of our guidance, and we’re again raising our full-year outlook.

第三季再次是超預期並上調展望的季度。我們在排除 FormSwift 後，實現連續第五個季度的營收成長；付費用戶淨新增也再度成長；營業利益率較我們的指引高出近 4 個百分點；我們也再次上調全年展望。

Our three-pillar growth strategy is taking shape. With the new Dropbox experience and DBX Agent Cloud, supported by our planned acquisition of Tigris, we’re opening an entirely new avenue of growth in object storage built on top of a profitable, cash-generative core business.

我們的三支柱成長策略正在成形。透過全新的 Dropbox 體驗與 DBX Agent Cloud，並在我們計劃收購 Tigris 的支持下，我們正以一個具獲利能力、能產生現金的核心業務為基礎，在其上開闢物件儲存這條全新的成長途徑。

We’re also building a more efficient Dropbox to support it. This year, we consolidated our Dash and core FSS teams as we brought Dash Intelligence into Dropbox. We rebalanced our go-to-market organization in Q3, and we’ve now announced a workforce reduction to further deduplicate resource investments and concentrate more intensely on our three-pillar strategy. Together, these actions allow us to invest behind our strategy while expanding operating margin, and I’ll quantify that impact when I discuss 2027.

我們也在打造更高效率的 Dropbox 以支援這一切。今年，隨著我們將 Dash Intelligence 整合進 Dropbox，我們整併了 Dash 與核心 FSS 團隊。我們在第三季重新平衡上市推廣（go-to-market）組織，並且現已宣布縮減人力，以進一步去除重複的資源投入，並更集中火力在三支柱策略上。綜合而言，這些行動讓我們能在擴大營業利益率的同時，投資於我們的策略；我會在談到 2027 年時量化其影響。

These initiatives are exciting, but they’re early, and we’ll scale our investment in step with the results they deliver. That same discipline shaped how we are approaching our acquisition of Tigris. Tigris is small but growing rapidly. They are already a proven player in object storage, have an established AI developer flywheel, have landed thousands of customers including large AI-native platforms, and have an experienced team already operating in this market.

這些計畫令人振奮，但仍處於早期階段，我們將依其帶來的成果逐步擴大投資。同樣的紀律也形塑了我們對收購 Tigris 的做法。Tigris 規模不大，但成長非常快速。他們已是物件儲存領域經驗證的參與者，建立了成熟的 AI 開發者飛輪，已獲得數千名客戶（包含大型 AI 原生平台），並擁有一支在此市場已具營運經驗的團隊。

This acquisition enables us to accelerate our Pillar 3 infrastructure strategy and capitalize on this emerging opportunity now. With that, let me turn to our Q3 results. Unless otherwise indicated, all income statement figures mentioned are non-GAAP and exclude stock-based compensation, amortization of purchased intangibles, certain acquisition-related expenses, workforce reduction expenses, and net gains on equity investments. Our non-GAAP net income also includes the income tax effect of the aforementioned adjustments.

這項收購使我們能加速第三支柱的基礎設施策略，並把握當前這個新興機會。接下來，我來說明第三季業績。除非另有說明，所有提及的損益表數字皆為非 GAAP，並排除以股份為基礎的薪酬、已購無形資產攤銷、特定併購相關費用、人力縮減費用，以及股權投資的淨收益。我們的非 GAAP 淨利亦包含上述調整項目的所得稅影響。

In Q3, revenue increased 0.5% year over year to $637.3 million. Excluding FormSwift, revenue grew 1% year over year, ahead of the roughly flat growth implied by our guidance. On a constant currency basis, revenue, excluding FormSwift, increased 0.2%. The outperformance was driven primarily by self-serve teams, where simplified pricing and packaging drove stronger upsells.

第三季營收年增 0.5%，達 6.373 億美元。排除 FormSwift 後，營收年增 1%，優於我們指引所暗示的大致持平成長。以固定匯率計算，排除 FormSwift 後的營收增加 0.2%。優於預期主要由自助式團隊（self-serve teams）帶動，簡化的定價與方案組合推動更強勁的升級銷售（upsells）。

Total ARR was $2.555 billion, up 0.8% year over year, and no longer includes any FormSwift ARR following its sunset at the end of the quarter. As a result, ARR declined $10 million sequentially. Excluding FormSwift, ARR grew 1.9% year over year, or 0.4% on a constant currency basis -- excluding FormSwift, ARR grew $5 million sequentially, led by self-serve teams.

總 ARR 為 25.55 億美元，年增 0.8%，且在本季末 FormSwift 結束服務後，ARR 已不再包含任何 FormSwift ARR。因此，ARR 較前一季減少 1,000 萬美元。排除 FormSwift 後，ARR 年增 1.9%，以固定匯率計算為年增 0.4%——排除 FormSwift 後，ARR 較前一季增加 500 萬美元，主要由自助式團隊帶動。

We exited the quarter with 18.2 million paying users, a sequential increase of approximately 20,000 users. Excluding FormSwift, we added approximately 60,000 paying users, led by our Simple plan and self-serve teams. Average revenue per paying user was $138.76, compared to $139.07 in the year-ago quarter, as growth in our lower-priced Simple plan was largely offset by FX tailwinds and a mix shift to monthly plans.

本季末付費用戶為 1,820 萬，較前一季約增加 2 萬名。排除 FormSwift 後，我們約新增 6 萬名付費用戶，主要由 Simple 方案與自助式團隊帶動。每位付費用戶平均營收為 138.76 美元，去年同期為 139.07 美元；其中，較低價格的 Simple 方案成長大致被匯率順風與方案組合轉向月繳方案所抵銷。

Gross margin was 81.8%, up roughly 40 basis points year over year. Infrastructure efficiencies and the move of more of our storage onto our own data centers more than offset the added AI compute costs of rolling out the new Dropbox experience. Operating margin was 42.2%, ahead of our guidance of approximately 38.5%, driven by higher revenue, lower headcount costs following our go-to-market changes, and the timing of marketing spend, which will shift to Q4 alongside the new Dropbox experience launch.

毛利率為 81.8%，年增約 40 個基點。基礎設施效率提升，以及將更多儲存遷移至自有資料中心，已足以抵銷推出新版 Dropbox 體驗所增加的 AI 運算成本。營業利益率為 42.2%，高於我們約 38.5% 的指引，主要受更高營收、上市推廣調整後人力成本下降，以及行銷費用投入時點影響；該行銷投入將隨新版 Dropbox 體驗上線而轉至第四季。

Net income was $197.8 million, compared to $196.7 million in the year-ago quarter. Diluted EPS was $0.88 compared to $0.74 in the year-ago quarter, based on 225.1 million diluted weighted average shares outstanding compared to 265.1 million in the year-ago period.

淨利為 1.978 億美元，去年同期為 1.967 億美元。稀釋後每股盈餘為 0.88 美元，去年同期為 0.74 美元；稀釋後加權平均流通股數為 2.251 億股，去年同期為 2.651 億股。

Cash flow from operations was $259.2 million, compared to $302.1 million in the year-ago period, primarily reflecting higher interest payments related to our term loan facility. Capital expenditures were $7.4 million. Unlevered free cash flow was $303.9 million, and unlevered free cash flow per share was $1.35, up 14% year over year.

營運活動現金流為2.592億美元，較去年同期的3.021億美元下降，主要反映與我們定期貸款融資安排相關的利息支付增加。資本支出為740萬美元。無槓桿自由現金流為3.039億美元，每股無槓桿自由現金流為1.35美元，年增14%。

We ended the quarter with cash and short-term investments of $1.142 billion, and our $400 million revolving credit facility remains undrawn. In Q3, we repurchased approximately 4.7 million shares for approximately $150 million, and we had approximately $1.235 billion remaining under our authorization.

本季末我們的現金及短期投資為11.42億美元，且我們4億美元的循環信用額度仍未動用。在第三季，我們以約1.5億美元回購約470萬股股票，且在授權額度下尚餘約12.35億美元可用。

I’ll now offer our outlook for Q4 and the full-year 2026. Please note that our outlook excludes the announced acquisition of Tigris because we have not yet closed the transaction.

接下來我將提供我們對2026年第四季及2026全年展望。請注意，我們的展望不包含已宣布收購Tigris，因為交易尚未完成交割。

For the fourth quarter of 2026, we expect total revenue to be in the range of $630 million to $633 million. As a reminder, we fully discontinued FormSwift at the end of Q3, so it contributes no revenue in Q4 compared to approximately $3 million in Q3. Excluding FormSwift, our Q4 guidance implies 0.2% year-over-year growth at the midpoint.

對於2026年第四季，我們預期總營收介於6.30億至6.33億美元。提醒一下，我們已於第三季末全面停止FormSwift，因此其在第四季不再貢獻營收，相較之下第三季約貢獻300萬美元。排除FormSwift後，我們第四季指引在中位數隱含年增0.2%。

We are expecting a currency tailwind of approximately $3 million, as we have continued to see FX tailwinds moderate. On a constant currency basis, we expect total revenue to be in the range of $627 million to $630 million. We expect our non-GAAP operating margin to be approximately 40%. This includes a partial quarter’s benefit from the reduction-in-force action announced last week and excludes approximately $22.5 million of expenses related to our workforce reduction.

我們預期匯率將帶來約300萬美元的順風，因為我們持續看到外匯順風趨於緩和。以固定匯率計，我們預期總營收介於6.27億至6.30億美元。我們預期非GAAP營業利益率約為40%。此包含上週宣布的人力精簡措施帶來的部分季度效益，並排除約2,250萬美元與人力縮減相關的費用。

As additional context to our expectations for Q4, when we provided our Q3 guidance, we expected operating margin to step up sequentially from Q3 to Q4. However, as I mentioned, we saw delayed spend in Q3, particularly marketing spend, which we now plan to spend in Q4 as we continue to expand the rollout of our new Dropbox experience. And Q4 reflects only a partial quarter of savings from our workforce reduction. As a result, we expect Q4 operating margin to be slightly lower than Q3.

作為第四季預期的補充背景，當我們提供第三季指引時，我們預期營業利益率將自第三季到第四季逐季提升。然而，如我所提到，我們在第三季看到支出延後，特別是行銷支出；我們現在計畫在第四季支出，因為我們將持續擴大新版Dropbox體驗的推出。此外，第四季僅反映人力縮減帶來的部分季度節省。因此，我們預期第四季營業利益率將略低於第三季。

And finally, we expect diluted weighted average shares outstanding to be in the range of 224 million to 229 million shares.

最後，我們預期稀釋後加權平均流通股數介於2.24億至2.29億股。

For the full-year 2026, we are raising the midpoint of our as-reported revenue guidance range by $12 million, now expecting a range of $2.528 billion to $2.531 billion. Excluding FormSwift, this implies 120 basis points of year-over-year growth at the midpoint. We are expecting a currency tailwind of approximately $28 million. On a constant currency basis, we are raising the midpoint by $15 million, now expecting a range of $2.5 million to $2.503 billion. We expect gross margin to be approximately 81.5%.

對於2026全年，我們將按報告口徑的營收指引區間中位數上調1,200萬美元，現預期介於25.28億至25.31億美元。排除FormSwift後，這在中位數隱含年增120個基點。我們預期匯率順風約2,800萬美元。以固定匯率計，我們將中位數上調1,500萬美元，現預期介於25.00億至25.03億美元。我們預期毛利率約為81.5%。

We are raising our non-GAAP operating margin guidance to 40.5%, up 25 basis points from the midpoint of our prior guidance. Our prior guidance already assumed savings from our workforce reduction, so this increase reflects our Q3 outperformance.

我們將非GAAP營業利益率指引上調至40.5%，較先前指引中位數提高25個基點。我們先前的指引已假設人力縮減帶來的節省，因此此次上調反映我們第三季表現優於預期。

We are also raising our unlevered free cash flow guidance, which we now expect to be at or above $1.075 billion, an increase of $5 million. This accounts for our severance expectations, where we ultimately expect to pay approximately $30 million in severance and benefits as related to our workforce reduction, with roughly $20 million expected to be paid in 2026 and the remainder to be paid in 2027.

我們也上調無槓桿自由現金流指引，現預期達到或高於10.75億美元，增加500萬美元。此已納入我們對資遣的預期：我們最終預計將支付約3,000萬美元的資遣費與福利（與人力縮減相關），其中約2,000萬美元預計於2026年支付，其餘將於2027年支付。

We continue to expect CapEx to be in the range of $20 million to $25 million, and additions to finance lease lines to be approximately 4% of revenue. Finally, we expect diluted weighted average shares outstanding to be in the range of 228 million to 233 million shares, which reflects our 30-day trailing average stock price.

我們仍預期資本支出（CapEx）介於2,000萬至2,500萬美元，且融資租賃新增額度約為營收的4%。最後，我們預期稀釋後加權平均流通股數介於2.28億至2.33億股，反映我們30日追蹤平均股價。

Now I’ll provide additional context as it relates to guidance. We continue to expect positive net new paying user growth for 2026. Regarding gross margin, our Q4 outlook reflects rolling out the new Dropbox experience to all eligible self-serve teams by year-end, which increases AI compute costs, partially offset by the infrastructure efficiencies we have been delivering. We expect some quarter-to-quarter variability as rollout and adoption progress.

接下來我將就指引提供更多背景說明。我們仍預期2026年付費用戶淨新增將為正成長。就毛利率而言，我們對第四季的展望反映：我們將在年底前把新版Dropbox體驗推送給所有符合資格的自助式（self-serve）團隊，這會增加AI運算成本，但部分將被我們持續交付的基礎設施效率所抵銷。隨著推出與採用進展，我們預期各季之間會有一定波動。

And finally, some early thoughts on 2027. We’ll provide formal guidance, including for Tigris, in February. Excluding the acquisition of Tigris, we expect 2027 revenue to be roughly flat.

最後，談一些對2027年的初步看法。我們將於2月提供正式指引，並將包含Tigris。若不計Tigris收購，我們預期2027年營收大致持平。

First, we have now fully discontinued the sale of FormSwift, which contributed $13 million to revenue in 2026 and will thus have no contribution in 2027. Second, the new Dropbox experience and workflow products are early in the rollout, and we believe monetization will build gradually. Finally, we have continued to see FX tailwinds moderate and expect this to continue. We will also be more prepared to guide the contribution from our new initiatives when we have more evidence in February.

第一，我們現已全面停止銷售FormSwift；其在2026年貢獻1,300萬美元營收，因此在2027年將不再有任何貢獻。第二，新版Dropbox體驗與工作流程產品仍處於推出初期，我們相信變現將逐步建立。最後，我們持續看到外匯順風趨於緩和，並預期此趨勢將延續。當我們在2月取得更多證據時，也將更有準備就新計畫的貢獻提供指引。

We expect approximately 150 basis points of operating margin expansion in 2027, reflecting the full-year benefit of the efficiency measures I described, consolidating our Dash and core FSS teams, rebalancing our go-to-market organization, and our recently announced workforce reduction. That expansion is net of reinvesting a portion of those savings behind our three-pillar growth strategy.

我們預期2027年營業利益率將擴張約150個基點，反映我所描述的效率措施在全年發揮效益、整併我們的Dash與核心FSS團隊、重新平衡我們的市場推進（go-to-market）組織，以及近期宣布的人力縮減。該擴張已扣除將部分節省再投資於我們三支柱成長策略的影響。

As a reminder, Tigris has not yet closed and isn’t included in these expectations. If it does close, we may reinvest some of these savings to support it, and we’ll reflect that in our guidance in February.

提醒一下，Tigris尚未完成交割，因此未納入這些預期。若交易完成，我們可能會將部分節省再投資以支持其發展，並會在2月的指引中反映。

Our workforce reduction will also lower stock-based compensation, and beyond that, we continue to look for ways to use our equity more efficiently and bring stock-based compensation down as a percentage of revenue. We expect unlevered free cash flow to expand in line with operating margin, driven by the same efficiency measures.

我們的人力縮減也將降低以股份為基礎的薪酬（stock-based compensation），此外，我們仍持續尋找更有效率使用股權的方式，並將以股份為基礎的薪酬占營收比重降低。我們預期無槓桿自由現金流將隨營業利益率同步擴張，並由相同的效率措施所驅動。

Turning to Tigris, the transaction hasn’t closed yet, and we’ll provide a more fulsome update when we report in February. Tigris is a small but fast-growing business that is not material to our total revenue in the trailing 12-month period. They already serve thousands of customers, including several large companies and AI-native platform players. Tigris runs at a modest burn, and given how quickly this market is moving, we see it as a time-sensitive opportunity to accelerate our time in the market as we extend our infrastructure to sell object storage and power AI workloads.

談到Tigris，交易尚未完成交割，我們將在2月發布財報時提供更完整的更新。Tigris是一家規模不大但成長快速的企業，在過去12個月期間對我們總營收並不具重大性。他們已服務數千名客戶，包括數家大型公司以及AI原生平台業者。Tigris目前以適度的燒錢速度運作；鑑於此市場變化迅速，我們視其為一項具時間敏感性的機會，可加速我們進入市場的速度，並在我們擴展基礎設施以銷售物件儲存（object storage）並支援AI工作負載之際，搶占先機。

As it relates to their capacity needs, we expect our recent and ongoing infrastructure efficiency initiatives will unlock sufficient storage capacity to absorb Tigris’ existing customer footprint and support its expected growth in 2027.

就其容量需求而言，我們預期近期及持續進行的基礎設施效率計畫將釋放足夠的儲存容量，以承接Tigris現有客戶規模，並支援其在2027年的預期成長。

To bring it together, Q3 was another beat-and-raise quarter. With the new AI-powered Dropbox experience, DBX Agent Cloud, and the acquisition of Tigris, we’ve opened an entirely new avenue of growth as a trusted independent content foundation for the AI era, and we’ve rebalanced our resources to invest behind that strategy while expanding margins and cash flow.

總結而言，第三季再次是一個「優於預期並上調指引」的季度。透過全新AI驅動的Dropbox體驗、DBX Agent Cloud，以及收購Tigris，我們作為AI時代值得信賴的獨立內容基礎，開啟了全新的成長途徑；同時我們也重新平衡資源以投資於該策略，並擴大獲利率與現金流。

We still need to demonstrate sustained growth, and that’s the standard we’re holding ourselves to. But we believe we are well positioned to pursue the opportunity ahead.

我們仍需要證明能夠持續成長，而這就是我們對自己所要求的標準。但我們相信，我們已具備良好條件去把握眼前的機會。

With that, operator, please open the line for questions.

接下來，接線員，請開放提問線路。

Brett Huff, Stevens Inc.

Brett Huff，Stevens Inc.。

Awesome. Nice to finally be covering you all. Appreciate the question and congrats on a good quarter. So on Tigris and Agent Cloud, we've gone to a few questions already about the rationale for Dropbox entering the development storage market. Can you kind of walk us through what the customer problem Agent Cloud is built to serve and why existing options kind of fall short of serving those customers today?

太棒了。很高興終於開始追蹤你們。感謝讓我提問，也恭喜你們交出一個不錯的季度成績。關於 Tigris 和 Agent Cloud，我們之前已經問過幾個問題，主要是 Dropbox 進入開發者儲存市場的理由。你能否帶我們了解一下 Agent Cloud 要解決的客戶痛點是什麼，以及為什麼現有選項在服務這些客戶時有些不足？

Thank you for that question. First of all, we're really excited about this acquisition with Tigris. When we look at the opportunity, we see fundamentally that more agents and more AI means more storage needs, whether you're building applications and that's requiring images and videos to training data, to really building any application that has needs for large amounts of storage.

謝謝你的提問。首先，我們對這次與 Tigris 的收購感到非常興奮。當我們看這個機會時，我們從根本上看到：更多代理（agents）與更多 AI，意味著更多的儲存需求——無論你是在開發應用程式，需要從圖片、影片到訓練資料，乃至於任何需要大量儲存的應用。

We see this when we look at our own infrastructure. And when we looked at this opportunity, we found a really incredible synergy here. We have a company with Tigris that has a big developer motion. They have thousands of customers. These customers range from self-serve customers all the way to some of the most known AI companies that use their product.

我們在檢視自家基礎設施時也看到了這一點。而當我們評估這個機會時，我們發現這裡有非常驚人的綜效。Tigris 這家公司具備很強的開發者推動力（developer motion）。他們有數千名客戶。這些客戶從自助式（self-serve）客戶一路到一些最知名、使用其產品的 AI 公司。

And we believe that our combination together with the exabyte scale efficient storage that we have and their capabilities, we can actually provide something really differentiated first and foremost for the companies and developers using it.

我們相信，結合我們具備的 exabyte 規模高效率儲存，以及他們的能力，我們其實能提供一個非常差異化的方案，首先且最重要的是，對使用它的公司與開發者而言。

Tigris has done an incredible job of building something from day one intended for AI for object storage. So they have example features like snapshots and forking. What this means is if you build an AI application, your agents will do things faster with lower latency and better.

Tigris 從第一天起就做得非常出色，打造出一個以 AI 為目的、用於物件儲存（object storage）的產品。因此他們有一些範例功能，例如快照（snapshots）與分叉（forking）。這代表如果你在打造 AI 應用，你的代理（agents）可以以更低延遲、更快速度、更好的方式完成工作。

We also believe in the neutrality. We're a company that really doesn't tie you to any specific cloud. You can take your content anywhere. So when we talk about us being a trusted, independent, secure, reliable content foundation layer for the AI era, we believe that's the way to do it.

我們也相信中立性（neutrality）。我們是一家不會把你綁定在任何特定雲端的公司。你可以把你的內容帶到任何地方。所以當我們談到在 AI 時代成為一個值得信賴、獨立、安全、可靠的內容基礎層（content foundation layer）時，我們相信這就是實現方式。

That's super helpful. We appreciate it. And then if I could just sneak a follow-up in there. So we've seen some -- from some other software companies that SMBs, their customers are getting a little bit more price sensitive. That doesn't seem a good case for you all, but just kind of curious on what you're seeing in the healthier paying user base and what are some of the things getting those paying users over the hump to get on the Dropbox platform?

這非常有幫助。謝謝你們。如果我可以再追問一題。我們看到一些其他軟體公司提到，中小企業（SMB）客戶變得對價格更敏感。你們看起來不像是這種情況，但我想了解你們在較健康的付費用戶基礎上觀察到什麼，以及有哪些因素能幫助這些付費用戶跨過門檻、開始使用 Dropbox 平台？

Is it specifically around AI usage?

是特別跟 AI 使用有關嗎？

Yeah, just what are some of like the AI ROI proof points that customers are seeing where it's an easy button to click to get on Dropbox?

對，就是客戶看到哪些 AI 投資報酬（ROI）的證據點，讓他們很容易就按下按鈕、選擇上 Dropbox？

I think I'll start by just giving a little bit of context. Fundamentally, we believe in giving customers the choice. We believe that you can use Dropbox without AI. You could use Dropbox inside a cloud or ChatGPT or you could use AI within Dropbox. So customer choice for us is first and foremost very important.

我想我先提供一些背景。從根本上，我們相信要給客戶選擇權。你可以在不使用 AI 的情況下使用 Dropbox；你也可以在雲端或 ChatGPT 裡使用 Dropbox；或者你也可以在 Dropbox 內使用 AI。因此，對我們而言，客戶選擇權是第一優先、最重要的。

And we believe that our biggest focus is to deliver value. We see this more of as a value equation. The more we provide meaningful time savings if we provide these SMBs with the ability to build their business or unlock more potential or save time from the mundane tasks, we believe that value translates into monetization. And that's our biggest focus right now, which is how do we actually tangibly take things that are mundane and take time from these SMBs and say we've automated a lot of it with AI and as a result, they find value.

我們也相信，我們最大的重點是交付價值。我們更把這視為一個價值方程式。我們提供越多有意義的時間節省——如果我們能讓這些中小企業有能力建立他們的業務、釋放更多潛力，或把時間從瑣碎工作中解放出來——我們相信這些價值會轉化為變現。而這正是我們目前最大的重點：我們如何把那些瑣碎、耗時的事情具體地拿出來，並告訴這些中小企業：我們已用 AI 自動化了其中很大一部分，因此他們能感受到價值。

So our focus is actually first and foremost on value creation, whether they do that within Dropbox or within ChatGPT and Claude, we believe that we're there to support them.

所以我們的重點首先是創造價值；不論他們是在 Dropbox 內做到，或是在 ChatGPT、Claude 裡做到，我們相信我們都會在那裡支持他們。

It's Ross. Thanks for joining the call. I just wanted to add, we're not able to discern a macro impact on our business right now. And as Ashraf talked about, for us, it's really about execution to our strategy that drives more value for our customers, that allow us to really capitalize on our flywheel, 700 million registered users, 18-plus million paying users, all the sharing, all the monthly actives.

我是 Ross，謝謝加入這通電話。我想補充一下，我們目前無法辨識總體環境（macro）對我們業務造成的影響。如同 Ashraf 所說，對我們而言，關鍵在於把策略執行到位，為客戶帶來更多價值，讓我們能真正發揮飛輪效應：7 億註冊用戶、超過 1,800 萬付費用戶，以及所有分享行為、所有月活躍用戶。

And so it's really incumbent on us to deliver more value on our products to drive more conversion, to drive more upsell, to drive more retention, and that's all part of the strategy. But, seeing some of the macro things that you alluded to, we haven't seen that pattern directly impacting our business.

因此，我們有責任在產品上交付更多價值，以推動更多轉換、更多升級銷售（upsell）、更高留存，而這些都是策略的一部分。但就你提到的一些總體因素而言，我們尚未看到那種模式直接影響我們的業務。

Rishi Jaluria, RBC.

Rishi Jaluria，RBC。

Maybe I want to think about 2027 and kind of the preliminary outlook there with all the moving pieces, appreciate you giving us a little bit of an early look. Maybe can you talk to us about how we should be thinking about contribution from the new AI tooling to ramp, what is the shape of that look like, kind of any new pricing methodology alongside that?

我想把焦點放在 2027 年，以及在所有變動因素之下的初步展望，謝謝你們願意先讓我們提前了解一些。你們能否談談我們應該如何看待新的 AI 工具帶來的貢獻與爬坡（ramp），它的形狀大概會是什麼樣子，以及是否會搭配任何新的定價方法？

Just trying to kind of get comfort around some of the building blocks, obviously you've turned the corner on stopping declines in growth, but obviously looking for kind of the building blocks to get to future growth from here. So maybe walk us through that and I've got a quick follow-up.

我只是想對一些組成要素更有把握；顯然你們已經走過了成長下滑的轉折點，但接下來也在尋找從這裡通往未來成長的基石。所以也許請你們帶我們走一遍，我還有一個簡短的追問。

Of course, and thank you for the question. Maybe I'll start with a little bit of a recap. Last quarter, we talked about our three pillars for growth.

當然，謝謝你的提問。也許我先簡單回顧一下。上個季度，我們談到三大成長支柱。

The first pillar is about continuously working on our foundation, driving that funnel improvement while also adding AI capabilities within Dropbox to unlock more value for our customers. The second pillar is about actually unlocking more workflows for specific customer segments that present themselves in an additional product that they can buy. And the third one is unlocking all the infrastructure we built via APIs for developers and companies.

第一個支柱是持續強化我們的基礎，推動漏斗改善，同時在 Dropbox 內加入 AI 能力，為客戶解鎖更多價值。第二個支柱是為特定客群解鎖更多工作流程，並以他們可以購買的額外產品形式呈現。第三個支柱是透過 API，把我們建立的所有基礎設施開放給開發者與企業。

And today, where we are is with Pillar 1, we're on our fifth consecutive quarter of growth, fourth consecutive quarter of customer growth and we're basically saying we're achieving that consistently. That's part of one pillar. So as you look at the first pillar, we still need to enable AI for the rest of customers and drive adoption and drive value.

而今天我們所處的位置是：在支柱一方面，我們已連續第五個季度成長、連續第四個季度客戶成長，我們基本上是在說我們能夠持續一致地達成。這是其中一個支柱的一部分。因此，當你看第一個支柱時，我們仍需要讓其餘客戶也能啟用 AI，並推動採用、推動價值。

And we're launching the first application of the second pillar, as I mentioned, for select customers next week. And we just announced the acquisition of Tigris to kick off our DBX Agent Cloud.

而如我提到的，我們將在下週為部分客戶推出第二個支柱的第一個應用。同時，我們也剛宣布收購 Tigris，以啟動我們的 DBX Agent Cloud。

Now, when we look at it, we see that we've planted all the seeds for a lot of amazing opportunities that we fundamentally believe will drive growth. So as we look at next year, we're not going to speculate early how these bets will pan out. We do believe in them. We do believe they're going to drive growth.

現在，當我們回頭看，我們看到自己已經種下了許多種子，帶來很多我們從根本上相信能推動成長的絕佳機會。因此，展望明年，我們不會太早去推測這些押注會如何發展。我們確實相信它們。我們也相信它們將推動成長。

But do expect as time goes by, we'll more concrete data to share. But I believe we have the right bets. It's the bets we talked about, and I believe they'll drive growth for the company.

但也請期待，隨著時間推進，我們會有更具體的數據可以分享。不過我相信我們做了正確的押注。這些就是我們談過的押注，而我相信它們會推動公司成長。

And Rishi, it's Ross. I can add a little bit there because you're probably going to ask about 2027. We said early color to think about flat revenue growth. And I just want to give you our philosophy.

還有，Rishi，我是 Ross。我可以在這裡補充一點，因為你可能會問到 2027 年。我們先前給的初步看法是，營收成長大致持平。我也想跟你說明一下我們的理念。

Number one is a reminder that FormSwift will sunset in Q3 and therefore, there'll be $0 of FormSwift in '27. So this year, we had three quarters of FormSwift , next year we'll $0. Also, we're seeing moderating FX tailwinds this year and expect that to moderate further next year, so that will have some headwind.

第一點提醒：FormSwift 會在第三季結束（sunset），因此 2027 年 FormSwift 的營收會是 0 美元。所以今年我們有三個季度的 FormSwift，明年就會是 0。另外，我們今年看到外匯（FX）順風在放緩，預期明年會進一步放緩，因此會帶來一些逆風。

And then I'll just echo what Ashraf said, we're really excited about the strategy, multi-pillared,-vectored approach that gives us confidence that we can execute on growth. And our philosophy is to drive growth and also sustain or improve our margins. And as we talked about, we're giving more margins by talking about adding 150 basis points next year.

接著我也呼應 Ashraf 說的，我們對這個策略非常興奮；這是一個多支柱、多向量的做法，讓我們有信心能夠落實成長。我們的理念是推動成長，同時維持或提升利潤率。如同我們談到的，我們也透過說明明年要增加 150 個基點的方式，提供更多關於利潤率的資訊。

So we just want to stick to our philosophy of the seeds are planted. We're really like where we are, but we don't want to get too excited and carried away on building that in just yet. When we get to February, we'll give you a more formal guidance that's influenced by everything we know at that time.

所以我們想堅持我們的理念：種子已經種下。我們很喜歡目前的位置，但也不想現在就過度興奮、太早把它完全納入模型。等到二月時，我們會基於當時掌握的一切資訊，提供更正式的指引。

All right, very helpful. And then maybe I just -- I'm not sure if you think about that third pillar, can you maybe talk a little bit about -- I get it's early, obviously a really exciting opportunity, but how should you be thinking about what those conversations that you've been having look like? Any potential kind of use cases?

好，非常有幫助。接著也許我想問——我不確定你們怎麼看第三支柱——你能不能談一下：我知道現在還早，顯然這是個很令人興奮的機會，但我們應該如何理解你們目前正在進行的那些對話長什麼樣子？有沒有任何潛在的使用情境（use cases）？

And I know you just announced the Tigris acquisition this morning. Is that something that fits more broadly into that strategy?

另外我知道你們今天早上剛宣布收購 Tigris。這是否更廣泛地符合那個策略？

Yeah. I'll give you a little bit of color, a little bit beyond this specific acquisition and more about the DBX Agent Cloud that we've been talking about.

是的。我可以提供一些補充說明，不只針對這筆特定收購，也談談我們一直在提的 DBX Agent Cloud。

We spent 20 years building infrastructure that is enabling humans to interact, collaborate, govern, secure their content, apply permissions and really do that at scale in a multiplayer way. And what we envision in the future is more agents means they need more of that for them to be able to produce results. And so what we believe is that object storage is the start and what we're really excited about is Tigris' vision and our vision matched.

我們花了 20 年打造基礎設施，讓人們能夠互動、協作、治理並保護他們的內容，套用權限，並且真正以多人協作的方式在規模化情境下運作。我們對未來的想像是：代理（agents）越多，就越需要這些能力，才能產出結果。因此我們相信，物件儲存（object storage）是起點，而我們真正感到興奮的是：Tigris 的願景與我們的願景一致。

When we kicked off these conversations, we believe that object storage is a start and it's a thing that exists, but where we've taken it is a governance layer of content. So your agents don't just store stuff, they're able to know who has permissions to it, who is able to share it, can they share it, apply security on top of it.

當我們開始這些對話時，我們認為物件儲存是一個起點、也是既有的事物，但我們把它延伸成內容的治理層（governance layer）。因此你的代理不只是把東西存起來，它們還能知道誰擁有存取權限、誰能分享、是否可以分享，並在其上套用安全性。

So as you think about the additional layers on top of content, we have a lot of capabilities and APIs that we can add, and it becomes a new flywheel. We think of Tigris, they have thousands of developers. We have APIs that we can offer to these developers and companies, and what it means is more value to these developers by hosting your content in one place. You can now govern it. You can process it. You can edit it, and we believe that that expands the value for Tigris and for ourselves combined.

所以當你思考內容之上的額外層次時，我們有很多能力與 API 可以加上去，這會形成新的飛輪。我們看 Tigris，他們有數千名開發者。我們有 API 可以提供給這些開發者與公司；這代表把內容託管在同一個地方，能為開發者帶來更多價值。你現在可以治理它。你可以處理它。你可以編輯它，而我們相信這會擴大 Tigris 與我們結合後的價值。

Matthew Bullock, Bank of America.

Matthew Bullock，美國銀行（Bank of America）。

Maybe just to start on pricing and packaging, it would be great to get an updated strategy for the new Dropbox experience you're rolling out. And the Agent Cloud, assuming that the Tigris deal closes, can you just help us think through how, if successful, those initiatives could influence user growth, ARPU, and whether or not the best metrics to evaluate the health of the business change it all post-acquisition?

也許先從定價與方案（pricing and packaging）開始：如果能更新一下你們正在推出的新 Dropbox 體驗的策略會很棒。另外在 Agent Cloud 方面，假設 Tigris 交易完成，你能否幫我們梳理一下：如果這些計畫成功，可能會如何影響用戶成長、ARPU（每用戶平均收入），以及在收購後，用來評估公司健康狀況的最佳指標是否會改變？

Yeah. Matt, it's Ross. So first, new Dropbox DBX experience, the way we're looking at, so we've got that rolled out to about 70,000 teams. We expect to roll that out to the majority teams by the end of Q4. We're seeing -- we talked a little bit about what we're seeing in our early metrics, more than 90% are staying in the new experience. You give the option to roll back, but more than 90% are staying in the experience.

好的。Matt，我是 Ross。先談新的 Dropbox（DBX）體驗：我們目前已經把它推送到大約 70,000 個團隊。我們預期在第四季末前，會推送到大多數團隊。我們看到——我們也稍微談過早期指標——超過 90% 的用戶會留在新體驗中。我們提供回退選項，但超過 90% 都選擇留在新體驗。

And we talked about more than 20% began using AI tokens. We've also sold AI token packs. So again, it's early. It's been limited time in market, but we're really pleased with executing to where we want to go.

我們也提到，超過 20% 的用戶開始使用 AI tokens。我們也已經售出 AI token 套裝。所以再說一次，現在還早。上市時間還不長，但我們對於執行進展、以及朝我們想去的方向前進，感到非常滿意。

And the goal there is. I would say twofold. One is as adoption rises and value is proven, we want to sell AI token packs to our customers. And then two, it's one of the biggest upgrades to the Dropbox platform we've had in many years. It enriches the platform alongside other things we're doing like our Replay solution, the Pillar 3 stuff, all the strategy stuff. And we think that that helps accrue to retention as well. So better monetization through selling AI tokens as well as through retention. So those are the big things on the Dropbox experience.

那裡的目標——我會說有兩個。第一，隨著採用率提升、價值被驗證，我們希望向客戶銷售 AI token 套裝。第二，這是 Dropbox 平台多年來最大的一次升級之一。它讓平台更豐富，並與我們正在做的其他事情相輔相成，例如 Replay 解決方案、第三支柱相關內容、以及整體策略。我們認為這也有助於提升留存。所以，一方面透過銷售 AI tokens 帶來更好的變現，另一方面也透過留存帶來效益。以上是 Dropbox 體驗的重點。

And then on Agent Cloud. So this will be a separate flywheel that we're building now and it will be a sales-led motion. We already were out in the market for the last couple of quarters talking to prospects and customers. We actually were able to build pipeline, get POCs. We even signed up a deal that was before we acquired Tigris.

接著談 Agent Cloud。這會是我們現在正在打造的另一個獨立飛輪，而且會是以銷售驅動（sales-led）的推進方式。過去幾個季度我們已經在市場上與潛在客戶和現有客戶交流。我們其實已經能建立銷售管線、進行 POC（概念驗證）。我們甚至在收購 Tigris 之前就簽下了一筆交易。

And our goal there is to now plug Tigris in, which is a rapidly growing company, thousands of customers. It's got some of the large AI platforms like Railway and Fall. It's doing really well and that would be from pricing and packaging is selling object storage, but also we have a vision of adding other services on top to create this file system for agents. And that type of offering would have software on top that would come at a higher gross margin. So that's an overview of what we're looking to do with Agent Cloud and the new Dropbox experience.

我們的目標是把 Tigris 納入其中；Tigris 是一家快速成長的公司，擁有數千名客戶。它服務一些大型 AI 平台，例如 Railway 和 Fall，表現非常好。從定價與方案來看，會先銷售物件儲存，但我們也有願景在其上加入其他服務，為代理打造一個檔案系統。而這類型的產品會在上層提供軟體能力，毛利率會更高。以上是我們對 Agent Cloud 與新 Dropbox 體驗的整體規劃概述。

Really, really helpful. Thank you for laying that all out. And then maybe just one housekeeping question. On the early look at 2027, we're backing out the FormSwift contribution and currency. But could you just help us think through what that implies for kind of an adjusted constant currency growth rate in terms of the early look for Tigris?

非常、非常有幫助。謝謝你把這些都說清楚。接著可能還有一個例行問題。關於 2027 年的初步看法，我們會把 FormSwift 的貢獻與匯率因素剔除。但你能否幫我們理解：就 Tigris 的早期展望而言，這對於調整後的固定匯率（constant currency）成長率大概意味著什麼？

Yeah, I think -- so again, it's an early signal on 2027. And a reminder to everybody, we're not including Tigris acquisition because it hasn't closed. What we said earlier was expect flat growth and we're talking about margin expansion.

好的，我想——再說一次，這只是對 2027 年的早期訊號。也提醒大家，我們沒有把 Tigris 收購納入，因為交易尚未完成。我們先前說的是預期成長持平，同時我們在談利潤率擴張。

And yeah, you got the point about FormSwift and FX. I would say that right now, what we're trying to convey is we're feeling really good about the strategy. We delivered five quarters of growth at FormSwift. We flipped net new paying users to positive. We're growing teams again. We feel really good about that and we feel really good about this multi-pillared strategy and where we can take it.

對，你提到 FormSwift 和 FX 的點很到位。我會說目前我們想傳達的是：我們對策略感覺非常好。我們在 FormSwift 交出了連續五個季度的成長。我們把淨新增付費用戶翻轉為正成長。我們的團隊（teams）業務再次成長。我們對此感覺非常好，也對這個多支柱策略以及我們能把它帶到哪裡，感覺非常好。

But just like right now, this morning we launched Agent Cloud this morning before we launched the new experience. We talked about rolling out replay to a select base of customers. And so we just want to get a few more months down the road into February when we provide formal guidance to be able to bake that in.

但就像現在一樣，今天早上我們在推出全新體驗之前，先行推出了 Agent Cloud。我們談到會先向一部分客戶推出重播（replay）功能。因此，我們希望再往後推進幾個月，到 2 月提供正式指引時，能把這些因素納入考量。

So we'll provide more detail specific to 2027 when we get there. But for now our philosophies are focused on growth and we're focused on sustaining or improving margin. And right now, we're giving an extra 150 basis points of margin next year.

所以，等到時機成熟時，我們會提供更多針對 2027 年的具體細節。但就目前而言，我們的理念聚焦在成長，同時也聚焦於維持或提升利潤率。而目前，我們在明年額外提供 150 個基點的利潤率。

Jaiden Patel, JPMorgan.

Jaiden Patel，摩根大通（JPMorgan）。

On Tigris, you mentioned thousands of customers and rapid growth from developers. Can you help us size the business? And then as you bring it into Dropbox's infrastructure, how should we think about the margin and potentially capital intensity of Agent Cloud relative to the core?

關於 Tigris，你提到有數千名客戶，且來自開發者的成長很快。你能否協助我們衡量這項業務的規模？另外，當你們把它整合進 Dropbox 的基礎設施後，我們應該如何看待 Agent Cloud 相較於核心業務的利潤率，以及可能的資本密集度？

Yeah. Thanks, Jaiden. So I'm just going to be careful about sizing the business just yet again because we're limited what we can say given we haven't closed it yet and we will provide more details appropriately when we get to February.

好的。謝謝你，Jaiden。我現在仍會謹慎談論業務規模，因為在交易尚未完成之前，我們能說的內容有限；我們會在 2 月適當地提供更多細節。

And I'll just reiterate, they are rapidly growing ARR, of course, revenue is catching up to the rapid growth of ARR. There are thousands of customers. They do include some of the big AI natives, Railway and Fall, other big customers, lots of AI native customers. So we're very impressed with the solution that they put in market., the developer motion that's targeted AI native is a differentiated technology and how it accelerates our Pillar 3 storage infrastructure strategy, and we'll provide more on how to think about growth in ARR when we get to February.

我再重申一次，他們的 ARR 正在快速成長；當然，營收也正在追趕 ARR 的快速成長。客戶有數千家。其中包括一些大型 AI 原生公司，例如 Railway 和 Fall，以及其他大型客戶，還有許多 AI 原生客戶。因此，我們對他們推向市場的解決方案印象非常深刻；其面向 AI 原生的開發者推動模式（developer motion）是一項差異化技術，也能加速我們「支柱 3」的儲存基礎設施策略。我們會在 2 月進一步說明如何看待 ARR 的成長。

On the second part of the question, which was around capital intensity. The other thing we were very impressed is there's about a little bit over 20 employees and very low burn on the business. And as we look at bringing it in, we do look at investing in areas. We think this is a big market opportunity. We want to capture that market opportunity. So we think about investing more. But even in the aggregate with that additional investment, we still think it will be a modest burn.

至於問題的第二部分，也就是資本密集度。我們同樣非常印象深刻的是，他們員工大約略多於 20 人，且業務燒錢率（burn）非常低。當我們考慮把它納入後，也會評估在某些領域加大投資。我們認為這是一個很大的市場機會。我們希望把握這個市場機會。因此我們會考慮投入更多。但即使把這些額外投資合併來看，我們仍認為燒錢會是溫和的。

And the good news is all the work that we've done on our infrastructure has freed up pretty significant capacity. So we have the capacity just kind of sitting there that we want to go out and sell. It's like gold in there that everybody wants and we want to get out there and sell it.

好消息是，我們在基礎設施上所做的所有工作，已釋放出相當可觀的產能。所以我們有一些產能就這樣閒置著，而我們希望把它拿出去賣。那裡面就像黃金一樣，人人都想要，而我們也想把它推出去賣掉。

And so what I'm trying to say is we don't have to go out like next year and go procure a bunch of new capacity like a new startup would. We have significant scale. We have significant excess capacity that we can put to work to support Tigris' growth.

所以我想表達的是，我們不必像新創公司那樣，明年就得出去採購一大堆新的產能。我們具備相當的規模。我們有大量的閒置產能，可以投入運用來支援 Tigris 的成長。

George Kurosawa, Citi.

George Kurosawa，花旗（Citi）。

Great. I'm on for Steve Anders. Yeah, I want to dig in on this Pillar 3 strategy. We rewind the clock earlier in this year. This was very much exploratory at the time. Obviously, since then, if this acquisition goes through, I believe this will be the largest acquisition you guys have done. So if you could just talk through the work that you guys have done, the signals you've seen that give you the confidence to make this move?

很好。我代替 Steve Anders 提問。我想深入了解這個「支柱 3」策略。如果把時間倒回今年稍早，當時這更多還是在探索階段。顯然，自那之後，如果這筆收購案順利完成，我相信這將會是你們做過最大的一筆收購。所以你們能否談談你們做了哪些工作、看到了哪些訊號，讓你們有信心做出這個決定？

Thanks for the question. So earlier, we talked about how we're exploring this. What we did was we actually built pipeline. We talked to customers. We actually built proof of concepts with some of these customers with the capabilities that we have.

謝謝你的問題。所以先前我們談到我們正在探索這件事。我們實際做的是建立銷售管線（pipeline）。我們與客戶交流。我們也確實用我們現有的能力，和其中一些客戶一起建立了概念驗證（POC）。

And just to be clear, Tigris has a lot more capabilities when it comes to object storage because they built it from scratch for AI-native developers. And so despite not having that, we were able to build a pipeline. We built proof of concepts. We actually got inbound requests from customers, and we got really excited.

先說清楚，Tigris 在物件儲存方面有更多能力，因為他們是為 AI 原生開發者從零打造的。因此，儘管我們當時還沒有那些能力，我們仍然能建立銷售管線。我們做了概念驗證。我們也確實收到客戶主動上門的需求，這讓我們非常興奮。

Obviously, there was a lot of really good signal that got us to believe that's the right move for us. But also as we started exploring further the Tigris data, what we found is they have built an incredible business with -- like we talked about, it's growing. And they have both a developer motion and an upmarket motion where they have large customers as well.

顯然，有很多非常正面的訊號讓我們相信，這對我們而言是正確的行動。同時，當我們進一步探索 Tigris 的數據時，我們發現他們打造了一個令人驚豔的業務——如同我們所說，它正在成長。而且他們同時具備開發者推動模式與向上市場（upmarket）推動模式，也就是他們也擁有大型客戶。

So the synergy here is, A, we're able to add APIs to the developer motion and unlock more potential out of infrastructure and APIs already built. Two, they wanted to go more and do more in GTM and upmarket and we have a team that can help with that. And third, we have capacity like Ross mentioned, that is sitting there that we freed up from all our efficiency work and we believe we can unlock that and drive growth with that.

因此，這裡的綜效在於：第一，我們能把 API 加到開發者推動模式中，並釋放既有已建置的基礎設施與 API 的更多潛力。第二，他們希望在 GTM（Go-to-Market）與向上市場方面做得更多，而我們有團隊可以協助。第三，如 Ross 提到的，我們有因效率提升工作而釋放出、目前閒置的產能；我們相信可以把它釋放出來並以此推動成長。

And so when we looked at it was a match and a really big kind of synergy from what we're seeing and what they're seeing and we got excited both about the signal we got and looking at the data they have.

因此，當我們審視這件事時，這是一個契合點；從我們看到的與他們看到的角度來看，都有非常大的綜效。我們對所獲得的訊號，以及他們所擁有的數據，都感到興奮。

Okay, that's great color. And then as a follow-up, if we say, and this is a little early, we haven't closed the deal yet, but if we just think about maybe qualitatively. What signals or milestones would you be looking to see kind of gates that you'd be passing through that would give you confidence to continue investing behind this opportunity versus maybe where this isn't the right fit for the business going forward? What are the gate posts you guys are looking to see as we progress to next year?

好的，這些說明很有幫助。接著追問一下：如果我們說——而且現在談可能還有點早，因為交易尚未完成——但若以定性角度來看。你們會想看到哪些訊號或里程碑，作為你們會通過的「關卡」，讓你們有信心持續加碼投資這個機會；相對地，在哪些情況下可能代表這並非未來對公司而言的合適方向？隨著我們邁向明年，你們在觀察哪些關鍵檢核點？

I think the answer is pretty simple. We want to see growth from this pillar. We believe in it. We believe growth will come. We believe we'll be able to unlock both a self-serve developer motion and an upmarket motion. Sell not just one API, but many. And we believe that that's the signal. We want to see growth come from that pillar.

我認為答案很簡單。我們希望看到這個支柱帶來成長。我們相信它。我們相信成長會到來。我們相信我們能同時解鎖自助式的開發者推動模式與向上市場推動模式。不只賣一個 API，而是賣很多個。而我們相信這就是訊號。我們希望看到成長來自這個支柱。

Thank you. This does conclude the question-and-answer session of today's program. I'd like to hand the program back to Sarah Schubach for any further remarks.

謝謝。今天節目的問答環節到此結束。我想把節目交回給 Sarah Schubach，請她做進一步的結語。

Thanks, everyone, for joining us today. Looking forward to speaking with you next quarter.

謝謝各位今天加入我們。期待在下個季度與各位交流。

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.

各位女士、先生，感謝您參與今天的會議。本節目到此結束。您現在可以中斷連線。祝您有美好的一天。