Serve Robotics Inc. (SERV) 2026 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by. My name is Desiree, and I will be your conference operator today. At this time, I would like to welcome everyone to Serve Robotics, Inc., second-quarter 2026 financial results and conference call. (Operator Instructions) I would now like to turn the call over to Steve Webb.

各位女士、先生，感謝您稍候。我是 Desiree，今天將擔任本次電話會議的接線員。此刻，我謹代表大家歡迎各位參加 Serve Robotics, Inc. 2026 年第二季財務業績與電話會議。(接線員指示) 現在我想將電話交給 Steve Webb。

Thank you, Operator. Welcome to Serve Robotics' second-quarter 2026 earnings call. With me today are Serve's co-founder and CEO, Ali Kashani; and our CFO, Brian Read.

謝謝，接線員。歡迎參加 Serve Robotics 2026 年第二季財報電話會議。今天與我一同出席的有 Serve 的共同創辦人兼執行長 Ali Kashani，以及我們的財務長 Brian Read。

During today's call, we may present both GAAP and non-GAAP financial measures. If needed, a reconciliation of GAAP and non-GAAP measures can be found in our earnings release filed earlier today. Certain statements in this call are forward-looking statements. You should not place undue reliance on forward-looking statements. Actual results may differ materially from these forward-looking statements, and we do not undertake any obligation to update any forward-looking statements we make today, except as required by law.

在今天的電話會議中，我們可能會同時呈現 GAAP 與非 GAAP 財務衡量指標。如有需要，GAAP 與非 GAAP 指標的調節表可見於我們今日稍早提交的財報新聞稿中。本次電話會議中的部分陳述屬於前瞻性陳述。您不應過度依賴前瞻性陳述。實際結果可能與這些前瞻性陳述存在重大差異；除法律要求外，我們不承擔更新今日所作任何前瞻性陳述的義務。

For more information about factors that may cause actual results to differ materially from forward-looking statements, please refer to the press release we issued today as well as the risks and uncertainty described in our most recent annual report on Form 10-K, as supplemented by our most recent quarterly report on Form 10-Q, and in our other reports and filings made with the SEC.

如需了解可能導致實際結果與前瞻性陳述存在重大差異的因素，請參閱我們今日發布的新聞稿，以及我們最近一期 Form 10-K 年度報告（並由最近一期 Form 10-Q 季度報告補充）中所描述的風險與不確定性，並參閱我們向 SEC 提交的其他報告與文件。

We published our quarterly financial press release and our updated corporate presentation to our investor relations website earlier today, and we ask you to review those documents if you haven't already.

我們已於今日稍早在投資人關係網站發布本季財務新聞稿以及更新後的公司簡報，並請您若尚未閱讀，務必查閱這些文件。

With that, let me hand it over to Ali.

接下來，讓我把時間交給 Ali。

Thank you, Steve, and good afternoon, everyone. We have important updates to share with you today. First, I want to give you an update about our Uber partnership and then share our Q2 results and update our full year 2026 guidance. We will discuss what took place in Q2 that has led to the new guidance and also what we are investing in and some of the exciting updates that are coming down the pipe. Let's start with the Uber partnership.

謝謝你，Steve，各位下午好。今天我們有一些重要更新要與各位分享。首先，我想就我們與 Uber 的合作關係提供最新進展，接著分享第二季業績並更新我們 2026 全年財測指引。我們將討論第二季發生了哪些事情導致新的指引，也會談到我們正在投入的項目，以及一些令人振奮、即將推出的更新。我們先從 Uber 的合作開始。

From the first quarter of 2022 through the first quarter of this year, delivery volume through Uber grew for 17 consecutive quarters. In Q2, that trend reversed for the first time. This was caused by lower than expected robot utilization.

從 2022 年第一季到今年第一季，透過 Uber 的配送量連續 17 個季度成長。在第二季，這個趨勢首次反轉。其原因是機器人使用率低於預期。

While customer and merchant demand has remained steady, and our fleet performance has been improving, we believe the reversal in Q2 was largely due to the changes in the operating model and the integration between the two companies. Our extensive discussions with Uber since the emergence of this trend in Q2 have clarified that we really have differing views about the operating model to scale our shared autonomous fleet.

雖然消費者與商家的需求維持穩定、我們的車隊表現也在改善，但我們認為第二季的反轉主要是由於營運模式的變更，以及兩家公司之間的整合所致。自第二季出現此趨勢以來，我們與 Uber 進行了廣泛討論，已釐清我們對於擴大雙方共享自動化車隊的營運模式確實存在不同看法。

This includes things like fleet coordination or merchant integration. Our experience across partners shows that having alignment on integration and operating models can really produce better outcomes from the same underlying technology and fleet. Case in point, in the same timeframe, we saw deliveries with another food delivery partner grow nearly 50% in a single quarter.

這包括例如車隊協調或商家整合等事項。我們在不同合作夥伴的經驗顯示，若能在整合與營運模式上達成一致，確實能在相同的底層技術與車隊條件下產生更好的成果。舉例來說，在同一期間，我們與另一家外送合作夥伴的配送量在單一季度內成長了將近 50%。

So based on the volume decline and some of these recent broader discussions with Uber, we don't currently expect that it would make sense to renew our agreements when it expires in early 2027. That's unless we can improve the operating model meaningfully.

因此，基於配送量下滑以及近期與 Uber 的一些更廣泛討論，我們目前不預期在協議於 2027 年初到期時續約會是合理的選擇。除非我們能在營運模式上做出有意義的改善。

This assessment was formed very recently and we are sharing it with you promptly. We will continue to engage with Uber and we are open to finding a path to continue working together, but I do want to be transparent with you about the conclusion that we have reached, at least with the information that we have today.

這項評估是在非常近期才形成的，我們也即時與各位分享。我們將持續與 Uber 互動，並對於尋找一條能繼續合作的路徑保持開放；但我也希望就我們目前在現有資訊下所得到的結論，對各位保持透明。

Ultimately, we need to focus our resources where we see the clearest path to high utilization and operational leverage so that we can unlock the most value for the communities that we serve. We believe that Serve will be in a stronger position because we'll be able to allocate our resources to stronger, more beneficial partnerships and activities that we think will generate better long-term returns and values for the company.

最終，我們需要把資源聚焦在我們看見最清晰路徑、能達到高使用率與營運槓桿的地方，從而為我們所服務的社區釋放最大價值。我們相信 Serve 將處於更強的位置，因為我們能把資源配置到更強、更具效益的合作關係與活動上，我們認為這些將為公司帶來更好的長期回報與價值。

I want to take a moment and say Uber has been a really important anchor partner for Serve. It's been a real privilege working with a company that has really reshaped urban transportation. Together we helped validate this category and build critical operating experience, and we've scaled our fleet to a level that matters. Working with Uber was a great way to bootstrap our platform, and we really value that partnership and what it enabled over the last 5 years.

我想花點時間說，Uber 一直是 Serve 非常重要的核心合作夥伴。能與一家真正重塑城市交通的公司合作，是一種榮幸。我們共同協助驗證了這個類別、建立了關鍵的營運經驗，並把我們的車隊規模擴大到具有實質意義的水準。與 Uber 的合作是我們啟動平台的絕佳方式，我們非常珍視這段合作關係，以及它在過去 5 年所帶來的助益。

Now this brings me to the financial impact of what changed in Q2. We did not meet our expectations in Q2 given that the delivery volume had declined. As a result, we are materially reducing our full-year revenue guidance. Our Q2 revenue was $3.2 million.

接下來談到第二季變化所帶來的財務影響。由於配送量下滑，我們第二季未達到原先的預期。因此，我們將大幅下調全年營收指引。我們第二季營收為 320 萬美元。

That's a 9% increase sequentially over Q1 and over 400% increase year over year. It is however below the level that we required to support our prior outlook. As such, we are lowering our full year 2026 revenue guidance from $26 million to a range of $9 to $10 million.

這較第一季季增 9%，且年增超過 400%。然而，這仍低於支撐我們先前展望所需的水準。因此，我們將 2026 全年營收指引由 2,600 萬美元下調至 900 萬至 1,000 萬美元的區間。

And we are matching this revenue update with real cost discipline across the second half operating and capital expenses. Brian will dig into that with you in more detail shortly. The principal driver of this change is the removal of the delivery volume growth that we had assumed for the second half.

同時，我們也以嚴格的成本紀律來配合這次營收更新，涵蓋下半年營運費用與資本支出。Brian 稍後會就此為各位做更詳細的說明。這項變動的主要驅動因素，是我們移除了原先假設下半年會出現的配送量成長。

The Q2 results no longer support that expected ramp, so we have removed it from our outlook. Even the Q2 results as well as our success in diversifying our revenue, Uber represented a limited share of our Q2 revenue.

第二季的結果已不再支持原先預期的加速爬坡，因此我們已將其自展望中移除。即便如此，從第二季結果以及我們在營收多元化方面的成功來看，Uber 在我們第二季營收中所占比重其實有限。

The magnitude of the guidance change, therefore, reflects the removal of a substantial expected future ramp not the loss of a large existing revenue stream. Let me explain why we view this as a disciplined portfolio decision and not something that changes our conviction in automating last-mile delivery. Our revenue base is already diversified across delivery, advertising and hospital robotics.

因此，指引調整幅度之所以如此之大，反映的是移除了一項原先預期的、相當可觀的未來爬坡，而非失去一條龐大的既有營收來源。我來解釋為何我們將此視為一項有紀律的投資組合決策，而不是會動搖我們對自動化最後一哩配送的信念。我們的營收基礎已在配送、廣告與醫院機器人等領域實現多元化。

As mentioned, our other delivery marketplace channel, DoorDash, grew nearly 50% sequentially last quarter. I'm also happy to share that we'll be announcing another major delivery marketplace partner in the coming weeks.

如前所述，我們另一個外送平台通路 DoorDash 在上季的配送量季增將近 50%。我也很高興分享，我們將在未來幾週宣布另一家主要的外送平台合作夥伴。

Our advertising revenues accounted for nearly 50% of our robotic food delivery revenues last quarter. This is despite the headwinds of geopolitical macro pressure on advertising spending. And last but not least, our hospital robotics business continues to generate contracted recurring revenue at attractive margins.

我們的廣告收入在上季占我們機器人餐飲外送收入的近 50%。即便面臨地緣政治宏觀壓力對廣告支出造成的逆風，仍是如此。最後但同樣重要的是，我們的醫院機器人業務持續以具吸引力的利潤率產生已簽約的經常性收入。

So far this year, we have signed 7 multi-year contract extensions with our hospital customers and 2 new hospitals, demonstrating continued customer demand for our health care automation platform. We are therefore making a disciplined portfolio decision.

截至今年目前為止，我們已與醫院客戶簽署 7 份多年期合約展延，並新增 2 家新醫院，展現客戶對我們醫療自動化平台的需求持續存在。因此，我們正在做出一項嚴謹的投資組合決策。

We are a platform now operating at meaningful scale. We are allocating fleet capacity, capital and operating attention toward opportunities with clearer demand signals, higher expected utilizations and attractive unit economics and stronger alignment of visions and incentives. We believe this is going to make us a stronger company in the long term.

我們如今是一個已在具意義的規模下運作的平台。我們正將車隊產能、資本與營運關注度配置到需求訊號更清晰、預期使用率更高、單位經濟效益更具吸引力，且願景與誘因更一致的機會上。我們相信，這將在長期使我們成為更強大的公司。

We already have several commercial and product initiatives underway that broaden our distribution, increase our merchant accessibility, expand our direct demand and improve the capabilities of our autonomy platform. These initiatives were in development long before Q2 as part of our plan diversification strategy, and now they're starting to bear fruit.

我們已經有多項商業與產品計畫在推進中，可擴大我們的分銷、提升商戶可及性、擴增我們的直接需求，並強化我們自動駕駛平台的能力。這些計畫早在第二季之前就已作為我們方案多元化策略的一部分在開發中，而現在正開始開花結果。

On August 17, we plan to provide our 2026 summer announcement, which includes updates across 4 areas: a new delivery marketplace partnership, 2 new market launches, a merchant integration products and new technology advances. Based on our analysis of our markets, back of house integration requirements really constrain a significant portion of otherwise addressable restaurant order volume.

8 月 17 日，我們計畫發布 2026 年夏季公告，將涵蓋 4 個面向的更新：一項新的外送市集合作夥伴關係、2 個新市場上線、商戶整合產品，以及新的技術進展。根據我們對市場的分析，後場（back of house）整合需求確實限制了原本可觸及的餐廳訂單量中的相當大一部分。

Almost 2/3 of delivery orders in our operating areas can't benefit from robotic last-mile delivery due to back of house integration barriers. So our new product, Beacon, is a standalone countertop device that connects customers and restaurants with Serve Robotics directly. Because it has its own cellular connectivity, it requires nothing from the restaurant beyond a consistent source of power.

在我們營運區域內，將近 2/3 的外送訂單因後場整合障礙而無法受益於機器人最後一哩外送。因此，我們的新產品 Beacon 是一款獨立的檯面裝置，可讓顧客與餐廳直接與 Serve Robotics 連結。由於它具備自有的行動網路連線，餐廳除了穩定的電源供應外不需要提供任何其他條件。

That means we are not dependent on a restaurant's internet or existing point-of-sale system anymore, which has historically been a source of integration friction across the industry. With Beacon, our aim is to work with most restaurants regardless of the infrastructure, including merchants that aren't connected to third-party delivery platforms. This is really exciting.

這表示我們不再依賴餐廳的網路或既有的 POS（銷售點）系統；而這在過去一直是整個產業整合摩擦的來源。透過 Beacon，我們的目標是不論基礎設施如何，都能與大多數餐廳合作，包括未連接第三方外送平台的商戶。這真的令人振奮。

Later this fall, we also expect to introduce an additional product designed to expand direct customer demand and broaden the types of goods and use cases that our network can serve. We are reimagining how things move around cities, not just food from restaurants, but anything from anywhere to anyone.

今年秋季稍晚，我們也預期推出另一項產品，旨在擴大直接的顧客需求，並拓寬我們網路可服務的商品類型與使用情境。我們正在重新想像城市中物品如何流動——不只是餐廳的食物，而是任何東西從任何地方送到任何人手上。

Beyond our solutions for merchants and customers, we are also in late stages of exciting new partnerships that we believe can support materially higher robot utilization. That means the same robots generating multiple times the value for our partners and customers. In recent months we are seeing inbound interest from major companies across a range of industries from food services to logistics and beyond.

除了面向商戶與顧客的解決方案外，我們也正處於令人振奮的新合作夥伴關係的後期階段，我們相信這些合作能實質提升機器人的使用率。這意味著同一批機器人能為我們的合作夥伴與顧客創造數倍的價值。近幾個月，我們看到來自多個產業的大型公司主動洽詢，範圍從餐飲服務到物流等更多領域。

We are advancing a number of commercial programs designed around denser order allocation, simpler merchant integration and materially higher seed utilization. We will announce each program as it reaches the appropriate contractual and launch milestones.

我們正在推進多項商業計畫，聚焦於更密集的訂單分配、更簡化的商戶整合，以及實質更高的種子使用率。我們將在各項計畫達到適當的合約與上線里程碑時逐一對外公告。

Of course, I can't really share a preview of announcements without talking about our technology. We also have an announcement coming later this year about our autonomy stack. We wanna highlight some major milestones we have achieved in creating new powerful AI models that are making our robots safer, faster, smarter and more reliable and more capable than ever before. So let me leave you with 4 points. First, our previous guidance assumed continuous growth in Uber delivery volume during the second half of the year.

當然，如果不談我們的技術，我其實無法預告這些公告。我們也將在今年稍晚就我們的自動駕駛技術堆疊（autonomy stack）發布一項公告。我們想強調在打造新的強大 AI 模型方面所達成的一些重大里程碑，這些模型正讓我們的機器人比以往更安全、更快速、更聰明、更可靠，也更有能力。所以我用 4 點作結。第一，我們先前的指引假設 Uber 外送量在下半年會持續成長。

The ramp we expected did not materialize in Q2, and we have removed it from our outlook. This was primarily caused by changes in the operating model and integration of our fleet in this particular partnership and not because of any sudden decrease in customer demand or our delivery quality.

我們預期的爬坡並未在第二季實現，因此我們已將其自展望中移除。這主要是由於在此特定合作夥伴關係中，我們的營運模式與車隊整合方式有所變更所致，而非因顧客需求突然下降或我們的配送品質出現問題。

Second, Uber has been a really important anchor partner in building Serve, but absent a meaningful change in the operating model, we do not currently believe that we will renew the agreement after it expires in early 2027. We'll continue working constructively with Uber, of course, and remain open to a new path. Third, our conviction in last-mile autonomy is higher than ever given the diverse traction we are realizing.

第二，Uber 一直是建立 Serve 的重要錨定合作夥伴，但若營運模式沒有出現具意義的改變，我們目前不認為會在協議於 2027 年初到期後續約。當然，我們會持續以建設性的方式與 Uber 合作，並對新的可能路徑保持開放。第三，鑑於我們正在實現的多元牽引力，我們對最後一哩自動化的信念比以往更強。

The momentum we are seeing across multiple delivery channels really reinforces that distribution, merchant integration and fleet operating models are really central to utilization and economics. And finally, we now have 2,000 robots distributed across more than 40 cities nationwide, a diversified revenue base, more than $240 million in liquidity at the end of Q2 and multiple commercial and product initiatives already underway.

我們在多個配送通路所看到的動能，確實強化了這樣的觀點：分銷、商戶整合與車隊營運模式是使用率與經濟效益的核心。最後，我們目前在全國超過 40 個城市部署了 2,000 台機器人，擁有多元化的收入基礎，第二季末流動性超過 2.4 億美元，且多項商業與產品計畫已在推進中。

This includes a new delivery marketplace partnership, a new product that help us reach more merchants, our continued momentum in hospital robotics, and much more. We are focusing our platform and our capital on opportunities with the clearest path to higher utilization, attractive unit economics and durable growth. With the partnerships and initiatives in the pipeline, we feel really good about the potential for revenue to scale, and we will update you in due course as we continue executing on our roadmap.

這包括一項新的外送市集合作夥伴關係、一款可協助我們觸及更多商戶的新產品、我們在醫院機器人業務的持續動能，以及更多內容。我們正將平台與資本聚焦於具備最清晰路徑可提升使用率、單位經濟效益具吸引力且可持續成長的機會。隨著合作夥伴關係與各項計畫在管線中推進，我們對收入擴張的潛力感到非常樂觀；在我們持續執行路線圖的同時，也會在適當時機向各位更新。

We want to bring the value of last-mile autonomy to more customers and merchants faster and share more of that value with them with compelling economics for all involved. This is a more focused route to the same large ambition we've always had: building last-mile autonomy that redefines urban logistics.

我們希望更快地把最後一哩自動化的價值帶給更多顧客與商戶，並在對所有參與者都具吸引力的經濟效益下，與他們分享更多價值。這是一條更聚焦的路徑，通往我們一直以來同樣宏大的目標：打造能重新定義城市物流的最後一哩自動化。

With that, let me hand it over to Brian.

接下來，我把時間交給 Brian。

Thank you, Ali. Good afternoon, everyone. This was a pivotal quarter for us, and I want to explain how it's reflected in the numbers and how we are running the business. I want to frame this as 2 things, a strategic update and the financial update. Strategically, we're positioning our offerings to pass more value directly to merchants and customers.

謝謝你，Ali。各位下午好。這一季對我們而言是關鍵的一季，我想說明這如何反映在數字上，以及我們如何經營這項業務。我想用兩個面向來說明：策略更新與財務更新。在策略上，我們正調整我們的產品與服務，將更多價值直接傳遞給商戶與顧客。

That reduces our reliance on any single delivery channel, and it opens the door to other partnerships Ali described. Eventually, we're evaluating and updating our cost base to match, which I'll walk through in a moment. Together, these give us levers to run the business efficiently.

這降低了我們對任何單一配送通路的依賴，也為 Ali 所描述的其他合作夥伴關係打開大門。最終，我們也在評估並更新我們的成本基礎以相互匹配，我稍後會說明。綜合而言，這些都讓我們擁有能有效率經營業務的槓桿。

Our priorities for the year have not changed: make each robot more productive, grow revenue per robot and per hour, grow the recurring part of our revenue and turn all of that into a stronger financial model. Total revenue for Q2 was $3.2 million compared to $3 million in Q1 and up 400% year over year.

我們今年的優先事項沒有改變：讓每台機器人更具生產力、提升每台機器人與每小時的收入、提高我們收入中經常性部分的占比，並將這一切轉化為更強健的財務模型。第二季總收入為 320 萬美元，相較第一季的 300 萬美元增加，年增 400%。

That total reflects 2 very different trends underneath it. Delivery revenue declined meaningfully in Q2 compared to Q1. That was a real in-quarter decline, not just a slower future ramp. At the same time, our other channels grew enough to more than offset it. Total revenue was still up sequentially, daily active robots were held steady and software revenue was once again nearly $1 million.

該總額反映出其下方兩個截然不同的趨勢。與第一季相比，第二季的配送營收明顯下滑。這是當季內的實際下滑，而不只是未來爬坡速度放緩。同時，我們其他通路的成長幅度足以且超過抵消這一影響。總營收仍呈現季增，每日活躍機器人維持穩定，而軟體營收再次接近100萬美元。

All of this highlights the diversification we've built beyond food delivery. Recurring revenue was over 50% of total revenue this quarter. Our health care business continues to deliver contracted, multi-year revenue at strong margins, and that mix shift is the real contributor to where we see this business heading. Gross loss for the quarter was approximately $8.8 million, and gross margin was negative 271%. I'd point to one thing in particular.

這些都凸顯了我們在餐飲外送之外所建立的多元化布局。本季經常性收入占總營收超過50%。我們的醫療照護業務持續以強勁毛利率貢獻已簽約的多年期收入，而這樣的組合轉移，正是我們所看到此業務發展方向的真正推動力。本季毛損約為880萬美元，毛利率為負271%。我想特別指出其中一點。

Fleet gross margin improved sequentially even as we absorbed the Uber decline Ali just walked through. That tells us this progress is coming from real cost discipline and operational efficiency, not from a clean gross quarter.

即使我們吸收了Ali剛剛說明的Uber下滑影響，車隊毛利率仍較前一季改善。這告訴我們，這項進展來自真正的成本紀律與營運效率，而不是因為某一季剛好出現「乾淨」的毛利表現。

I want to reinforce that this remains our focus. We believe the path to gross margin positivity is inevitable. More revenue per robot per operating hour, improved operational productivity and a growing mix of recurring software and platform revenues.

我想再次強調，這仍是我們的重點。我們相信，走向毛利轉正的路徑是必然的。包括：每台機器人在每個營運小時創造更多收入、提升營運生產力，以及經常性軟體與平台收入占比持續提高。

GAAP operating expenses were $57.3 million in Q2, excluding stock-based compensation of $14.7 million and amortization of intangible assets and acquisition-related expenses of $2.2 million. Non-GAAP operating expenses were approximately $40.4 million. R&D remains our largest investment area, and that's by design. GAAP R&D expense was $20.3 million, or $14.9 million excluding stock-based compensation. This goes towards autonomy development, AI model training, fleet software, data infrastructure and integrating across our platform.

第二季GAAP營業費用為5,730萬美元，其中不含1,470萬美元的股份基礎給付，以及220萬美元的無形資產攤銷與併購相關費用。非GAAP營業費用約為4,040萬美元。研發仍是我們最大的投資領域，這是刻意為之。GAAP研發費用為2,030萬美元；若不含股份基礎給付則為1,490萬美元。這些投入用於自動化（autonomy）開發、AI模型訓練、車隊軟體、資料基礎設施，以及跨平台整合。

G&A expense was $24.8 million, or $14.2 million non-GAAP, operations expense was $7.9 million, or about $7.4 million non-GAAP. Sales and marketing expense was $4.3 million or about $3.9 million non-GAAP. Our investment philosophy is anchored in ensuring that every dollar goes toward revenue quality, margin improvement and platform differentiation. GAAP net loss for the quarter was $64 million, or negative $0.80 per share. Non-GAAP net loss was $47.1 million, or negative $0.59 per share.

G&A費用為2,480萬美元；非GAAP為1,420萬美元；營運費用為790萬美元；非GAAP約為740萬美元。銷售與行銷費用為430萬美元；非GAAP約為390萬美元。我們的投資理念以確保每一美元都用在提升收入品質、改善毛利率與強化平台差異化為核心。本季GAAP淨損為6,400萬美元，每股虧損0.80美元。非GAAP淨損為4,710萬美元，每股虧損0.59美元。

Capital expenditures were approximately $1 million in the quarter before the benefit from approximately $3.6 million in tariff refunds received in the period. We ended the quarter with more than $240 million in cash and marketable securities.

本季資本支出約為100萬美元（尚未計入本期收到約360萬美元關稅退稅的效益）。季末我們持有超過2.4億美元的現金及有價證券。

That's a real advantage and enables us to make the decisions we made this quarter from a position of strength. Turning to our outlook, we're revising our full year 2026 revenue guidance to approximately $9 million to $10 million, down from the $26 million we guided earlier this year. Even at this revised level, we expect annual revenues to grow nearly 3.5x year-over-year.

這是一項真正的優勢，使我們能以更強勢的立場做出本季所做的決策。接著談展望，我們將2026年全年營收指引下修至約900萬至1,000萬美元，低於我們今年稍早所給出的2,600萬美元指引。即使在此修正後的水準，我們仍預期年度營收將年增接近3.5倍。

Two things drove this update. First, the delivery decline wasn't just a future impact as our Q2 results already reflect the decline. Second, and larger, our prior guidance assumed a substantial increase in Uber delivery volume in the second half.

這次更新主要由兩個因素驅動。第一，配送下滑不只是未來的影響，因為我們第二季的結果已反映出這一下滑。第二，而且影響更大的是，我們先前的指引假設下半年Uber配送量會大幅增加。

That ramp is not materializing, and we've removed it from our outlook entirely. To be clear, our fleet size hasn't changed, and our robots can be deployed to direct merchant relationships, other verticals and the new delivery marketplace platform Ali mentioned, which we expect to have more to share on soon.

這樣的爬坡並未出現，我們已將其完全自展望中移除。需要說明的是，我們的車隊規模並未改變，我們的機器人可部署至直接商戶合作、其他垂直領域，以及Ali提到的新配送市集平台；我們預期很快會分享更多資訊。

Each of these channels represents an opportunity to reallocate that capacity toward better economics. Turning to advertising, even with the same macro pressure on ad budgets that Ali referenced, it accounted for nearly half of our robotic food delivery revenues this quarter and at an attractive margin.

上述每一個通路都代表一個機會，讓我們把這些產能重新配置到更好的經濟效益上。再談廣告，即使面臨Ali所提到的廣告預算同樣的總體壓力，本季廣告仍占我們機器人餐飲外送營收近一半，且毛利率具吸引力。

Our healthcare business continues to be durable, recurring revenue, with higher margin, high retention and longer-term contracts. We're taking action on cost and capital to match this revised plan. We've reduced planned 2026 capital expenditures from about $25 million down to approximately $15 million to $17 million.

我們的醫療照護業務持續展現韌性，屬於經常性收入，毛利率更高、留存率高，且合約期限更長。我們正採取成本與資本面的行動，以符合這份修正後的計畫。我們已將原規劃的2026年資本支出由約2,500萬美元下調至約1,500萬至1,700萬美元。

And we're lowering our 2026 non-GAAP operating expense outlook from $160 million to $170 million down to approximately from $140 million to $150 million. Reducing our OpEx guidance gives us the runway to execute our updated plan.

同時，我們也將2026年非GAAP營業費用展望由1.6億至1.7億美元，下調至約1.4億至1.5億美元。下修OpEx指引，讓我們有足夠的跑道來執行更新後的計畫。

You should expect this discipline to show up increasingly through the second half of the year. It's coming from a few places: headcount discipline, optimized deployment infrastructure spend and tighter discretionary spent. Overall, we're narrowing investment to what directly supports autonomy performance, utilization, recurring revenue and gross margin improvement.

你們應該會預期，這種紀律在下半年會愈來愈明顯地反映出來。它來自幾個方面：人力編制的紀律、優化部署基礎設施支出，以及更嚴格的可自由裁量支出控管。整體而言，我們將投資收斂到能直接支持自動化表現、使用率、經常性收入與毛利改善的項目。

We're also working through the Diligent Robotics integration, looking for opportunities to consolidate overlapping G&A and shared services across the combined company. One thing we are deliberately investing in is core autonomy and software. These are the areas that most directly improve robot productivity, customer outcomes and long-term unit economics, which will remain our focus.

我們也正在推進與Diligent Robotics的整合，尋找機會整併合併後公司在G&A與共享服務上的重疊部分。我們刻意加碼投資的一項重點是核心自動化與軟體。這些領域最能直接提升機器人生產力、客戶成果與長期單位經濟效益，並將持續是我們的重點。

This includes our next-generation autonomy platform, which we expect to meaningfully improve per-unit economics and expand the geography our fleet can serve. We are not stepping back from the core business.

其中包括我們的下一代自動化平台，我們預期它將顯著改善單位經濟效益，並擴大我們車隊可服務的地理範圍。我們並未從核心業務退縮。

We're executing with tighter prioritization and a clear focus on operating leverage because we believe that's the right way to build the business and protect long-term shareholder value. Let me close here. The trend we saw in Q2 led us to make hard, clarifying decisions since the quarter closed. concentrating our capital where demand is the clearest and taking a hard look at low-margin, low-control channels to reduce the cost base to match. The near-term revenue number is lower because of that.

我們以更嚴謹的優先順序與清晰的營運槓桿聚焦來執行，因為我們相信這是打造業務並保護長期股東價值的正確方式。我在此做個結語。我們在第二季看到的趨勢，促使我們在季末後做出艱難但釐清方向的決策：把資本集中在需求最明確的地方，並嚴格檢視低毛利、低可控性的通路，以相應降低成本基礎。因此，短期營收數字會更低。

Serve is building a robotics platform, not a single-use delivery fleet. Even with this update, we expect revenue to grow meaningfully this year, and the revenue base underneath that number is more diversified, higher quality and higher margin than it was a year ago. The investments we are making, anatomy, software and our proprietary data are the ones that compound and provide long-term enduring value. We believe that is what turns our early lead in physical AI into a durable operating and financial model.

Serve正在打造的是一個機器人平台，而不是單一用途的配送車隊。即使有這次更新，我們仍預期今年營收將顯著成長，而支撐該數字的營收基礎，相較一年前更加多元、品質更高且毛利率更佳。我們正在投入的項目——自動化、軟體以及我們的專有資料——是能夠累積複利並提供長期持久價值的資產。我們相信，這正是把我們在實體AI（physical AI）的早期領先優勢，轉化為可持續的營運與財務模型的關鍵。

With that, we will open the lines for Q&A.

接下來，我們將開放電話線進行問答。

(Operator Instructions) Michael Latimore, Northland Capital Markets.

(接線員指示) Michael Latimore，Northland Capital Markets。

Ali, can you just elaborate on -- a little bit more on just like what happened in the second quarter? Why do you think utilization was not hitting goals? Was there some specific events that occurred? Maybe Uber stopped allocating resources to you for some reason? Was there any discussion or concern over rev share? Maybe a little more elaboration would be great.

Ali，你能否再多說明一下——第二季到底發生了什麼事？你認為為什麼使用率沒有達到目標？是否有發生某些特定事件？例如 Uber 是否因為某些原因停止把資源分配給你們？是否有針對分潤（rev share）進行任何討論或產生疑慮？如果能再多補充一些會很棒。

Yes, Mike. Yes, it's a good question. I mean -- but needless to say, I want to be respectful of partner confidentiality and all that. But I think primarily -- I mean, first thing I would say, there are a lot of investments happening on both sides into the platform. So Uber is investing in theirs, we are investing in ours.

是的，Mike。是的，這是個好問題。我的意思是——但不用說，我希望尊重合作夥伴的保密性等等。不過我認為主要——我的意思是，第一點我會說，雙方都在平台上進行大量投資。所以 Uber 在投資他們的平台，我們也在投資我們的平台。

And at the same time, a lot of decisions are being made about how do you allocate orders, how do you organize fleets? Who makes which decisions? Like there're a lot of very interesting problems to solve when you start introducing autonomous fleets and mixing them with human fleets.

同時，也有很多決策正在進行：如何分配訂單、如何組織車隊？由誰來做哪些決定？當你開始導入自動化車隊並把它與人力車隊混合運作時，會有很多非常有趣的問題需要解決。

So these are, I think, part of the process that's now taking place because these things are real and they're on the streets and doing their jobs. And when it came to those kind of decisions, I think we were not perfectly aligned and the decline kind of highlighted that it brought those discussions to surface, so we decided we should be transparent with everybody that we don't believe that this would be kind of renewed.

所以我認為這些是目前正在發生的流程的一部分，因為這些事情是真實的——它們已經在街上工作、在執行任務。而當涉及這類決策時，我認為我們並沒有完全一致；這次下滑某種程度上凸顯了這點，也把相關討論帶到檯面上，因此我們決定對大家保持透明：我們不認為這會以這種方式續約。

But of course, we are still engaged and we will continue talking. And if there are better paths, we would look for it.

但當然，我們仍在互動，也會持續對話。如果有更好的路徑，我們也會尋找。

And then in terms of just market reach going forward, I mean, you talked about DoorDash, sounds like there is another kind of overlay app you're going to be working with. I guess, which channel will be more important here? Will it be the DoorDash and this new company partnership,, or will it be direct to merchants? This direct to merchants, how do you kind of reach the end customer and market to them effectively?

另外，就未來的市場觸及而言，我的意思是，你提到 DoorDash，聽起來你們還會與另一個類似覆蓋式（overlay）的 App 合作。我想問，哪個通路會更重要？會是 DoorDash 和這個新公司的合作夥伴關係，還是直接面向商家？若是直接面向商家，你們要如何觸及終端客戶並有效地向他們行銷？

Yes, look, we want to have -- as I said, we want to have control over our destiny, which means we don't want to be kind of in the wake turbulence of bigger partners all the time. And I think the use of, for example, Beacon, the product that I mentioned, to be able to really work with any restaurant is one of those pieces.

是的，你看，我們希望——如我所說，我們希望掌握自己的命運，這代表我們不想一直處在大型合作夥伴尾流擾動（wake turbulence）的影響之下。我認為例如我提到的 Beacon 這個產品，能夠真正與任何餐廳合作，就是其中一個關鍵拼圖。

It's an investment we've been making for a while, so we're going to bring that and highlight that soon. But it's a really exciting product. There are other things in the pipeline in that same order.

這是我們投入了一段時間的投資，所以我們很快會把它推出並強調它。但這真的是一個令人興奮的產品。同一條產品線上還有其他項目正在開發中。

But we're also working with partners. So I mentioned DoorDash. It's not just that we grew 50% with them in Q1, in fact, between June and July, we grew another 50%. So it's doing really well. We would keep investing there, but we would also want to have channels that allows us to work with anybody directly and honestly enable new types of use cases.

但我們也在與合作夥伴合作。所以我提到 DoorDash。不只是我們在第一季與他們合作成長了 50%，事實上，在 6 月到 7 月之間，我們又再成長了 50%。所以表現非常好。我們會持續在那裡投資，但同時也希望擁有能讓我們直接與任何人合作的通路，並坦白說，能夠支援新的使用情境。

It's something that I think if we are just working with platforms we won't be able to do. I think we are much more incentivized to want to innovate and try new things. We are more incentivized to want to pass a lot of the value to restaurants and customers directly. So I think it's really important that we really invest in that. And I think the decision we made this quarter is going to allow us to put energy into that.

我認為如果我們只與平台合作，就做不到這件事。我認為我們有更強的動機去創新、去嘗試新事物。我們也更有動機把大量價值直接回饋給餐廳與消費者。所以我認為在這方面加大投資非常重要。而我認為我們本季做出的決定，將讓我們能把精力投入到這裡。

And then just last one, so any guidance on software revenue in the second half? And also, is Diligent still on track for about $7 million this year?

最後一個問題，下半年軟體營收有任何指引嗎？另外，Diligent 今年是否仍按計畫約 700 萬美元？

Yes, Mike. So obviously you would have seen the updated outlook for revenue at the $9 million to $10 million number. We saw strength in Q2 for the software continuing. I think in the back half of the year, right, we're going to see the continued headwinds that we observe in the back half of Q2. So from a software standpoint, it will be a little softer in the back half, but the strength of Diligent, as you mentioned, will continue.

是的，Mike。所以很明顯你已經看到我們更新後的營收展望，為 900 萬到 1,000 萬美元。我們看到第二季軟體業務持續走強。我認為在下半年，對吧，我們會看到我們在第二季後半段觀察到的逆風持續存在。因此就軟體而言，下半年會稍微疲弱一些，但如你所提到的，Diligent 的強勁表現會持續。

So we're not going to get into the split between Diligent, but they're definitely helping the positive mix, right, in the margin story from a top line perspective. So happy to provide some more clarity there, but I think software will be little softer as we get throughout the back half of the year.

所以我們不會進一步拆分 Diligent 的占比，但它們確實在正向的組合（mix）上提供幫助，對吧，從營收端的角度來看，對毛利敘事是有利的。我們也樂於提供更多清晰度，但我認為隨著進入下半年，軟體會稍微疲弱一些。

Colin Rusch, Oppenheimer & Company.

Colin Rusch，Oppenheimer & Company。

Can you talk a little bit about the KPIs that you're looking at for the efficiency of the autonomy and how we should think about the cadence of learning that you guys have been able to engage in here in the first half?

你能否談談你們用來衡量自動化效率的 KPI，以及我們應該如何看待你們在上半年能夠進行學習的節奏（cadence）？

Yes. Great question. I think that's one of the areas that we really want to share more in the next couple months, hopefully. We talked about our new generation autonomy. You all know about the acquisition we made in that space.

是的。好問題。我認為這是我們希望在接下來幾個月能分享更多的領域之一。我們談到我們的新一代自動化。各位也知道我們在那個領域做了收購。

Ultimately, the most important KPI is are the robots getting better, are getting -- they're getting faster, they're safer than anything else out there, they are more reliable and they get the delivery done in such a way that it allows us to grow our revenue and also improve margins. So that's the top line.

最終，最重要的 KPI 是：機器人是否變得更好——是否變得更快、比市面上任何其他方案都更安全、更可靠，並且能以一種讓我們得以提升營收、同時改善毛利的方式完成配送。所以這是最核心的。

Once we share more about the new generation autonomy, we would have more color on what we are tracking and how do we measure progress. But there's a lot of areas of investment I'm really excited about. This new type of autonomy architecture with end-to-end models where you have a lot less Lossiness in the information where the models are making a lot smarter decisions, intuitive decisions.

一旦我們分享更多關於新一代自動化的內容，我們就能更具體說明我們在追蹤什麼，以及如何衡量進展。但有很多投資領域讓我非常興奮。這種新的自動化架構，採用端到端模型，資訊傳遞的損失（lossiness）更少，模型能做出更聰明、更直覺的決策。

I think it's a pretty exciting area that we are going to talk about more. And we are already seeing the improvements in the fleet as a result. So we'll kind of share how we are looking at it again soon.

我認為這是一個相當令人興奮的領域，我們之後會談更多。而且我們已經因而在車隊上看到改善。所以我們很快會再分享我們如何看待這件事。

And then with the hospital robots and Diligent, can you talk a little bit about the potential for accelerating deployments and potential optimization of the design of those bots in kind of a timeframe on how we should think about both of those dynamics?

接著談到醫院機器人以及 Diligent，您能否稍微談談在加速部署方面的潛力，以及在這些機器人的設計優化方面的潛力？也請就我們應該如何看待這兩項動態的時間框架，分享一下看法。

Yes, so in a way, and I think I've mentioned this in the past, it felt to us that Diligent was basically a clutch behind we were at Serve in terms of being ready for scale. So you make the robots work, you prove the model, you prove it with real customers, then you want to cost it down, make the hardware reliable, so that you can scale rapidly.

是的，所以某種程度上，而且我想我過去也提過，對我們而言，Diligent 基本上就像是我們在 Serve 這邊、就規模化準備程度而言的「離合器」在後面跟著。也就是說，你先讓機器人能運作、驗證商業模式、用真實客戶來證明，接著你會想要把成本降下來、讓硬體更可靠，如此才能快速擴張。

And that's kind of when we got to our 2,000 robots, only after we made those investments. That's exactly what Diligent is doing right now with the new hardware that they're working on. So again, we are going to have a lot more to say about this very soon, but that's the kind of investment that has taken place in the last year, effectively, to position us to really push that growth in the following years.

而我們大概就是在做完那些投資之後，才達到 2,000 台機器人的規模。這正是 Diligent 目前正在做的事——他們正在開發新的硬體。所以我們很快會就此分享更多內容，但過去一年實際上所進行的就是這類投資，目的是讓我們在接下來幾年能真正推動成長。

Jeffrey Cohen, Ladenburg Thalmann.

Jeffrey Cohen，Ladenburg Thalmann。

I guess firstly, could you talk about the advertising business and advertising revenue a little bit and drill into that as far as are you wrapping the robots and some of the customers, are they more local or more national? And any pricing on that associated with the robots and any kind of forward-looking statements as far as what you would expect for the back half of the year?

我想先請您談談廣告業務與廣告營收，並深入說明一下：你們是否會在機器人上做包膜（wrap），以及客戶是偏在地還是偏全國性？另外，與機器人相關的任何定價資訊，以及對於今年下半年你們預期的任何前瞻性看法？

Yes, I'll let Brian chime in on some parts of the questions. I think on the advertising, we're seeing pretty good traction. I mentioned earlier in the year, after the war began, we noticed some softness in advertising spending, but we really worked hard to make up for it. That's why something like 50% of our sidewalk revenue came from advertising and it's looking really good. We are very kind of satisfied with what we're seeing.

好的，我先請 Brian 針對部分問題補充。就廣告而言，我們看到相當不錯的進展。我在年初提過，戰爭爆發後，我們注意到廣告支出有些轉弱，但我們確實很努力把缺口補回來。這也是為什麼我們的人行道（sidewalk）營收大約有 50% 來自廣告，而且目前看起來非常好。我們對目前看到的狀況感到相當滿意。

To answer your question, we see both local and national campaigns right now. So -- and again, this is something that I think you're going to hear about a lot soon. And Brian can take the rest of the question.

回答您的問題，我們目前同時看到在地與全國性的活動。所以——而且我想這也是您很快會常聽到的內容。接下來其餘問題讓 Brian 回答。

Yes, Jeff. So I think with pricing, right, our teams have been doing a great job with the expanded fleet of looking at how we can do multi-city campaigns or multi-market, multi-neighborhood campaigns as well. So the opportunity continues to grow, and I think that's excellent from a top-line perspective for margin and the product mix that we're seeing as we move forward.

是的，Jeff。我想在定價方面，我們的團隊隨著車隊規模擴大，做得非常好，正在研究我們如何執行跨城市活動，或跨市場、跨社區的活動。因此機會持續成長，從營收端（top-line）、毛利以及我們往前看到的產品組合來看，我認為這非常正面。

To answer your direct question, right, they're mostly wraps. We're also seeing some growth within the experiential side of the business and a lot of inbound interest from clients that are interested in having the robots for various experiences. For a split on the revenue, not something we get into for guiding to advertising versus deliveries, but we'll anchor back to the updated guidance talked about in the script.

直接回答您的問題，目前主要是包膜（wraps）。我們也看到體驗式（experiential）業務面向有所成長，並且有很多客戶主動洽詢，希望在各種體驗活動中使用機器人。至於營收拆分，我們不會在指引中細分廣告與配送的占比，但我們會回到講稿中提到的更新後指引作為依據。

Okay. That's helpful. And then as a follow-up, I know you had some news out last month on NoScrubs and some introductions of non-food. So any update there that you could talk about at this point in time as far as any non-food updates, perhaps stuff to think about and geographies to think about? I appreciate it.

好的。這很有幫助。接著追問一下，我知道你們上個月有關於 NoScrubs 的消息，也有一些非食品品類的導入。所以目前有沒有任何更新可以分享，例如非食品的進展、值得關注的事項或地理區域？謝謝。

Yes, so one of them that we announced recently was NoScrubs. This is laundry. It's kind of an obvious application. The product that I mentioned earlier, Beacon, it really allows us to work with any merchant. Any place that wants to work with robots would be to do that without any backend integration barriers the way we've had to deal with so far.

好的，我們最近宣布的其中一個就是 NoScrubs。這是洗衣服務。這算是相當直觀的應用。我先前提到的產品 Beacon，確實讓我們能與任何商家合作。任何想要使用機器人的地方，都可以在不需要後端整合障礙的情況下做到——而這正是我們迄今必須處理的問題。

I think that's one of the pieces of the puzzle that we've been working on to be able to really open the kind of use cases that we can do. Because ultimately we are not building robots to just deliver burritos. This is last-mile infrastructure for cities.

我認為這是我們一直在努力拼上的其中一塊拼圖，讓我們能真正擴大可做的使用情境。因為歸根結底，我們打造機器人不是只為了送墨西哥捲餅。這是城市的最後一哩路基礎設施。

And part of why we want to be able to put our resources beyond things like just food marketplaces is to open up those opportunities. I mentioned there's another product as well that we'll be announcing that actually allows customers to also reach out to us directly.

而我們之所以希望把資源投入到不僅僅是食品平台這類領域之外，是為了打開那些機會。我也提到我們還有另一個產品將會宣布，它實際上也能讓消費者直接聯繫我們。

So you can -- if you have direct integration ability with merchants and with customers, I think that those are the building blocks to open a lot of interesting other use cases. Now, I would say I think those things take time because food is something you eat 3 times a day.

所以你可以——如果你具備與商家以及與消費者的直接整合能力，我認為這些就是打開許多其他有趣使用情境的基石。不過我會說，這些事情需要時間，因為食物是你一天吃三次的東西。

There's a lot of habits and demands for it that's already formed. We have to make these investments to open up these new use cases and opportunities. And that's what exactly we are doing. We want to kind of make the investments so that over time these things become a part of our, in a way, portfolio mix.

對它的習慣與需求早已形成。我們必須進行這些投資，才能開啟新的使用情境與機會。而這正是我們正在做的事。我們希望進行這些投資，讓這些項目隨著時間推移，某種程度上成為我們投資組合（portfolio）結構的一部分。

(Operator Instructions) There are no more further questions at this time. Ladies and gentlemen, that concludes today's call. Thank you all for joining in. You may now disconnect.

（接線生指示）目前沒有其他問題。各位女士、先生，今天的電話會議到此結束。感謝各位參與。您現在可以掛線。