Roper Technologies, Inc. (ROP) 2026 Q2 法說會逐字稿

Good morning. The Roper Technologies conference call will now begin. Today's call is being recorded. (Operator Instructions)

早安。Roper Technologies 的電話法說會現在開始。今天的通話將被錄音。(接線員指示)

I would now like to turn the call over to Zack Moxcey, Vice President, Investor Relations. Please go ahead.

現在我想把電話交給投資人關係副總裁 Zack Moxcey。請開始。

Good morning, and thank you all for joining us as we discuss the second quarter financial results for Roper Technologies. Joining me on the call this morning are Neil Hunn, President and Chief Executive Officer; Jason Conley, Executive Vice President and Chief Financial Officer; Brandon Cross, Vice President and Chief Accounting Officer; and Shannon O'Callaghan, Senior Vice President of Finance.

早安，感謝各位加入我們，一同討論 Roper Technologies 第二季財務結果。今天早上與我一同參與電話會議的有：總裁兼執行長 Neil Hunn；執行副總裁兼財務長 Jason Conley；副總裁兼首席會計長 Brandon Cross；以及財務資深副總裁 Shannon O'Callaghan。

Earlier this morning, we issued a press release announcing our financial results. The press release also includes replay information for today's call. We prepared slides to accompany today's call, which are available through the webcast and are also available on our website.

今天稍早，我們發布新聞稿公告財務結果。新聞稿也包含今日電話會議的重播資訊。我們準備了投影片搭配今天的電話會議，可透過網路直播取得，也可在我們的網站上取得。

Now if you please turn to page 2. We begin with our Safe Harbor statement. During the course of today's call, we will make forward-looking statements, which are subject to risks and uncertainties as described on this page, in our press release and in our SEC filings. You should listen to today's call in the context of that information.

現在請翻到第 2 頁。我們先從「安全港」聲明開始。在今天的電話會議過程中，我們將作出前瞻性陳述，這些陳述受本頁、我們的新聞稿以及向 SEC 提交文件中所述之風險與不確定性影響。請在該等資訊的脈絡下收聽今天的電話會議。

And now please turn to page 3. Today, we will discuss our results primarily on an adjusted non-GAAP basis. For the second quarter, the difference between our GAAP results and adjusted results consists of the following items: amortization of acquisition-related intangible assets, financial impacts associated with our minority investment in Indicor, and lastly, cash collected on Outgo's beneficial interest, which is the residual amount owed to Outgo after it sells receivables to a third party and is classified within cash flows from investing activities. Reconciliations can be found in our press release and in the appendix of our presentation on our website.

接著請翻到第 3 頁。今天我們將主要以調整後的非 GAAP 基礎來討論結果。就第二季而言，我們 GAAP 結果與調整後結果之差異包含以下項目：併購相關無形資產攤銷、與我們對 Indicor 的少數股權投資相關之財務影響，以及最後，Outgo 受益權（beneficial interest）所收取的現金；該受益權係 Outgo 將應收帳款出售予第三方後，Outgo 仍應收取的剩餘金額，並分類於投資活動之現金流量。對帳表可在我們的新聞稿以及網站上簡報的附錄中找到。

Now, if you please turn to page four, I'll hand the call over to Neil. After our prepared remarks, we will take questions from our telephone participants. Neil?

現在請翻到第 4 頁，我把電話交給 Neil。在我們的準備發言之後，我們將接受電話與會者的提問。Neil？

Thank you, Zack. And thanks to everyone for joining us this morning. You can see our agenda on page 4. So let's get right into it. Next slide, please.

謝謝你，Zack。也感謝各位今天早上加入我們。各位可以在第 4 頁看到我們的議程。那我們就直接開始。請下一張投影片。

As we turn to page 5, I want to highlight three takeaways for today's call. First, we delivered solid results and are raising our outlook for the year. Second, AI momentum continues to build across the enterprise at an accelerating pace. And third, we continue to execute our capital allocation opportunities with our long-standing discipline. Let me double-click each point.

翻到第 5 頁，我想強調今天電話會議的三個重點。第一，我們交出穩健的成果，並上調全年展望。第二，AI 動能持續在全企業加速累積。第三，我們持續以長期以來的紀律執行資本配置機會。我逐點說明。

First, second quarter results were solid and ahead of expectations. Total revenue grew 9%, organic revenue grew 5%, EBITDA grew 5%, and free cash flow grew 11%. Importantly, enterprise gross retention remained strong consistently in the mid-90s. On that foundation, enterprise software bookings remained solid with core bookings up mid-single digit plus on a TTM basis.

第一，第二季結果穩健且優於預期。總營收成長 9%，有機營收成長 5%，EBITDA 成長 5%，自由現金流成長 11%。重要的是，企業整體的總留存率（gross retention）持續強勁，穩定維持在 90% 中段。在此基礎上，企業軟體訂單（bookings）仍然穩健，以過去十二個月（TTM）計算，核心訂單成長為中個位數以上。

At a high level and across the portfolio, Deltek had an encouraging second quarter, but we're not ready to call a turn in the GovCon market. Relative to Neptune, they had better first half performance than originally expected. And finally, at DAT, we're seeing improving freight market conditions for the first time in many years.

從整體與投資組合角度來看，Deltek 第二季表現令人鼓舞，但我們尚未準備好宣稱 GovCon 市場已出現反轉。至於 Neptune，上半年表現優於原先預期。最後，在 DAT，我們多年來首次看到貨運市場狀況正在改善。

On the back of this quarter's performance, we're raising our full-year DEPS guidance for the second time this year to a range of $22.15 to $22.30. That is up $0.30 at the midpoint and up $0.80 our original guide in January. In addition, we're raising our full-year revenue growth outlook, with total revenue now expected to be north of 8% and organic growth expected to be in the 6% range.

基於本季表現，我們今年第二次上調全年 DEPS 指引至 22.15 至 22.30 美元區間；以中位數計上調 0.30 美元，較我們 1 月的原始指引上調 0.80 美元。此外，我們也上調全年營收成長展望，總營收目前預期將超過 8%，有機成長預期約在 6% 左右。

Second, we continue to build momentum around AI and the automation of tasks and workflows within our end markets. During the quarter, our product release cadence accelerated, and I'd like to highlight both the breadth of what we're doing and the vertical market depth of what each of our businesses is delivering on the AI and agentic product front.

第二，我們持續在 AI 以及於終端市場中任務與工作流程自動化方面累積動能。本季我們的產品發布節奏加快，我想強調我們所做工作的廣度，以及各事業在 AI 與代理式（agentic）產品方面所交付的垂直市場深度。

Since we last spoke, Deltek released agented capabilities across both its GovCon and private sector solutions, including RFP development and month-end financial close. Vertafore released its AI velocity platform with six SKUs, which I'll get into more depth and to later. Strata released in beta is AI-enabled financial Decision Intelligence product. Aderant shipped its first generation of agents across collections, billing appeals, talent and time capture. Procare shipped its first of many agentic features Room Runner, which helps daycare operators optimize enrollment strategies.

自上次通話以來，Deltek 在其 GovCon 與私部門解決方案中推出代理式能力，包括 RFP 開發與月末財務結帳。Vertafore 發布其 AI velocity 平台，包含六個 SKU，我稍後會更深入說明。Strata 以 beta 形式發布其 AI 驅動的財務決策智慧（Decision Intelligence）產品。Aderant 在催收、帳單申訴、人才與工時擷取等領域推出第一代代理。Procare 推出眾多代理式功能中的第一個：Room Runner，協助托育機構營運者最佳化招生策略。

CentralReach deployed its next generation of AI solutions focused on clinical documentation quality and audit readiness. DAT continued the evolution of its ML-based freight match automation capabilities, including its first proactive AI recommendation engine, which services load opportunities to carriers who have not otherwise come across them. And SoftWriters, iTrade, Foundry, and ConstructConnect rounded out this quarter's AI and agentic product releases. So I know this was a long list, but that's precisely the point.

CentralReach 部署其下一代 AI 解決方案，聚焦於臨床文件品質與稽核就緒度。DAT 持續演進其以機器學習為基礎的貨運媒合自動化能力，包括其第一個主動式 AI 推薦引擎，向承運商推送其原本不會接觸到的載貨機會。SoftWriters、iTrade、Foundry 與 ConstructConnect 也完成了本季的 AI 與代理式產品發布。我知道這份清單很長，但這正是重點所在。

We continue to see very positive signals from our businesses and their ability to build and ship agentic features and products. Our right to win is grounded in the same advantages that have always made our vertical software businesses strong: deep domain expertise, embedded customer workflows, unique data assets, high trust customer relationships, and scale distribution.

我們持續從各事業及其打造並交付代理式功能與產品的能力中看到非常正面的訊號。我們的致勝優勢（right to win）根基於那些一向使我們垂直軟體事業強大的優勢：深厚的領域專業、嵌入式客戶工作流程、獨特的資料資產、高信任度的客戶關係，以及規模化的分銷能力。

Product velocity is also increasing within our businesses, each company develops AI capabilities close to their customers and their workflows while our AI accelerator team further increases velocity and scales reusable patterns across the portfolio. This model is working very well. Demand is strong, but these products are iterative and in many cases, they change how customers do their work, so the pace of adoption is still developing.

各事業的產品速度也在提升；每家公司都在貼近其客戶與工作流程的地方開發 AI 能力，同時我們的 AI 加速器團隊進一步提升速度，並在整個投資組合中擴大可重複使用的模式。這個模式運作得非常好。需求強勁，但這些產品是迭代式的，而且在許多情況下會改變客戶的工作方式，因此採用速度仍在發展中。

What is powerful, though, is that these products will improve with use. As adoption expands, the products get better, knowledge graphs compound and the value to customers increase. That creates a flywheel. More value drives more adoption, more adoption improves the product and the cycle accelerates. So while we're still early in the commercialization curve, we like the progress we're seeing.

不過，強大之處在於這些產品會隨著使用而改善。隨著採用擴大，產品會變得更好，知識圖譜會累積增長，對客戶的價值也會提升。這會形成飛輪效應。更多價值帶來更多採用，更多採用改善產品，循環加速。因此，雖然我們仍處於商業化曲線的早期階段，但我們喜歡目前看到的進展。

And third, we continue to execute with the same consistent disciplined capital allocation approach that has defined Roper for years. During the quarter, we repurchased another 3.6 million shares, and we expect to soon receive proceeds from Indicor's divestment of its instrumentation businesses to AMETEK.

第三，我們持續以多年來定義 Roper 的一致且嚴謹的資本配置方式執行。本季我們再回購 360 萬股，並預期不久將收到 Indicor 將其儀器事業出售給 AMETEK 的處分款項。

So let me turn the call over to Jason, so he can walk you through the details of both of these and our financials and balance sheet position. Jason?

接下來我把電話交給 Jason，請他帶各位了解這兩項事項的細節，以及我們的財務表現與資產負債表狀況。Jason？

Thanks, Neil. And good morning, everyone. I'll first take you through our second quarter financial performance, beginning on slide 6.

謝謝，Neil。各位早安。我先從第 6 張投影片開始，帶各位回顧我們第二季的財務表現。

At a high level, this was a solid second quarter. We finished above the high end of our DEPS guidance range, and organic growth was slightly above expectations. Revenue of $2.1 billion was up 9%. Organic growth of 5%, and acquisitions contributing 3 points. Organic recurring revenue across our Softwares segments grew 7% in the quarter, consistent with our Q1 performance. Further, we expect organic recurring revenue growth to inflect higher as we move into the second half.

從整體來看，這是一個穩健的第二季。我們的表現高於DEPS指引區間的上緣，而有機成長也略高於預期。營收為21億美元，年增9%。有機成長5%，併購貢獻3個百分點。本季我們軟體各事業群的有機經常性收入年增7%，與第一季表現一致。此外，隨著我們進入下半年，我們預期有機經常性收入成長將出現更高的上行拐點。

EBITDA of $815 million -- it was $815 million with EBITDA margin of 38.6%. Core EBITDA margin was down 70 basis points. The reduction was mainly driven by lower margins in our TEP segment, given similar dynamics to Q1, which are: we had higher input costs at Neptune and amidst more reoccurring consumables at NDI and Vertafore. As we move into the second half, we expect margin improvement at Neptune and easy comparisons in the segment.

EBITDA為8.15億美元——也就是8.15億美元，EBITDA利潤率為38.6%。核心EBITDA利潤率年減70個基點。下降主要由TEP事業群利潤率較低所致，原因與第一季類似：Neptune的投入成本較高，以及NDI與Vertafore的經常性耗材占比提高。隨著我們進入下半年，我們預期Neptune的利潤率將改善，且該事業群的比較基期也將更為有利。

Importantly, in the quarter, core margins across our Software segments were down a modest 10 basis points, which includes investment in our AI team. Diluted earnings per share of $5.38 was above our guided range of $5.25 to $5.30 and up 10% over prior year. This outperformance was driven by the combination of solid organic growth and additional buyback activity, which generated $0.03 of accretion in the quarter relative to our guidance.

重要的是，本季我們軟體各事業群的核心利潤率僅小幅下滑10個基點，其中包含對AI團隊的投資。稀釋後每股盈餘為5.38美元，高於我們指引區間5.25至5.30美元，且較去年同期成長10%。此一優於預期的表現，來自穩健的有機成長與額外的庫藏股回購活動的共同作用；相較於我們的指引，回購在本季帶來0.03美元的增益。

Free cash flow was $447 million, up 11% over prior year. On a trailing 12-month basis, free cash flow is now at $2.6 billion and has compounded at 18% over the past three years, or 15% adjusted for Section 174. So growth in free cash flow, coupled with share repurchase activity, resulted in free cash flow per share growth of 19% in the quarter.

自由現金流為4.47億美元，較去年同期成長11%。以過去12個月計，自由現金流目前為26億美元，過去三年複合成長率為18%，若就第174條（Section 174）調整後則為15%。因此，自由現金流的成長，加上股份回購活動，使得本季每股自由現金流成長19%。

Now if you turn with me to slide 7, I'll walk you through our financial position and capital deployment. We exited the second quarter at 3.4 times net debt to EBITDA, up from 3.1 times at the end of Q1, reflecting the capital deployed towards share repurchases in the quarter. We finished with $365 million of cash and $2.9 billion drawn on our $3.5 billion revolver.

現在請和我一起看第7頁投影片，我將帶各位了解我們的財務狀況與資本配置。我們在第二季末的淨負債／EBITDA為3.4倍，高於第一季末的3.1倍，反映本季用於股份回購的資本投入。期末現金為3.65億美元，並在35億美元的循環信貸額度中動用29億美元。

As Neil mentioned, during the quarter, we repurchased 3.6 million of shares for $1.2 billion, at an average price of approximately $341 per share. That brings our cumulative repurchases to 9 million shares, spending $3.2 billion since this program began. Said differently, we have repurchased over 8% of our shares in eight months, which gets us back to our 2013 share count position.

如Neil所提到，本季我們以12億美元回購360萬股，平均每股價格約341美元。自該計畫開始以來，我們累計回購900萬股，合計支出32億美元。換句話說，我們在八個月內回購了超過8%的股份，使我們的流通股數回到2013年的水準。

Also, Indicor announced an agreement to sell its instrumentation businesses to AMETEK, which is expected to close in the second half of this year. As a minority holder in Indicor, we expect gross proceeds of approximately $1.4 billion or an estimated $1.2 billion after tax, which further strengthens our balance sheet and deployment capacity.

此外，Indicor宣布已達成協議，將其儀器儀表業務出售給AMETEK，預計於今年下半年完成交割。作為Indicor的少數股權持有人，我們預期可取得約14億美元的總收益，或稅後估計約12億美元，這將進一步強化我們的資產負債表與資本部署能力。

Of note, the flow control businesses remain in the Indicor portfolio and will provide additional liquidity following a competitive sale process, which is not yet factored into our capacity framework. With good visibility into forward free cash flow and the expected Indicor instrumentation proceeds, we retain more than $5 billion of annualized capacity for capital deployment.

值得注意的是，流體控制業務仍保留在Indicor投資組合中，並將在競爭性出售流程後提供額外流動性；目前這部分尚未納入我們的能力框架。在對未來自由現金流具備良好能見度，以及預期可取得Indicor儀器儀表出售收益的情況下，我們仍保有每年超過50億美元（年化）的資本部署能力。

We expect M&A activity to break loose later this year and into 2027 and plan to prepare our balance sheet to take advantage of these opportunities. That said, we will remain patient and disciplined and deploy capital to the highest risk-adjusted return investments.

我們預期併購（M&A）活動將在今年稍晚至2027年期間明顯升溫，我們也計畫調整資產負債表，以把握這些機會。即便如此，我們仍將保持耐心與紀律，將資本投入於風險調整後報酬最高的投資。

With that, I'll turn it back over to Neil to discuss segment performance and our outlook. Neil?

接下來我把時間交還給Neil，請他說明各事業群表現與我們的展望。Neil？

Thanks, Jason. As we turn to page 9, let's review our Application Software segment. Revenue grew 8% in total, with organic revenue growth of 5%, slightly better than our expectations. EBITDA margins were 42.8% and core margins declined 20 basis points year-over-year. Organic recurring and reoccurring revenue, which represents about 85% of the segment, continued to grow in the mid-single digit plus range while nonrecurring revenue was down low single digits.

謝謝，Jason。接著看第9頁，我們來回顧應用軟體（Application Software）事業群。營收總計成長8%，其中有機營收成長5%，略優於我們的預期。EBITDA利潤率為42.8%，核心利潤率年減20個基點。有機經常性與經常性（reoccurring）收入約占該事業群的85%，持續以中個位數以上的幅度成長；非經常性收入則下滑低個位數。

Looking across the segment, a couple of themes spike out. First, our SaaS transitions continue to advance meaningfully. Several of our larger businesses made real progress on ground to cloud conversions and on bringing new cloud native products to market. Over time, this should continue to be a positive trend for this segment. And second, as discussed earlier, AI momentum continues to build across the portfolio.

從整個事業群來看，有幾個主題特別突出。第一，我們的SaaS轉型持續取得顯著進展。我們幾個較大型的業務在地端轉雲端（ground-to-cloud）轉換，以及將新的雲原生產品推向市場方面，都取得實質進展。長期而言，這應會持續成為該事業群的正向趨勢。第二，如先前所述，AI動能在整個投資組合中持續累積。

Turning to business highlights. Aderant was once again excellent in the quarter. The business continues to win in the market to drive accelerating adoption of its cloud offerings and build momentum with its genetic AI solutions. Deltek was solid in the quarter, driven by strength in its private sector solutions. GovCon was decent, and we saw some encouraging signs. That said, it is still too early to call it a trend.

接著談業務亮點。Aderant本季再次表現出色。該業務持續在市場上勝出，推動其雲端產品採用加速，並在其生成式AI解決方案上建立動能。Deltek本季表現穩健，主要受其私部門解決方案強勁帶動。GovCon表現尚可，我們也看到一些令人鼓舞的跡象。不過，現在仍言之過早，尚不足以稱之為趨勢。

Before upgrading our GovCon outlook, we'd like to see the sales pipeline convert consistently over the next few quarters. We do expect improvement over time, supported by the market tailwinds from the OBBB, but we're not ready to call that inflection just yet.

在上調我們對GovCon的展望之前，我們希望看到未來幾季銷售管線能夠持續、穩定地轉化為訂單。我們確實預期在OBBB帶來的市場順風支持下，隨時間推移將有所改善，但我們尚未準備好認定拐點已經到來。

Vertafore delivered another solid quarter with continued ARR growth. More importantly, faster product delivery is turning its AI strategy into tangible customer solutions. It's agentic AI strategy targets five areas of insurance distribution: digital servicing, smart submissions, accounting automation, producer workflows, and business intelligence. It starts with high-volume, labor-intensive work where customer value is immediate, that expands into more complex higher value workflows.

Vertafore本季再度交出穩健成績，ARR持續成長。更重要的是，更快的產品交付正把其AI策略轉化為具體的客戶解決方案。其代理式AI（agentic AI）策略鎖定保險分銷的五大領域：數位化服務、智慧投保申請（smart submissions）、會計自動化、業務員工作流程，以及商業智慧。該策略從高量、勞力密集的工作切入，客戶價值可立即體現，並進一步擴展到更複雜、價值更高的工作流程。

Velocity, Vertafore's purpose-built AI platform, enables rapid agent development directly within systems customers already use. Recently launched with six initial agent SKUs, it establishes the foundation for continued innovation. Vertafore's high right to win comes from applications that define and run their customers' core workloads. Thousands of customer environments are configured around specific data, products, rules, permissions, and processes. This allows agents to operate inside live environments with unique customer-by-customer context to perform real work accurately and securely.

Velocity是Vertafore專為此打造的AI平台，可直接在客戶既有使用的系統內快速開發代理（agent）。近期以六個初始代理SKU推出，奠定持續創新的基礎。Vertafore高度的致勝優勢（right to win）來自其應用程式能定義並運行客戶的核心工作負載。數千個客戶環境皆依特定資料、產品、規則、權限與流程進行配置。這使代理能在真實運行環境中，帶著每位客戶獨有的情境脈絡，在準確且安全的前提下完成實際工作。

Three products demonstrate the strategy's breadth and value. The reconciliation agent ingest carrier statements, matches transactions, and direct employees only to exceptions, reducing up to an hour of work to minutes. The submission processing agent converts unstructured e-mails and documents into underwriting ready data, reducing processing time from roughly an hour to about two minutes.

有三項產品展現了該策略的廣度與價值。對帳代理可匯入保險公司對帳單、比對交易，並僅將例外情況交由員工處理，將最多一小時的工作縮短至數分鐘。投保申請處理代理可將非結構化的電子郵件與文件轉換為核保可用資料，將處理時間從約一小時縮短至約兩分鐘。

The Porter Launcher agent converts static PDF applications into digital submission experiences, enabling MGAs to launch new programs dramatically faster and compete more effectively. These are not features that merely make work faster. They make meaningful portions of the workflow, improving speed and accuracy, increasing capacity, enabling growth without proportional headcount. That deepens Vertafore's customer value and expand this opportunity beyond software spend, into the much larger pool of labor sense.

Porter Launcher代理可將靜態PDF申請表轉換為數位化投保申請體驗，使MGAs能以大幅更快的速度推出新方案，並更有效地競爭。這些不只是讓工作更快的功能。它們能承擔工作流程中具實質份量的部分，提升速度與準確性、增加產能，並在不按比例增加人力的情況下支持成長。這將加深Vertafore對客戶的價值，並把機會從軟體支出擴展到更大的勞動力支出池。

Illumia delivered a strong second quarter with revenue ahead of expectations as it wins share and cross-sells into the combined CBORD and Transact customer base. Strata also performed well with its AI strategy accelerating faster than anticipated.

Illumia 第二季表現強勁，營收優於預期，因其持續提升市占，並在合併後的 CBORD 與 Transact 客戶基礎中推動交叉銷售。Strata 亦表現出色，其 AI 策略的加速進展快於原先預期。

Procare remains a work in progress. The team is advancing a broader product strategy, with RoomRunner released, strong early usage, and more innovation in the pipeline. At the same time, Joe and the team are shoring up the core product, go-to-market execution, and implementation capability.

Procare 仍在持續推進中。團隊正在推動更廣泛的產品策略，RoomRunner 已發布、早期使用情況強勁，且更多創新正在研發管線中。同時，Joe 與團隊也在強化核心產品、上市（go-to-market）執行，以及導入實施能力。

And finally, CentralReach continues to execute at a high level, delivering very strong growth and meaningful margin improvement. As a reminder, CentralReach turns organic next quarter, and after one year, the business is ahead of our deal model. As we turn to our outlook for application software, we expect organic growth for the second half of the year to be in the mid-single-digit plus range.

最後，CentralReach 持續以高水準執行，帶來非常強勁的成長與顯著的利潤率改善。提醒一下，CentralReach 將於下季轉為有機成長（organic），且在一年後，該業務表現已優於我們的交易模型（deal model）。展望應用軟體（application software），我們預期下半年有機成長將落在中個位數以上（mid-single-digit plus）的區間。

Please turn us to page 10. Total revenue in our network software segment grew 12%, and organic revenue grew 4% in the quarter. Organic recurring growth was strong, growing high singles, was offset by weaker reoccurring and non-recurring revenue at MHA and iPipeline.

請翻到第 10 頁。我們的網路軟體（network software）部門總營收成長 12%，本季有機營收成長 4%。有機經常性（recurring）成長強勁，達高個位數，但被 MHA 與 iPipeline 較弱的經常性與非經常性（non-recurring）營收所抵銷。

EBITDA margins were 50.9%, down 370 basis points year-over-year, while core margins improved slightly, up 30 basis points. The gap reflects two dynamics: the acquisition of Subsplash, a faster growth business with a lower but steadily improving margin profile, and our ongoing investment at DAT, particularly Convoy.

EBITDA 利潤率為 50.9%，年減 370 個基點，而核心利潤率小幅改善，上升 30 個基點。差距反映兩項因素：收購 Subsplash（成長更快、利潤率較低但持續改善的業務），以及我們在 DAT（尤其是 Convoy）持續投入。

Turning to the business level highlights. DAT grew nicely in the quarter and continues to execute at high fidelity. We're also seeing encouraging growth in our digital freight marketplace solutions, with increasing adoption and engagement.

接著看各業務層級重點。DAT 本季成長良好，並持續以高一致性執行。我們也看到數位貨運市集解決方案出現令人鼓舞的成長，採用率與參與度持續提升。

Broader market indicators are improving as well. Carrier additions have increased steadily throughout the year, spot pricing is strengthening relative to contract rates, and carrier rejection rates are rising. We still need to see load volumes improve before becoming fully bullish on the recovery. The 2026 benefit is tracking modestly ahead of our prior expectations, and the indicators remain encouraging as we look towards 2027.

更廣泛的市場指標也在改善。承運商新增數量全年穩步增加，即期（spot）價格相對於合約運價走強，且承運商拒載率正在上升。在我們對復甦轉為完全樂觀之前，仍需看到貨載量（load volumes）改善。2026 年的效益追蹤結果略優於我們先前預期，展望 2027 年，相關指標仍令人鼓舞。

ConstructConnect had another solid quarter with continued growth of its AI-based takeoff solution. The business is expanding AI deeper into preconstruction workflow, helping customers reduce manual effort, move faster through project documents, and improve decision speed.

ConstructConnect 再度交出穩健的一季，其 AI 驅動的算量（takeoff）解決方案持續成長。該業務正將 AI 更深入導入施工前（preconstruction）工作流程，協助客戶降低人工投入、更快處理專案文件，並提升決策速度。

Foundry continued to deliver year-over-year ARR growth with meaningful AI innovation across SmartRoto and Griptape. SmartRato automates a historically manual part of the visual effects workflow, while Griptape extends Foundry's position in AI orchestration across production and post-production environments.

Foundry 持續實現年對年 ARR 成長，並在 SmartRoto 與 Griptape 上推出具意義的 AI 創新。SmartRato 可自動化視覺特效工作流程中一項歷來高度仰賴人工的環節，而 Griptape 則強化 Foundry 在製作與後期製作環境中 AI 編排（orchestration）的領先地位。

Subsplash delivered another strong quarter with double-digit software growth, even faster payments growth and meaningful margin expansion. The business also continued to extend AI across its church engagement and giving platform, taking its AI assistant live to all customers.

Subsplash 再度交出強勁的一季，軟體營收維持雙位數成長，支付（payments）成長更快，且利潤率顯著擴張。該業務也持續將 AI 延伸至其教會互動與奉獻平台，並已將其 AI 助理正式上線提供給所有客戶。

Finally, I joined our SoftWriters team, which provides the core operating system for long-term care pharmacies for a site visit at L.A. Script Pharmacy. I spent time understanding LI Script's strategic and operational priorities and where SoftWriters can help accelerate the business. LI Script serves more than 32,000 beds across 150 facilities, so small workflow improvements create significant value.

最後，我與我們的 SoftWriters 團隊（為長期照護藥局提供核心作業系統）一同前往 L.A. Script Pharmacy 進行實地參訪。我花時間了解 LI Script 的策略與營運優先事項，以及 SoftWriters 可如何協助加速其業務。LI Script 服務超過 150 家機構、合計逾 32,000 張床位，因此即使是小幅的工作流程改善也能創造顯著價值。

I sat alongside pharmacists processing prescription orders with and without SoftWriters' AI-powered order entry and review solution. Without automation, a new order can require up to 90 seconds of manual entry and review. Our solution reads the prescription, populates information in the existing workflow, and lets the pharmacist focus on validation rather than transcription. Processing time falls to about 18 seconds per order, an 80% reduction.

我坐在藥師旁，觀察在有與沒有 SoftWriters 的 AI 驅動處方訂單輸入與審核解決方案下，處理處方訂單的差異。在沒有自動化的情況下，一筆新訂單最多可能需要 90 秒的人工輸入與審核。我們的解決方案可讀取處方、在既有工作流程中自動帶入資訊，讓藥師能專注於驗證而非抄錄。每筆訂單處理時間降至約 18 秒，縮短 80%。

The benefit extends beyond lower labor costs. It gives pharmacists and technicians more capacity for quality assurance, clinical review, answering nurses' questions, supporting customers, and absorbing volume without proportional headcount. That means better patient care, better customer service, and a more scalable path to growth.

效益不僅止於降低人力成本。它讓藥師與技術員有更多產能投入品質保證、臨床審查、回覆護理人員問題、支援客戶，並在不按比例增加人頭的情況下吸收更多量能。這意味著更好的病患照護、更好的客戶服務，以及更具可擴展性的成長路徑。

Equally important, the automation is delivered in context at the precise point of work with extremely low latency. SoftWriters derive proprietary data, domain-specific models, dense workflow integration, customer trust, and distribution were evident. It showed how our vertical market moats enable differentiated AI that automates meaningful customer work. Order entry is just the beginning of SoftWriters' workflow automation opportunity.

同樣重要的是，自動化是在工作發生的精準節點、以情境化方式交付，且延遲極低。SoftWriters 的專有資料、領域專屬模型、深度工作流程整合、客戶信任與通路優勢都十分明顯。這展示了我們在垂直市場的護城河如何支撐差異化 AI，進而自動化客戶的關鍵工作。訂單輸入只是 SoftWriters 工作流程自動化機會的起點。

Finally, the monetization model is also attractive and in addition to the existing software spend. SoftWriters prices the solution transactionally, directly aligning revenue with orders automated and labor savings delivered. Customers achieve a clear return while SoftWriters participates in the value created with revenue scaling as adoption and transaction volumes grow. Thank you to the LI Script team for hosting me and for the partnership.

最後，其變現模式也具吸引力，且是在既有軟體支出之外的增量。SoftWriters 以交易（transactional）方式定價，讓營收與自動化的訂單量及所帶來的人力節省直接對齊。客戶可獲得明確的投資報酬，而 SoftWriters 也能分享所創造的價值，並隨採用率與交易量成長而擴大營收規模。感謝 LI Script 團隊接待我並促成合作夥伴關係。

Now let me turn to our outlook for network software. We expect organic growth for the second half of the year to increase versus the first half and be in the mid-single-digit plus range.

接著我來談談我們對網路軟體的展望。我們預期下半年有機成長將較上半年加速，並落在中個位數以上（mid-single-digit plus）的區間。

Now please turn to page 11, and let's review our TEP segment. Total revenue and organic revenues grew 7%. Results were better than anticipated at Neptune, NDI, and Verathon. EBITDA margins were 34.5%, down 220 basis points year-over-year, reflecting the same dynamics discussed last quarter. Input cost pressure at Neptune, principally bronze ingot inflation and a mix shift at both NDI and Verathon towards faster-growing consumables, which carry lower gross margins, though notably more durable and reoccurring revenue profiles.

現在請翻到第 11 頁，我們來檢視 TEP 部門。總營收與有機營收成長 7%。Neptune、NDI 與 Verathon 的表現優於預期。EBITDA 利潤率為 34.5%，年減 220 個基點，反映與上季討論相同的因素。Neptune 面臨投入成本壓力，主要來自青銅錠（bronze ingot）通膨，以及 NDI 與 Verathon 的產品組合轉向成長更快的耗材（consumables）；耗材毛利率較低，但其營收型態更具耐久性且更偏經常性（recurring）。

Starting with NDI. The business had another very strong quarter, driven by sustained demand for its electromagnetic tracking solutions across cardiac, neurological, and orthopedic applications. EP remains a compelling multiyear growth opportunity with procedure volumes increasing, leading OEMs launching new tracking-enabled catheter platforms, and NDI uniquely positioned at the sensor layer.

先從 NDI 開始。該業務再度交出非常強勁的一季，動能來自其電磁追蹤解決方案在心臟、神經與骨科應用上的持續需求。EP 仍是具吸引力的多年期成長機會：手術量增加、主要 OEM 推出新的支援追蹤功能的導管平台，且 NDI 在感測器層（sensor layer）具備獨特定位。

Neptune had an encouraging quarter. Revenue declined modestly year-over-year but came in ahead of our expectations as strong mechanical meter volumes and continued growth at UnionLink helped offset the anticipated decline in market unit volumes. UnionLink also continues to strengthen Neptune's position across the broader meter-to-cash workflow through its cloud-based utility billing and customer communication capabilities. Finally, the metering business remains on track to return to growth in the second half.

Neptune 本季表現令人鼓舞。營收年對年小幅下滑，但優於我們預期，因強勁的機械式水表出貨量與 UnionLink 的持續成長，抵銷了原先預期的市場單位量下滑。UnionLink 也透過其雲端公用事業帳單與客戶溝通能力，持續強化 Neptune 在更廣泛的抄表到收款（meter-to-cash）工作流程中的地位。最後，計量（metering）業務仍按計畫在下半年恢復成長。

Verathon delivered solid growth driven by strong demand for BFlex and GlideScope single-use offerings. We're also encouraged by the launch of the CFlex single-use cystoscope and BFlex specimen collection system, which further expands Verathon's single-use portfolio.

Verathon 交出穩健成長，主要由 BFlex 與 GlideScope 一次性（single-use）產品的強勁需求所帶動。我們也對 CFlex 一次性膀胱鏡（cystoscope）與 BFlex 標本採集系統的上市感到鼓舞，這將進一步擴大 Verathon 的一次性產品組合。

Finally, I want to recognize the teams at CIVCO, FMI, and IPA for their excellent work in the second quarter. Each team executed well and contributed to the segment's performance. Turning to our TEP outlook. We expect organic growth for the second half of the year to be in the high singles range and be a bit stronger in Q3.

最後，我要表揚 CIVCO、FMI 與 IPA 團隊在第二季的優異表現。各團隊執行到位，並對部門績效做出貢獻。接著談我們對 TEP 的展望。我們預期下半年有機成長將落在高個位數區間，且第三季（Q3）會略為更強。

With that, please turn us to page 13. On this page, let's walk through our increased full year revenue and DEPS guidance as well as our Q3 outlook. For the full year, we're raising adjusted DEPS guidance to a range of $22.15 to $22.30. That represents a $0.30 increase at the midpoint from our prior guide and an $0.80 increase from our original guide.

那麼，請各位翻到第 13 頁。在這一頁，我們將說明我們上調的全年營收與 DEPS 指引，以及我們對第三季的展望。就全年而言，我們把調整後 DEPS 指引上調至 22.15 美元至 22.30 美元的區間。這代表相較於先前指引的中點上調 0.30 美元，較我們最初指引上調 0.80 美元。

We're also increasing our full year total revenue growth guidance to north of 8% with organic revenue growth expected to be in the 6% area. Please note, this guidance outlook excludes any proceeds from Indicor's divestiture of its instrumentation businesses. For the back half of the year, we expect a tax rate in the 21% area.

我們也將全年總營收成長指引上調至 8% 以上，其中有機營收成長預期約在 6% 左右。請注意，此指引展望不包含 Indicor 出售其儀器業務所帶來的任何處分收益。就下半年而言，我們預期稅率約在 21% 左右。

To reiterate a few key assumptions from our segment commentary, there are some encouraging signs of stabilization at Deltek, but not enough yet to change our outlook. DAT's freight market is improving, and we now expect a modest benefit in '26 with the setup becoming increasingly encouraging for '27. And Neptune's first half performance was stronger than we modeled, contributing to our first half beat. For Q3, we're establishing adjusted DEPS guidance of $5.75 to $5.80.

重申我們在各事業部評論中的幾項關鍵假設：Deltek 出現一些令人鼓舞的穩定跡象，但目前仍不足以改變我們的展望。DAT 的貨運市場正在改善，我們現在預期在 26 年會有小幅助益，且 27 年的態勢愈來愈令人鼓舞。Neptune 上半年表現強於我們的模型預估，帶動我們上半年優於預期。就第三季而言，我們建立調整後 DEPS 指引為 5.75 美元至 5.80 美元。

Now please turn with us to page 14, and we'll open it up for your questions. We'll conclude with the same 3 points with which we started. First, we delivered solid second quarter results and are raising our outlook for the year. Retention remains strong. And based on our year-to-date performance, we're raising full year DEPS guidance for the second time this year. We're also raising our full year revenue growth outlook, both for total and organic revenue.

現在請各位翻到第 14 頁，我們將開放提問。我們將以與開場相同的 3 點作結。第一，我們交出穩健的第二季成果，並上調全年展望。留存率仍然強勁。而基於我們年初至今的表現，我們今年第二次上調全年 DEPS 指引。我們也上調全年營收成長展望，包含總營收與有機營收。

Second, AI momentum continues to accelerate across the enterprise. Our product release cadence accelerated meaningfully in the quarter, and our businesses are shifting agentic and AI-enabled capabilities into high-value vertical workflows. Though it's still early, and we will continue learning through the back half of the year, the signals are very positive. We have the vertical market knowledge, deeply embedded workflows, unique data advantages, high-trust customer relationships, and scale distribution to win in the AI era.

第二，AI 動能持續在整個企業加速。本季我們的產品發佈節奏明顯加快，各事業也正把代理式（agentic）與 AI 賦能能力導入高價值的垂直工作流程。雖然仍在早期，我們也會在下半年持續學習，但各項訊號非常正面。我們具備垂直市場知識、深度嵌入的工作流程、獨特的資料優勢、高信任度的客戶關係，以及規模化的分銷能力，足以在 AI 時代勝出。

Finally, our capital allocation framework remains unchanged. We'll deploy capital wherever we see the greatest opportunity for durable long-term cash flow per share compounding, whether through acquisitions or opportunistic share repurchases. Over the last eight months, that discipline led us to repurchase approximately $3.2 billion of equity, roughly 9 million shares or a bit more than 8% of shares outstanding, returning the share count to 2013 levels, and we're very pleased by that outcome.

最後，我們的資本配置框架維持不變。我們會把資本投入到我們認為最能帶來每股長期、可持續現金流複利成長的機會上，無論是透過併購或是機會性庫藏股回購。過去八個月，這項紀律促使我們回購約 32 億美元的股權，約 900 萬股、約略超過流通在外股數的 8%，使股數回到 2013 年水準；我們對這個結果非常滿意。

Looking ahead, we believe M&A is likely to offer the more attractive long-term compounding opportunity provided assets meet our strategic and risk-adjusted financial criteria. We expect the M&A market to improve over the coming quarters and are cautiously optimistic about deploying $5 billion or more towards acquisitions over the next 12 to 18 months.

展望未來，我們認為只要標的符合我們的策略與風險調整後的財務標準，併購（M&A）很可能提供更具吸引力的長期複利成長機會。我們預期未來幾季 M&A 市場將改善，並對於在未來 12 到 18 個月內投入 50 億美元或以上於併購持審慎樂觀態度。

Given this expected improvement and with leverage currently around 3.4 times, near-term capital deployment will favor deleveraging over opportunistic buybacks. We expect to delever quickly through strong second half cash flow and the Indicor proceeds rebuilding capacity to pursue high-quality acquisition candidates.

鑑於這項預期改善，且目前槓桿約為 3.4 倍，短期資本部署將更偏向去槓桿，而非機會性回購。我們預期透過下半年強勁的現金流，以及 Indicor 處分所得來快速去槓桿，並重建追求高品質併購標的的能力。

So in closing, the ingredients for accelerating cash flow per share compounding, our most important financial metric, continue to strengthen. Our portfolio is the strongest it has ever been. Organizational velocity is increasing. AI is expanding our addressable markets, and we expect our product innovation to translate into higher growth over time. Our capital deployment capacity and flexibility remain significant differentiators and most importantly, our discipline is unchanged.

因此在結語中，用以加速每股現金流複利成長（我們最重要的財務指標）的各項要素持續強化。我們的投資組合是史上最強。組織運轉速度正在提升。AI 正在擴大我們的可服務市場，我們預期產品創新將隨時間推動更高成長。我們的資本部署能力與彈性仍是重要差異化優勢；更重要的是，我們的紀律不變。

So with that, we'll open it up to your questions.

那麼，我們現在開放各位提問。

(Operator Instructions) Deane Dray, RBC.

(接線員指示) Deane Dray，RBC。

Congrats on all the AI, SKU launches.

恭喜你們在 AI 與 SKU 推出方面的成果。

Thank you, Deane. Before I answer your question, I just want to jump in and just say a couple of things, if that's okay, to you personally. Congrats on just a great career and also congrats on stepping away from the day to day.

謝謝你，Deane。在回答你的問題之前，如果可以的話，我想先插一句，對你個人說幾件事。恭喜你擁有非常出色的職涯，也恭喜你從日常工作中退下來。

I've come to know you -- we've come to know you for the better part of the last 15 years, lots of conversations, several road shows, and I think the thing that I take away the most from that in respect the most is the way you've done your job. Always fair, objective, principles, including the moments when we didn't see things the same way. So it's pretty amazing. So congrats on that.

在過去大約 15 年的大部分時間裡，我們認識你——我們一直與你互動，進行了很多對話、跑了好幾次路演；而我從中最深的體會、也最敬重的一點，是你做事的方式。你一向公允、客觀、講求原則，即使在我們看法不一致的時候也是如此。這真的很了不起。所以恭喜你。

I also appreciate you introducing me to a few leaders who become friends that I can call friends now. Great -- congrats on a great career and especially congrats on retirement. And I hope you get plenty of time with that family, especially those grandkids.

我也很感謝你介紹我認識幾位領導者，他們後來成了朋友——我現在可以稱他們為朋友。總之——恭喜你有一段精彩的職涯，尤其恭喜你退休。也希望你能有很多時間陪伴家人，特別是那些孫子孫女。

Thank you so much for all those kind words. It's really been my privilege to follow Roper over the years. And yes going back to the hallowed days of Brian Jellison. So I really appreciate all the help Neil, you and the team have given me and all the insight. I've got a ton of respect for you all, and I wish you continued success.

非常感謝你這些溫暖的話。多年來能追蹤 Roper 的發展，真的是我的榮幸。是的，還可以追溯到 Brian Jellison 的光輝年代。所以我非常感謝 Neil、你以及團隊給我的所有協助與洞見。我非常敬重你們，也祝福你們持續成功。

Thank you so much. All right, let's get to your questions. So congrats.

非常感謝。好，那我們來談你的問題。再次恭喜。

I don't know how I pivot to buybacks, but I'm going to. Just the idea here, this is now the second quarter of some sizable buybacks, and I'd really be interested in hearing how you look at buybacks on a capital allocation versus a CRI-accretive deal equivalent what that would need to look like. And I think the key here is on a risk-adjusted basis. But if you could just share that insight, that would be great.

我不知道要怎麼把話題轉到回購，但我就這麼做了。這已經是連續第二季進行相當規模的回購，我很想聽聽你們如何在資本配置上看待回購，相較於一筆在 CRI（資本報酬率）上具增益的交易，兩者需要呈現什麼樣的條件才會等價。我想關鍵在於風險調整後的基礎。如果你能分享一些看法就太好了。

That's right. Happy to do it. So we always, first of all, look at what we believe will be the best way to compound our cash flow per share over a longer arc, like a five- to seven-year period of time. We've always looked at it this way.

沒錯。很樂意。首先，我們一向會看我們認為在更長的時間弧線上——例如 5 到 7 年——最能提升每股現金流複利成長的方式。我們一直都是用這種方式來看。

Until the last eight months, the capital allocation mathematically always favored M&A. When you look at the returns, both a time zero return, a time five return, whether it's a year five EBITDA, or year five ROIC, to M&A. Obviously, with the last 12 months or so, 12 to 18 months with the current share price level of Roper Technologies in the broader software sector, all of a sudden, buybacks became significantly more attractive.

在過去八個月之前，從數學上看，資本配置幾乎總是偏向併購（M&A）。當你看報酬，不論是零時點報酬、五年期報酬，無論是第 5 年 EBITDA，或第 5 年 ROIC，併購都更占優勢。顯然，在過去大約 12 個月、或 12 到 18 個月，隨著 Roper Technologies 的股價水準以及更廣泛軟體產業的估值變化，回購突然變得顯著更具吸引力。

Now looking forward, we expect the private values to mirror those or come down to the public values. We're starting to see some very early signs of that. As we all know, public values are ultimately the gravitational force for all private companies. And when that happens, then the math turns quite interesting, quite accretive, quite more attractive towards M&A versus buyback. And so we'll always sort of judge those two between what's best for the long-term cash flow for share compounding.

而往前看，我們預期私募市場的估值會向公開市場估值靠攏，或下修至公開市場估值水準。我們已開始看到一些非常早期的跡象。眾所周知，公開市場估值最終是所有私營公司的「重力」。當那種情況發生時，數學就會變得相當有意思——相當具增益、也更偏向併購而非回購。因此，我們會在這兩者之間持續衡量，判斷哪一個最有利於長期的每股現金流複利成長。

That's really helpful. And could you just expand on the comment about the pipeline? I think the term was breaking loose, just the expectation on deal flow over the near term.

這真的很有幫助。另外你能否再延伸說明一下你提到的關於管線（pipeline）的評論？我想你用的詞是「開始鬆動」，也就是對近期交易流量（deal flow）的預期。

Yeah. I should also say just on -- there is -- on the last question, we always look at the buyback of our own stock versus an acquisition with a risk premium. We always obviously know ourselves better than anything we could know externally. So we have to sort of be compensated their returns on that. And so we obviously look at that. So we round out that answer.

是的。我也想補充一下——關於上一個問題，我們一直會拿回購自家股票與帶有風險溢價的併購做比較。我們當然總是比任何外部標的更了解自己。所以在那樣的比較下，我們必須在報酬上獲得相應補償。因此我們顯然會看這一點。所以把那個回答補完整。

On the -- and now I forgot your second question already.

至於——然後我現在已經忘了你第二個問題了。

Pipeline.

管線。

The pipeline. Sorry. Yeah. What I would say there is Janet and her team, they always have a lot of dialogue. The dialogue's turned decidedly more constructive over the last couple of months-ish in terms of the sponsors sort of understanding the valuation landscape. They've had three or four years of DPI pressure.

管線。抱歉。是的。我會說的是，Janet 和她的團隊一直都有很多對話。過去大概幾個月左右，這些對話在贊助方（sponsors）逐步理解估值環境方面，明顯變得更具建設性。他們承受了三到四年的 DPI 壓力。

Private credit sort of pressure doesn't help their situation with the public values have been here for, like I said, the better part of nine to 12 months, sort of a bounce back. Sort of their hope of a bounce back, this near-term bounce back is likely not going to happen. Sponsors just coming to grips, the new 15 is yesterday's 25 or whatever the quote is from a couple of sponsors that we've talked to privately.

私募信貸（private credit）的壓力也無助於他們的處境；而公共市場估值如我所說，在過去大約 9 到 12 個月的大部分時間裡，呈現某種反彈——他們所期待的這種短期反彈，很可能不會發生。贊助方正在逐步接受現實：新的 15 就是昨天的 25——或是我們私下與幾位贊助方聊到的那句話。

A lot of proprietary opportunities right now. Sponsors saying, hey, if we can get a deal done, we'll just do something with you on a one-off basis. I think that's a sign of where we are, both as a preferred buyer, but also from a macro perspective. That sort of informs our view of the pipeline. We'll always be disciplined, we'll always be patient. I wouldn't anticipate there's like a massive breakthrough of the pipeline in the next small number of months, but we want to prepare the balance sheet to be completely ready when that time comes.

目前有很多專有（proprietary）的機會。贊助方會說，嘿，如果我們能把交易做成，我們就跟你用一次性（one-off）的方式合作。我認為這反映了我們所處的位置：一方面我們是優先買家（preferred buyer），另一方面也從總體環境角度反映現況。這些都形塑了我們對管線的看法。我們永遠會保持紀律、永遠會有耐心。我不預期接下來幾個月管線會出現什麼巨大的突破，但我們希望在那個時點到來時，資產負債表能完全準備就緒。

I'd also just say in the channel, like, from the investment banking community, we're getting a lot of signals there. And then also those that do market diligence and consultants, their books are filled up right now quite a bit, so it's another signal.

我也想說，就通路而言——從投資銀行圈子那邊，我們收到很多訊號。另外，那些做市場盡職調查的人以及顧問，他們目前的案子排得很滿，所以這也是另一個訊號。

Terry Tillman, Truist.

Terry Tillman，Truist。

Neil, Jason, and Zack, can you all hear me okay? I've had a bad connection. Can you hear me?

Neil、Jason 和 Zack，你們都聽得到我嗎？我這邊連線不太好。你們聽得到嗎？

We hear you great.

我們聽得很清楚。

Wonderful. On Vertafore, seems like it is very high value. We do a lot of work in insurtech, and I'd love to kind of double-click on it. I think you all said continued ARR growth. I mean how is the growth compared to the overall software organic, which I think was 5% in the quarter. And then on top of that, I mean, with these six agents, have you started to size the potential uplift you could see? And then I had a follow-up.

太好了。關於 Vertafore，看起來它的價值非常高。我們在 insurtech 做了很多研究，我很想更深入了解。我想你們提到 ARR 持續成長。我的意思是，它的成長相較於整體軟體的有機成長如何？我記得本季大概是 5%。另外，針對這六個代理（agents），你們是否已開始估算可能帶來的提升幅度？然後我還有一個追問。

Yeah. So the ARR growth is a little bit higher than the segment. So trending well there.

是的。ARR 的成長率比該事業群（segment）略高一些。所以趨勢表現不錯。

Yeah, I can take the agentic opportunity. So we've worked with our partners at Bain. Amy and her team have done their own internal work about what the TAM expansion potential is at Vertafore for the agentic work in the five areas that we talked about in the prepared remarks, it's a doubling-ish of the market size.

我可以回答代理式（agentic）的機會。我們與 Bain 的合作夥伴一起做了研究。Amy 和她的團隊也針對 Vertafore 在我們於事先準備講稿中提到的五個領域、就代理式工作所帶來的 TAM 擴張潛力做了內部分析；大致上是市場規模接近倍增。

Now specific to these six agents, it's a distinct minority or a smaller portion of that doubling, but this is just the beginning. I mean this is -- think of the scaffolding and the architecture of the agentic platform is probably the biggest part of the release here and then the six SKUs on top, which means the acceleration of SKUs coming from this point forward should accelerate and we'll get -- we will sort of develop product and launch product to sort of fully cover that 2x market opportunity.

至於這六個代理本身，它只占那個倍增中的一個明顯少數、或較小的一部分，但這只是開始。我的意思是——把這次發布視為代理式平台的鷹架與架構，可能才是最重要的部分；其上再加上六個 SKU。這表示從現在起，SKU 推出的速度應該會加快，而我們也會——我們會持續開發並推出產品，以更完整覆蓋那個 2 倍的市場機會。

And by the way, these are informed guesses. I don't think anybody really knows the complete TAM expansion, but going in and sort of automating some of these tasks and giving our customers that efficiency certainly lets us sort of monetize some of the labor spend.

順帶一提，這些都是基於資訊的推估。我不認為有人真的知道 TAM 擴張的完整幅度，但透過自動化其中一些任務、並為客戶帶來效率提升，確實讓我們能在某種程度上把部分人力支出變現。

Yes. And then the follow-up is kind of a hard question, unfortunately, but it is kind of -- you said guesses because I agree it is all guesses for all of us. But this kind of clearing event and the timing of it really to take hold, like I know we talked last quarter, Neil, and I'd even mentioned maybe is there some sort of like FTE layer that you have to add? I mean you said scaffolding then the agents. When do you see like the light bulb, the switch flipping where, hey, we're going to really run big, large production workloads. Do you think it's the second half of this year? Or just anything on that?

是。那我的追問有點難問、也有點不太好意思，但大概是——你說是推估，我同意，對我們所有人來說都是推估。但這種「釐清事件」（clearing event）以及真正開始發酵的時間點——我知道我們上季談過，Neil，我甚至提到過是否需要增加某種像是 FTE 的層級。你提到鷹架、然後是代理。你什麼時候會看到那種「燈泡亮起、開關切換」的時刻：也就是，嘿，我們要真的開始跑很大、很大型的生產工作負載（production workloads）了。你覺得會在今年下半年嗎？或是對此有任何看法？

Yeah. I said -- I think we said last quarter and we'll repeat this quarter. We -- as an organization, we just had an amazing set of learnings on how to build these commercial-grade agentic products over the last 12 to 18 months, and that's just compounding. We had just an amazing CTO, CPO event this quarter. It was the largest in-person Roper event we've ever had, a couple of hundred people.

是的。我說——我想我們上季說過，這季也再重申一次。我們——作為一個組織，在過去 12 到 18 個月裡，對於如何打造這些商用等級（commercial-grade）的代理式產品，獲得了一系列非常驚人的學習，而這些學習正在持續累積、複利化。我們本季剛辦了一場非常精彩的 CTO、CPO 活動。那是我們 Roper 史上最大規模的線下活動，有兩三百人。

And our AI center team, Shane and his team did just an expert job of having a dialogue in this course and training session on how to build like production-grade agentic capabilities, not agent toys, but like truly production grade, and that was incredible. So we've come that learning curve, and we'll continue sort of springboarding sort of that knowledge compounding.

而我們的 AI 中心團隊，Shane 和他的團隊在這次課程與訓練中，針對如何打造生產等級（production-grade）的代理式能力——不是代理玩具（agent toys），而是真正的生產等級——進行對話與教學，做得非常專業，令人印象深刻。所以我們已經走過那段學習曲線，並且會持續以這些知識的複利效應作為跳板往前推進。

Now we're also on a commercialization learning curve. And so how do you -- how do we price, how do we deploy, how do we drive utilization. And we're getting -- we'll certainly get our teams together across Roper to sort of share those learnings.

同時，我們也正處於商業化的學習曲線上。也就是我們要如何定價、如何部署、如何推動使用率。而我們正在累積——我們也一定會把 Roper 各團隊聚在一起，去分享這些學習。

Ultimately, whether or not we call them forward deployed engineers or not, I don't know. But definitely, the customer service implementation resources will spend more time with our customers driving utilization, teaching them, going through change management, their workflows. And it's just the rate of adoption across each of our businesses is the large unknown. So will it be material revenue in the second half? It will not be for us this year. But there's momentum building unmistakably inside the organization, and we're excited about that.

最終，我們是否稱之為前線部署工程師（forward deployed engineers），我不確定。但可以肯定的是，客戶服務與導入（implementation）資源會花更多時間與客戶一起推動使用、教學、進行變革管理與工作流程調整。而各個事業的採用速度是最大的未知數。所以今年下半年會是重大的營收嗎？對我們而言，今年不會。但組織內部的動能正在明顯累積，我們對此感到興奮。

Brent Thill, Jefferies.

Brent Thill，Jefferies。

Just on M&A, I think you had mentioned you're hoping to see the thaw in the cold. But I guess what is -- what's underpinning the confidence in that? Is that just, hey, it hasn't thought forever? Or you're now seeing signs underneath that we can't see that you're just giving a signal that things are in the mix? Like what's different than, hey, we just -- we hope it thaws.

就併購（M&A）而言，我記得你提到你希望看到寒冬回暖、開始解凍。但我想問的是——支撐你這個信心的基礎是什麼？是因為覺得，嘿，不可能永遠都這麼冷嗎？還是你們現在看到一些我們看不到的底層跡象，所以你才在釋放訊號，表示事情正在醞釀中？也就是說，跟單純說「我們希望會解凍」相比，現在有什麼不一樣？

Yeah. I think it's similar to what we attempted to say a little bit ago, which is the conversations we're having with sponsors are now real conversations. They understand the valuation world that we live in. They've come to grips with that reality. I would say the sponsors that are multi-industry sponsors come to grips with that a little bit quicker than ones that are specialized in sectors and software sector.

是的。我想這跟我們剛才稍早試著表達的類似：我們現在跟贊助方（sponsors）進行的對話，已經是「真正的」對話了。他們理解我們所處的估值環境。他們已經接受並正視這個現實。我會說，多產業的贊助方比那些專注於特定產業、尤其是軟體領域的贊助方，更快接受這件事。

As Jason said, the consultants who do a lot of the diligence work are quite busy. The conversation we have with the intermediaries, the bankers, their pipelines are quite busy and quite full. And it's that that informs us. The signal is there. It's positive, but it's distant, but it takes more than a couple of minutes to prepare the balance sheet to play offense. And so that's what we're doing at the moment.

如 Jason 所說，做大量盡職調查工作的顧問目前非常忙。我們跟中介機構、投行的對話也顯示，他們的案源管線相當忙碌、也相當充實。這些情況就是我們判斷的依據。訊號已經出現。它是正面的，但還有一段距離；而且要把資產負債表準備好、轉為進攻姿態，不是幾分鐘就能完成的。所以我們現在正在做的就是這件事。

Okay. And sorry, for the first half of the year, what has been the capital you've deployed --

好的。另外不好意思，想問一下上半年你們部署的資本是多少——

Well, outside the buyback, there are two small bolt-ons. So in aggregate, about $50 million.

嗯，除了回購之外，還有兩筆小型的補強式併購（bolt-on）。合計大約 5,000 萬美元。

$50 million. Okay. Great. And then just real quick, Jason, for Deltek, can you just give us a sense of how things are going, what you're seeing into the second half of the year?

5,000萬美元。好的。很棒。然後再很快問一下，Jason，關於 Deltek，你能否談談目前進展如何、以及你對下半年的觀察？

Yes. So actually had a good second quarter. We had mentioned a large license deal that we had not put in our guidance that they actually executed on, so it drove a little bit of outperformance in the segment.

可以。我們第二季其實表現不錯。我們之前提到過一筆大型授權交易，當時沒有納入我們的指引，但對方實際上已經完成簽約執行，因此帶動該事業部門有一些超預期表現。

Pipelines look really strong. I think what Neil mentioned in his prepared remarks is that we're really waiting for some of those conversions to be more of a trend before we call it. Certainly, the contractors, the appropriations are starting to move into the contractor world. I think you'll see some of the public company earnings releases have been good. So we have really good signal that things are going to move in our direction, but we haven't called it yet.

案源管線看起來非常強勁。我想 Neil 在事先準備的發言裡提到的是：我們正在等待其中一些轉換（conversions）在趨勢上更明確之後，才會下定論。確實，承包商端的情況是，政府撥款（appropriations）開始逐步流向承包商體系。我想你也會看到一些上市公司的財報發布表現不錯。所以我們收到的訊號非常好，顯示事情會朝對我們有利的方向發展，但我們還沒有正式定調。

Joe Giordano, Cowen.

Joe Giordano，Cowen。

Can you like -- I know it probably varies a lot business to business. But when we talk about these AI solutions that you're deploying, can you like help us think about what the hit rate internally is? Like when you're drawing these things up on the whiteboard, like how many ideas is it taking to generate something that's actually getting into production versus like how these businesses operated in like a pre-AI world when you're thinking about improvements?

你能不能——我知道這可能因業務而差異很大。但當我們談到你們正在部署的這些 AI 解決方案時，你能不能幫我們理解一下內部的命中率（hit rate）？例如你們在白板上構思這些東西時，大概要多少個點子，才能產出真正進入生產環境（production）的成果？相較於 AI 出現之前，你們在思考改進時，這些業務的運作方式有什麼不同？

Yeah. I mean it's really appreciate that question. The hit rate right now is quite high. Now if you think about -- if I step back about really what the -- if you will, the transformation we're going through on the product side is our historical approach to software development being the market leader in each one of our 21 software verticals is sort of methodical and lower risk.

可以。我很感謝這個問題。目前的命中率其實相當高。如果我退一步看——也就是說，我們在產品端正在經歷的轉型，本質上是：我們過去作為 21 個軟體垂直領域各自的市場領導者，歷史上的軟體開發方法相對更有章法、風險也更低。

In the AI world, you want to go at a frenetic pace and actually work at a little bit higher risk because the cost to develop is much lower. So the risk of a mistake is much lower. So that's been the -- if you really sort of condense the learning over the last two years, that is the learning of the organization. And so we're moving at just great pace and the development here has been great.

在 AI 的世界裡，你會想用更快、更密集的節奏推進，並且在更高一點的風險下工作，因為開發成本低很多。所以犯錯的成本也低很多。因此，如果把過去兩年的學習濃縮起來，這就是組織層面的核心學習。所以我們正以非常快的速度前進，而這裡的開發進展也非常好。

Now the also the learnings have been you start with putting AI chat in your products, not a lot of value there, you can monetize. Then you sort of do lighter weight agents that have lower entropy. And then you get into like real commercial-grade differentiated high-value agentic workflows, which is what now is being released to the market.

另外，我們也學到：一開始把 AI 聊天功能放進產品裡，其實可變現的價值不大。接著你會做一些較輕量的代理（agents），其不確定性（entropy）較低。然後你才會進入真正商用等級、具差異化且高價值的代理式工作流程（agentic workflows），而這正是現在開始對市場發布的內容。

But the customer signal that we get is incredible. I think I put in on social and LinkedIn when Procare released their first agentic feature. And mind you, it's not monetized in this particular case, the first one. But within four hours released, I think 20% of the customer base engaged with the feature, unheard of engagement in the software landscape that quickly.

但我們從客戶端得到的訊號非常驚人。我想我在社群和 LinkedIn 上有提到，當 Procare 發布他們第一個代理式功能時——請注意，在這個案例中，第一個功能還沒有變現（monetized）。但在發布後四小時內，我記得有 20% 的客戶群就使用了這個功能；在軟體產業中，這麼短時間內出現這種互動率是前所未見的。

Strata, with their financial decision intelligence tool, they hosted a call and they had more people on the call seeing the demo of this product than when they merged with Syntellis to give you a sense of the energy in the market.

Strata 的財務決策智慧工具（financial decision intelligence tool）也是，他們辦了一場電話會議展示產品 demo，參與觀看 demo 的人數甚至比他們與 Syntellis 合併時還多，讓你感受一下市場的熱度。

The number of customers in the beta and the early adopter for some of the Vertafore agents, normally they're capped at a dozen or two, and Amy left it uncapped, and there were more than a couple hundred. The early signal is quite high. I'm just trying to give you a couple data points on this. If the hit rate is, if we miss on a few, that's okay because we've added a little bit of that, if you will, risk given the speed in which we're developing.

Vertafore 的一些代理產品，在 beta 與早期採用者（early adopter）階段的客戶數量，通常會被限制在十幾或二十來家；但 Amy 沒有設上限，結果超過了兩百家。早期訊號非常強。我只是想給你幾個數據點。如果命中率上有幾個沒打中也沒關係，因為考量到我們開發速度之快，我們也加入了一點你所說的風險承擔。

And then on the M&A side, as you evaluate these companies, like I know the multiples are weird and we're maybe talking about like what multiple should you pay on a trough type result. How willing are you to underwrite inflections in these businesses. So like maybe you're willing to pay a headline sticker price that's way a higher multiple than we normally see from you because you're willing to underwrite something in the business? Like how confident are you in ability to kind of pick that here in this world?

另外在併購（M&A）方面，當你們評估這些公司時——我知道倍數（multiples）現在很怪，我們可能在討論的是：在景氣谷底（trough）型的結果上，你應該付什麼倍數。你們願意在多大程度上承擔並押注這些業務的拐點（inflections）？也就是說，你們可能願意付出表面上的標價（headline sticker price），其倍數遠高於我們過去看到你們通常會付的倍數，因為你們願意承作（underwrite）該業務的某些改善或成長？在這個環境下，你們對自己挑選這種機會的能力有多大信心？

We've never really -- we've never been the buyer and fixer of an asset ever, at least intentionally, right? We think that we're buying winners that have good momentum behind them, and that's what we're focused on. And increasingly, that will become easier to discern an AI era because the targets will have AI growth and earnings, AI-related growth -- product growth and earnings.

我們從來沒有——我們從來都不是那種買進資產再把它修好（buyer and fixer）的買方，至少不是刻意如此，對吧？我們認為我們買的是本來就在贏、且背後有良好動能的標的，而這就是我們的重點。而且在 AI 時代，這會越來越容易辨識，因為標的會有 AI 成長與獲利、以及 AI 相關的成長——產品成長與獲利。

And so -- but the concept of a core business that has some headwind attached to it and then somehow we can own it and magically improve it, not really what we do, and I don't see us doing that in the future. We're going to buy the winners and make them better.

所以——但那種核心業務帶著一些逆風，然後我們買下來就能神奇地把它改善的概念，並不是我們在做的事，我也不認為未來我們會這麼做。我們會買贏家，然後把他們變得更好。

Clarke Jeffries, Piper Sandler.

Clarke Jeffries，Piper Sandler。

I wanted to ask kind of two interrelated questions around the core margin momentum in the business. Just where do we stand on the quarter-to-quarter movement of the input costs or the mix in TEP? I think core margins were down 70%, but with less contribution from software. So I think that implies that TEP improved from a core margin perspective, but love to get sort of the background on what you're seeing and what you expect for second half?

我想問兩個彼此相關的問題，關於公司核心毛利率動能。就單季之間的變動來看，TEP 的投入成本或組合（mix）目前處於什麼狀態？我記得核心毛利率下降了 70 個基點，但軟體的貢獻更少。所以我認為這暗示 TEP 從核心毛利率角度看是有所改善的，但希望你能說明一下你們看到的背景因素，以及你們對下半年的預期？

Sure. I'll take that. So yes, you're right. It's mostly concentrated in TEP. So core margins were down for the enterprise 70 basis points and TEP was down substantially more than that. And it's really the same dynamics that happened in the second quarter are consistent with what happened in the first, which is at Neptune, they had more of a mix of mechanical meters that shipped, and those have carried higher input costs.

當然。我接受。所以是的，你說得對。主要集中在 TEP。因此企業的核心毛利率下滑了 70 個基點，而 TEP 的下滑幅度明顯更大。而這其實與第二季發生的相同動態一致，也與第一季的情況相符：在 Neptune，他們出貨的機械式水表占比更高，而這些產品承擔了較高的投入成本。

A big input cost for them is ingot, which has -- ultimately has copper in it. And you've seen what's happened with the price of copper and fulfilling data center demand. So that will -- in the second half, we've got some offset to that will -- on the price side. So we'll have some better offset. And we'll also have more static meter shipments in the second half that don't have as much that input cost.

他們一項很大的投入成本是錠材（ingot），而它——最終含有銅。你也看到銅價的走勢，以及為滿足資料中心需求所帶來的影響。因此在下半年，我們會有一些抵銷——在價格端會有一些抵銷。所以我們會有更好的對沖。而且下半年我們也會有更多靜態水表的出貨，這些產品的投入成本沒有那麼高。

And then on -- and then we just have more mix of reoccurring consumables across our NDI and Verathon business, which we love because that's higher recurring and it's more secular and more predictable, but they do carry a little bit of lower margin. So that's really the primary thing that's going on there.

另外——我們在 NDI 與 Verathon 業務中，經常性耗材的組合占比也更高；我們很喜歡這點，因為它的經常性更高、更具長期趨勢性、也更可預測，但毛利率會略低一些。所以這就是那邊主要正在發生的事情。

And then that dynamic in the second half just comps a little bit better. So if you look at Q2, it's usually -- it's been our high watermark, but it was a little bit lower this quarter. And then as you roll into the second half, we'll have better comps in the segment.

然後下半年這個動態在同比比較上會稍微更有利一些。所以如果你看第二季，通常——它一直是我們的高水位，但這一季略低一些。接著進入下半年，我們在該分部的同比比較會更好。

Perfect. And then just on the other side of it with the -- I think you mentioned core margins in software being down based off of the investment in the AI team. Just curious how you expect the investment to play out. I would imagine that it's heavier today in application software versus network software. Do you have insight to when investment might peak? Just any insight on the relative magnitude of investment in the AI team maybe between the segments or the sort of next 12 months?

很好。那再從另一面來看——我想你提到軟體的核心毛利率下滑，是因為對 AI 團隊的投資。我只是好奇你預期這項投資會如何展開。我猜目前在應用軟體上的投入會比網路軟體更重。你是否能分享投資可能何時達到高峰？以及未來 12 個月，AI 團隊在各分部之間的相對投資規模之類的任何看法？

Yes. I think you're right, right? Most of the, I'd say, activity right now is in application, but all the businesses will have some level of interaction with the AI team. We're sort of -- I think we're at maybe the halfway point of where we think that investment is going to be in terms of headcount, but probably much higher on spend because we've started at the more senior level, and then we'll add some more entry level and more junior folks as we go through. So a lot of that is already in our second quarter base and won't be that much incremental as we go forward.

是的。我想你說得對，對吧？目前大多數——我會說——活動都在應用端，但所有業務都會在某種程度上與 AI 團隊互動。我們大概——我想在人力編制（headcount）方面，可能已經到達我們預期投資規模的半途，但在支出上可能高得多，因為我們是從較資深的層級開始，接著在推進過程中會再加入一些入門與較初階的人員。所以其中很多已經反映在我們第二季的基期裡，往後不會再有那麼多的增量。

Brian Peterson, Raymond James.

Brian Peterson，Raymond James。

Neil, I know you're addressing a lot of different markets, but I'm curious, in what of those markets do you think you have the earliest appetite or demand from customers for AI solutions? And then maybe what verticals do you think will take a little bit longer to kind of test out those solutions before they step in and buy?

Neil，我知道你們涵蓋很多不同市場，但我很好奇，在這些市場中，你認為哪一些市場的客戶對 AI 解決方案會最早出現意願或需求？另外，哪些垂直領域你覺得可能需要更久一些，會先測試這些解決方案，之後才會進一步採購？

Yes. It's -- we spent a fair amount of time thinking about this. And I think it's -- maybe try to answer that on like on two different axes. I mean there's certainly the industry end market. And I don't know per se if you're in insurance versus health care that the mindset is different in those end markets, one more aggressive, one more conservative. So I don't think that's really the dynamic.

是的。我們在這方面花了不少時間思考。我想可以嘗試從兩個不同的軸來回答。當然有產業終端市場的差異。但我不確定例如保險相較於醫療保健，這些終端市場的心態是否真的不同——一個更積極、一個更保守。所以我不認為那才是主要動態。

I think the other one is we can compare and contrast like CenterReach, which is autism therapy and freight matching automation at DAT. And we have incredible take rates, adoptions in autism therapy and methodical week month-over-month growth rates are quite nice at DAT, but they're not like at escape velocity yet.

我認為另一個面向是，我們可以對比像 CenterReach（自閉症治療）以及 DAT 的貨運配對自動化。在自閉症治療方面，我們的採用率與導入率非常驚人；而在 DAT，循序漸進、按週按月的成長率也相當不錯，但還沒有到達爆發式成長的臨界點。

And why is the difference? And my -- our discernment on that is pretty simple, which is in autism therapy, you have something like 800 million therapy hours demanded, 300 million therapy hours supplied. There's a line outside the door of every clinic for care. And so to the extent that we can deliver AI solutions to give therapists more time to see patients, they're seeing patients, families are getting the services, the clinics revenues and earnings are going up, just everything is aligned.

那差異在哪裡？而我——我們的判斷其實很簡單：在自閉症治療領域，大約有 8 億個治療小時的需求，但供給只有 3 億個治療小時。每一家診所門外都有人排隊等候照護。因此只要我們能提供 AI 解決方案，讓治療師有更多時間看診，病患就能被看診、家庭能獲得服務、診所的營收與獲利都會上升——一切都高度一致。

And the DAT example, we can automate the manual -- the dozen or so phone calls to broker a load into no phone calls and do it for a fraction of the labor cost. But there, you're having to do a change management in the workflow about how loads are brokered. There's a human element of that. There's a change management element of that.

而以 DAT 為例，我們可以把人工處理——大約十幾通電話才能撮合一筆貨載——變成零通電話，並且用遠低於人力成本的一小部分來完成。但在那裡，你必須在工作流程上做變革管理，因為貨載是如何被撮合的流程要改變。其中有人為因素。也有變革管理的因素。

And then there's going to be flavors of that in between. And like, for instance, Vertafore probably sits somewhere in the middle between those 2. So I think it's going to be more about just the level of human change more than it is like industry that drives sort of the rate of adoption.

而中間還會有各種不同的情況。例如 Vertafore 可能就介於這兩者之間。所以我認為，驅動採用速度的更關鍵因素，是人員變革的程度，而不是產業本身。

No, that's great color. And maybe just -- I know you mentioned some of your conversations with sponsors. Do you have any sense of where they are in terms of their AI investments? And if those are really ramping up, how do we think about that investment or margin cadence of the targets you would be looking at, assuming we do see a swing of activity in the next couple of quarters?

不，這些說明很有幫助。再問一下——我知道你提到與贊助方（sponsors）的部分對話。你是否能感覺到他們在 AI 投資方面目前處於什麼位置？如果那些投資真的正在加速，我們該如何看待你們會關注的標的在投資或毛利率節奏上的變化，假設未來幾季活動確實轉為更活躍？

Yes. This is probably unfair for us to characterize all sponsors the same. We'll try to do this out. We'll do it without naming names. But there's a small number of sponsors that when the gun went off a couple of years ago with AI, like us, decided to put all of their energy around AI into playing offense, driving product velocity, driving revenue growth, and that was the principal focus.

是的。我們把所有贊助方一概而論，可能不太公平。我們試著把它拆開來說。我們會在不點名的情況下說明。有少數贊助方在幾年前 AI 的槍聲一響時，就像我們一樣，決定把所有圍繞 AI 的能量用在進攻端：提升產品迭代速度、推動營收成長，而那是主要焦點。

The vast majority of sponsors when the government off said, my goodness, I've got an opportunity to take a bunch of cost out of my business, and I'll be able to capitalize that. And now they're realized, boy, I've got to get on the product journey because I can't sell a software business unless it has a viable AI sort of product road map in front of it.

絕大多數贊助方在槍聲一響時則想：天啊，我有機會從我的業務中砍掉一大堆成本，而且我還能把它資本化。而現在他們才意識到：糟了，我必須踏上產品之旅，因為如果一家軟體公司沒有一個可行的 AI 產品路線圖在前面，我就賣不掉它。

So as a general matter, I think we're ahead on the product side versus the sponsors, but they're catching up. I mean it's -- they know they have to do the agentic SKU delivery and show some defensible increase in growth rate to be able to sort of sell a business at a premium price, whatever the definition of premium is going forward. And -- but there -- it's a broad landscape, and you can imagine there's different shades to gray across the sponsors on the AI front. I hope that answered your question.

所以整體而言，我認為我們在產品端相較於贊助方是領先的，但他們正在追趕。我的意思是——他們知道必須做到代理式（agentic）的 SKU 交付，並展現某種可防禦的成長率提升，才能以溢價出售一家公司——不論未來對「溢價」的定義是什麼。而且——但這是一個很廣的版圖，你可以想像在 AI 方面，各家贊助方之間會有不同程度的差異。希望這回答了你的問題。

Daniel Jester, BMO Capital Markets.

Daniel Jester，BMO 資本市場。

I think we've maybe talked about this in a couple of different ways, but maybe I'll just double-click on it. I really appreciate the color around SoftWriters and sort of the detailed sort of in the prepared remarks about the opportunity there. And it sounds like there's incremental monetization from -- you had a whole bunch of features that you talked about that have been launched across the portfolio in the prepared remarks as well.

我想我們可能已經用幾種不同的方式談過這件事，但我或許再深入一點。我非常感謝你們對 SoftWriters 的補充說明，以及在事先準備的發言中對那裡機會的相當詳細的描述。而且聽起來，除了—你們在事先準備的發言中也提到，整個產品組合已推出一大堆功能—之外，還有額外的變現空間。

Can you help us maybe understand the direct monetization opportunities versus sort of the feature and functionality and quality of life improvements that you're launching? Just as we think about going into next year, how much potential monetization uplift from AI is coming in the pipeline?

你們能否協助我們理解，直接的變現機會，與你們正在推出的功能與功能性、以及生活品質（使用體驗）改善之間的差異？就我們展望明年而言，AI 在管線中帶來的潛在變現提升大概有多少？

Yes. So maybe I can spend a minute on how we're going to monetize and then I'll try to -- we won't be able to give you a lot of color on the exact '27 monetization amounts -- but I think let me walk through this, and I think it will become evident. So I think we talked about this last quarter, but the principal way we're going to monetize these agentic SKUs is not going to be the SoftWriters example, which is straight transactional. Our customers have been very clear. They need to have an amount that they can budget and they can sort of understand an envelope -- financial envelope in which they can operate in.

是的。所以我或許先花一點時間談談我們將如何變現，然後我再嘗試—我們無法就確切的『27 年變現金額提供太多細節—但我想我先帶大家走一遍，應該就會很清楚。所以我想我們上季談過，變現這些代理型（agentic）SKU 的主要方式，不會是 SoftWriters 那種直接按交易計價的例子。我們的客戶講得非常清楚。他們需要一個可以編列預算的金額，並且能夠理解他們可運作的範圍—也就是財務上的「預算框架」。

So I believe the vast majority of our agentic SKUs will be sold in, if you will, an agent layer, an orchestration layer, the customers will pay a subscription for access to that. And then based on the consumption of the agents and the value that we deliver, they will buy larger tiers of utilization, but it's not going to be straight consumption. Imagine it's going to be stair step consumption. That will be the majority of the way that we monetize. There'll be a few businesses, DAT SoftWriters that already have transactional pricing for the core of what they do that will transactionally price to AI solutions.

因此我相信，我們絕大多數的代理型 SKU 會以你可以稱之為代理層（agent layer）、編排層（orchestration layer）的形式來銷售，客戶會為取得該層的存取權而支付訂閱費。接著依據代理的使用量以及我們交付的價值，他們會購買更高的使用等級（tier），但不會是純粹按用量計費。可以想像成階梯式的用量。這將是我們變現的主要方式。也會有少數業務，例如 DAT SoftWriters，本來核心業務就採交易式定價，AI 解決方案也會採交易式定價。

And then a couple, the most notable will be Deltek is they're piling the majority of their AI agentic features into their cloud product, and they will sort of monetize it on the uplift of their on-premise to the cloud and drive pricing -- sort of additional pricing in the lift and shift. So based on some of the questions that have already been asked today, the big X factor, and we're just going to call it as we see it, is what's the rate of adoption at the customer level, right?

另外還有幾個，最值得注意的是 Deltek：他們把大多數 AI 代理型功能打包進他們的雲端產品，並且會透過把客戶從地端（on-premise）升級到雲端所帶來的提升來變現，並在搬遷上雲（lift and shift）過程中推動—也就是額外的定價提升。所以基於今天前面已經被問到的一些問題，最大的 X 因素—我們就實話實說—是客戶端的採用速度到底有多快，對吧？

We have the products. We're going to have more products in three months and more products in six months. And that we're on that train, and we're excited by that velocity. But what's the pace of adoption at the customer level? That's the unknown. And so until we get clear signal on that, we have to be unfortunately vague on the magnitude of impact in the near term.

我們有產品。三個月後會有更多產品，六個月後也會有更多產品。我們已經在那班列車上，並且對這樣的速度感到興奮。但客戶層級的採用節奏是什麼？那是未知數。因此在我們得到明確訊號之前，對於短期影響的幅度，我們不得不（很遺憾地）保持相對模糊。

No, that's really helpful context. I appreciate all that color. And then maybe just as my follow-up, I think in the prepared remarks, you talked about sort of ground to cloud conversions starting to improving. I guess maybe can we just spend a moment, I guess, why do you think that this is improving? And I guess, the confidence level that we should continue to see that improving for the rest of the year into next year.

不，這些背景非常有幫助。我很感謝這些補充說明。接著作為追問，我想在事先準備的發言中，你們提到從地端到雲端的轉換開始改善。我想我們花一點時間談談，你們為什麼認為這正在改善？以及我們對於今年剩餘時間到明年仍會持續改善的信心程度應該如何看待。

Sure, Dan. I can take that. So yes, I think we're seeing it -- we have seen it for a few years now at Aderant, and it's picked up velocity there. So that's a legal end market, and we started with some of the smaller firms and now we're starting to move into some of the larger firms, which is encouraging. our Power Plant business has moved their tax solution into the cloud, and we're getting great adoption there. And now they're moving their core accounting -- tax accounting module into the cloud or suite into the cloud, and the adoption is picking up there.

當然，Dan。我來回答。是的，我想我們在 Aderant 已經看到這個趨勢—其實已經看到好幾年了，而且那裡的速度正在加快。那是法律終端市場，我們先從一些較小的事務所開始，現在也開始往一些較大的事務所推進，這很令人鼓舞。我們的 Power Plant 業務已把他們的稅務解決方案移到雲端，在那裡的採用情況非常好。而現在他們正把核心會計—稅務會計模組移到雲端或把整套套件移到雲端，那裡的採用也正在加速。

And then lastly, as Neil talked about, at Deltek, they're our largest maintenance base today. Putting new features into the cloud is certainly going to increase that adoption and they actually just end of life on their point solution in the GovCon space, which they're going to only go back on that. So they've made -- they've declared that the customers are going to move into the cloud, and that will be a big sort of lift for us over the next five years or so.

最後，如 Neil 所提到，在 Deltek，他們目前是我們最大的維護（maintenance）客戶基礎。把新功能放到雲端當然會提高採用率，而且他們其實在 GovCon 領域的點狀解決方案已經走到產品生命週期終點（end of life），之後只會回到那個方案上。所以他們已經—已經明確宣示客戶將會移轉到雲端，而這在未來大約五年左右會為我們帶來相當大的推升。

Dylan Becker, William Blair.

Dylan Becker，William Blair。

Neil, maybe for you, obviously, a lot of conversation around AI and agents and how that's kind of TAM expansive as you go into the labor segment. But maybe wondering on the vertical approach, right, how this validates not only as the workflow system, your ability to identify those opportunities to expand the scope, but also validate the ROI and maybe improve the attach, right, so to draw against that expanded TAM, if that makes sense? And maybe adjacently, right, like why customers would buy from you versus somebody else with an agent as you are kind of that workflow orchestration layer?

Neil，這題可能給你。顯然大家談了很多 AI 和代理，以及當你們進入勞動（labor）環節時，這如何擴大 TAM。但我想從垂直領域策略的角度來問，這如何不僅驗證你們作為工作流程系統（workflow system）的定位、你們識別那些擴大範疇機會的能力，也能驗證 ROI、並可能提高附加率（attach），也就是為了把握那個擴大的 TAM（如果這樣說得通的話）？另外相鄰的問題是，當你們作為工作流程編排層（workflow orchestration layer）時，為什麼客戶會向你們買，而不是向其他也有代理的人買？

I'll try to attack all of that. Hold me accountable if I fall short. So the ROIs, going back to the earlier question about the hit rate, the ROIs are clear and demonstrable. I mean it's -- I think I said it some of the cases in the Vertafore going from an hour to minutes, you have people that spend hours a day doing tasks and now take 30 minutes. And so it's very, very clear about the value that the customer receives or $100 to $200 to broker load and we charge a fraction of that or it's a very hard, very easy to identify ROI. So we're quite bullish by that, and that's why the hit rate early on and the signals are quite positive.

我會試著把這些都回答到。如果我講得不夠完整，請你提醒我。關於 ROI，回到先前那個命中率（hit rate）的問題，ROI 是清楚且可被證明的。我的意思是—我想我在一些案例中提過，例如 Vertafore，把一小時縮短到幾分鐘；有人每天花好幾個小時做的工作，現在只要 30 分鐘。因此客戶獲得的價值非常、非常清楚；或者每筆保單（broker）上架（load）成本是 100 到 200 美元，而我們只收其中一小部分—這是非常硬、也非常容易辨識的 ROI。所以我們對此相當樂觀，這也是為什麼早期的命中率與訊號都相當正面。

So why us is a question that we spend a lot of time on, and we're -- the signal here is just getting clearer and clearer and clearer. It matters a lot. I mean, it's the classic moat framework, but it matters a lot when you have a system of record position, which is the most of what we do, you have the data, but importantly, you have the workflow depth.

至於為什麼是我們，這是我們花很多時間思考的問題，而這裡的訊號也越來越清楚、越來越清楚、越來越清楚。這非常重要。我的意思是，這是典型的護城河（moat）框架：當你擁有記錄系統（system of record）的地位—這是我們大多數業務所處的位置—你有資料，但更重要的是，你有工作流程的深度。

So think about that SoftWriters example. Like, yes, we know how to validate a pharmacy entry, order entry. There's the AI tool to do that but delivering it precisely in the workflow with no latency is incredibly important. There are the effects whereas we -- there's a learning curve and a network effect, if you have a flywheel effect in all these solutions as we get feedback, the models get better, the prompts get better, the result gets better. And so it more use -- gets more use. So there's a bit of a network effect there.

所以想想 SoftWriters 的例子。像是，沒錯，我們知道如何驗證藥局的輸入、訂單輸入。有 AI 工具可以做到，但要在工作流程中精準交付、且沒有延遲，極其重要。此外還有一些效應：當我們取得回饋時，模型會變得更好、提示詞（prompts）會變得更好、結果會變得更好；在所有這些解決方案中，會有學習曲線與網路效應，或者說有一種飛輪效應。因此使用量會更多—所以那裡確實有一些網路效應。

I think 18 of our 21 software businesses are operating in a regulated end market. So there's a regulatory burden that is -- that sort of then leads into like a very high level of customer trust. Candidly, like why a lot of vertical market software exists is because of you have end markets that are different for some reason, oftentimes from a regulated point of view, which makes trust. And then let's just not forget about distribution. I mean it's a huge deal when you can release a feature or a product and you have a customer base that is sort of asking for that.

我認為我們 21 家軟體業務中有 18 家是在受監管的終端市場中運作。因此會有監管負擔——而這也在某種程度上帶來非常高的客戶信任度。坦白說，許多垂直市場軟體之所以存在，是因為某些終端市場基於某些原因（往往是從監管角度）而有所不同，這就建立了信任。另外也別忘了通路。我的意思是，當你能發布一個功能或產品，而你有一個客戶基礎正好在要求那個東西時，這是非常重要的。

They didn't even know they needed it and they see it like, I need that right now, and so all of that comes together about why AI is a winning solution in these vertical markets. And we have just a high degree of -- a high and increasing level of confidence in that.

他們甚至不知道自己需要它，但一看到就會覺得「我現在就需要」，所以這些因素加總起來，就解釋了為什麼 AI 在這些垂直市場是一個勝出的解決方案。而我們對此有很高——而且持續提高的信心。

Perfect. That's really helpful. And maybe, Jason, just a quick one for you as well, too, as we kind of think about the outlook, right? We had a strong first half of the year. We're seeing a slight kind of acceleration and improvement in the second half. I guess, given a lot of kind of the agentic opportunity we're talking about, I know it's not immediately monetizable, but probably improving buyer sentiment.

很好。這真的很有幫助。也許，Jason，我也想快速問你一個，因為我們在思考展望，對吧？我們上半年表現強勁。我們看到下半年有些微的加速與改善。我想，考量我們談到的許多代理式（agentic）機會，我知道它不會立刻變現，但可能會改善買方情緒。

I guess, how would you attribute or weight strength in the first half, kind of the mechanics of normalization in some of those larger businesses versus maybe that incremental improving sentiment, if you will, kind of across those three buckets as we think about the outlook for the balance of the year?

我想問的是，你會如何歸因或權衡上半年的強勁表現——也就是某些較大型業務中正常化的機制——相較於也許那種邊際上改善的情緒（如果你願意這麼說），當我們從這三個面向來看今年剩餘期間的展望時？

Yes. I mean, obviously, we had a good first half. TEP outperformed the second quarter and most of our races has been based on the confidence of the first half. I think the second half is a lot of mechanics as we've talked about before, with Central Reach and SubSplash turning organic. And so we haven't really baked in anything meaningful for AI above what the businesses are delivering today. And so that's more of plan for '27 at this point.

是的。我的意思是，顯然我們上半年表現不錯。TEP 在第二季表現優於預期，而我們大多數的調升（races）是基於對上半年的信心。我認為下半年很大一部分是我們之前談過的機械性因素，因為 Central Reach 和 SubSplash 轉為有機成長。因此，除了各業務今天已經在交付的內容之外，我們並沒有把任何有意義的 AI 貢獻納入。所以就目前而言，那更像是 2027 年的計畫。

Joe Vruwink, Baird.

Joe Vruwink，Baird。

I'll just do one. But talking about the organic recurring software, I think it was up 7% in the quarter. You then have reoccurring down 2% and nonrecurring down 4%. I guess, one, how did the latter two compare to expectations? And then two, given some of the early reports in software have maybe alluded to like deal timing issues around license transactions. Are you seeing anything pop up there that is at all a leading indicator of demand interest ahead of the second half?

我就問一個。談到有機的經常性軟體收入，我想本季成長了 7%。接著你們的 recurring（經常性）下降 2%，nonrecurring（非經常性）下降 4%。我想問，第一，後面兩項與預期相比如何？第二，鑑於一些軟體公司早期的報告可能提到授權交易在成交時點上有些問題。你們是否看到任何相關跡象，可能作為下半年需求興趣的領先指標？

Sure, I can take that. So in application general, yes, nonrecurring was down a little bit. We expected that. Actually, I thought it was going to be a little bit worse because we didn't have the Deltek large license deal baked in. That was just really timing between PowerPlan and others around license and service activity. So nothing to really call out there in our network recurring was down a little bit at our AFA business. So on the call, just more large customers with lower unit economics, and that actually hits the revenue line. So really, that was the only two things to call out.

當然，我可以回答。就應用（application）整體而言，是的，非經常性收入確實小幅下滑。我們有預期到。其實我原本以為會更差一些，因為我們沒有把 Deltek 的大型授權交易納入。那主要只是 PowerPlan 和其他業務在授權與服務活動之間的時點差異。所以那邊沒有什麼特別需要點出的。至於我們的 network recurring（網路經常性）在 AFA 業務上小幅下滑。所以在電話會議上提到，主要是更多大型客戶、單位經濟較低，而這確實會影響營收線。所以真正需要點出的就這兩件事。

And then actually, iPipeline also in the network segment had some service timing. But -- so overall, though, I would just say that bookings continue to look favorable. We were up. We had a tough comp on the second quarter of last year, but we're still up on a TTM basis, mid-single-digit plus. So pipelines look really strong. I think we'd like to see -- we would expect to see that number creep up in terms of TTM bookings as we move into the second half of the year. So the environment has been good for us in terms of commercial activity.

另外，其實 iPipeline 在 network（網路）部門也有一些服務時點的影響。不過——整體而言，我會說 bookings（訂單）看起來仍然有利。我們是成長的。去年第二季的比較基期很硬，但以 TTM（過去十二個月）來看，我們仍是中個位數以上的成長。所以 pipeline（商機管道）看起來非常強勁。我想我們希望看到——也預期會看到——隨著進入下半年，TTM bookings 的數字會逐步上升。因此，就商業活動而言，環境對我們一直不錯。

And just to spike one thing, Jason said on the deal -- on the services timing, this is booked deals, and it was pushing up deliveries of deals versus bookings that pushed.

再補充一點，Jason 在交易——在服務時點上提到的是：這些是已簽訂（booked）的交易，差異在於交付時點往前推，而不是 bookings 被往後推。

George Kurosawa, Citi.

George Kurosawa，Citi。

A lot of discussion in the industry about rising AI costs, token budgets kind of swelling, anecdotes of token maxing. Maybe just when I think about your position, it seems like I'm curious if there's a part of the value prop if you think about the portfolio of businesses that is built around or emerging around the more efficient deployment of AI. And then, Jason, maybe just your approach to managing AI costs and governing those internally.

業界有很多討論在談 AI 成本上升、token 預算膨脹，以及 token 用到上限的案例。從你們的立場來看，我很好奇：如果看你們的業務組合，是否有一部分的價值主張是建立在——或正在形成於——更有效率地部署 AI？另外，Jason，也想請你談談你們在內部管理 AI 成本與治理的做法。

Sure. So yes, obviously, spend increased for us, just like everyone else. We're up about -- on an annualized basis, about 3 times since January. We've been really pushing -- we think what's going to go up by the end of the year, by the way. I think we've been pushing adoption at the enterprise level and getting the business to fully move to agentic coding, and we're in the early days of that.

當然。所以是的，我們的支出確實增加了，跟其他人一樣。以年化來看，自一月以來大約增加到 3 倍。我們一直在大力推動——順帶一提，我們認為到年底還會再上升。我想我們一直在推動企業層級的採用，並讓各業務全面轉向代理式（agentic）程式開發，而我們仍在早期階段。

I will say we had our CFO Summit a couple of months ago and all of us were talking about that. We've been sharing a lot of best practices and tools around controlling spend, be it gateways or auto routing or reporting controls. So that's fully in swing. I think the good news, it's kind of a hallmark of our model is that the businesses are driving local ownership, right, around balancing speed with control. And so that's sort of how we've managed it.

我會說，我們幾個月前辦了 CFO Summit，我們大家都在談這件事。我們一直在分享許多控制支出的最佳實務與工具，不論是 gateway、auto routing，或是報表控管。所以這些都已全面展開。我想好消息是，這也算是我們模式的一個特徵：各業務在本地端有所有權（local ownership），對吧，在速度與控管之間取得平衡。所以我們大致就是這樣管理的。

In terms of how it gets into the product, I mean, I think we use a lot of the -- we don't have to use the frontier models in our product. We're using some of the lower-level tools to do that. I mean you don't always have to use AI in every case. By the way, you can codify certain rules and don't have to continue to ping and do calls when you're spooling up the agents. So overall, we're obviously mindful of it, but I think we're in a good spot there in terms of managing the token spend.

至於如何導入產品，我的意思是，我們在產品中不一定要使用最前沿（frontier）的模型。我們會用一些較低階的工具來做到。我的意思是，你不一定在每個情境都必須用 AI。順帶一提，你可以把某些規則寫死（codify），在啟動代理（agents）時就不必一直去 ping、一直發起呼叫。所以整體而言，我們當然會留意，但我認為在管理 token 支出方面，我們目前處在不錯的位置。

Okay. Great. And then I did also want to ask about Aderant. It seems like the business has continued to see good momentum. The legal tech space is rapidly evolving, a lot of fast-growing privates in the space, maybe not directly competing with Aderant, but at least in the same world. The hourly billings model itself seems to be maybe under some level of scrutiny. Just how do you feel that business is positioned over the long term and kind of the durability of momentum there?

好的。很好。另外我也想問 Aderant。看起來這項業務持續有不錯的動能。法律科技領域正在快速演變，這個領域有很多成長很快的私人公司，可能不會直接與 Aderant 競爭，但至少在同一個生態圈。按時計費（hourly billings）模式本身似乎也正受到某種程度的檢視。你們如何看待這項業務的長期定位，以及目前動能的耐久性？

Yes. We -- the Aderant business is just awesome. Just so everybody frames, a lot of what you've talked about is sort of in the practice of law. the Harvey and Lagoras and Anthropic of the world about how you sort of make the practice law more efficient. Aderant is all about the business of law. So we're the ERP billing collection, cash cycle, time capture, all of that. And it's just been a great business. since we own the business, the market share has gone from like 35% to 65%. The growth rates tripled. It's just been a great business for us.

是的。我們——Aderant 這項業務真的非常出色。為了讓大家有個框架，你們談到的很多內容比較偏向於法律實務，也就是像 Harvey、Lagoras、Anthropic 這些公司在探討如何讓法律實務更有效率。Aderant 則完全聚焦在「法律業務經營」本身。所以我們做的是 ERP、開票與收款、現金循環、工時記錄等所有這些。而且這一直是一門很棒的生意。自從我們擁有這項業務以來，市占率大概從 35% 提升到 65%。成長率也變成原來的三倍。對我們來說，這一直是一項非常好的業務。

In terms of billing, yes, I mean, we -- there was -- we did a tuck-in last year, ProPricer at Aderant, which is specifically leaning into how do you help law firms think about strategically changing their pricing models to be more fixed fee or fixed fee with some overages, et cetera. It still has pretty low uptake across -- not ProPricer, just that pricing model, relatively low uptake across the industry.

就開票而言，是的，我的意思是，我們——去年在 Aderant 做了一個補強式併購（tuck-in），收購了 ProPricer，這項產品特別著力於：如何協助律所從策略面思考調整其定價模式，使其更偏向固定費用，或固定費用加上部分超額費用等。不過在整個產業中——不是指 ProPricer，而是指那種定價模式——目前採用率仍然相對偏低。

But you can see that happening in certain situations and Aderant that's a business of law question that we will do -- sorry, virtual price [VPD], sorry, I gave the wrong answer, wrong deal we did there. But we're excited by that opportunity to partner with our customers in that regard as this market evolves.

但你可以看到在某些情境下確實正在發生，而 Aderant——這是一個「法律業務經營」層面的問題——我們會做——抱歉，是 Virtual Price [VPD]，抱歉，我剛剛答錯了，講成了我們做的另一筆交易。但隨著市場演進，我們對於在這方面與客戶合作的機會感到很興奮。

A lot of the business, the practice-law leader folks, by the way, very much want to partner with us because we're the one that have the matter sort of number and matter details and they need to know what matter their AI solutions are working in so that they can be in the system the right way. So we partner with those folks as appropriate.

順帶一提，很多做法律實務（practice of law）那一端的領導者其實非常希望與我們合作，因為我們掌握案件（matter）的編號與案件細節，他們需要知道他們的 AI 解決方案正在處理哪一個案件，才能以正確的方式接入系統。所以我們會在適當的情況下與這些夥伴合作。

Josh Tilton, Wolfe Research.

Josh Tilton，Wolfe Research。

I'll keep it to one given the time. I guess if we step back from like a high-level perspective, it felt like we came into the year with an outlook for the year that had some conservatism in it for the different things going on with Deltek, DAT and Neptune. We're halfway through the year. The full year guide is going up a little bit. It sounds like there are signs of improvement in all three of those businesses.

考量時間我就問一題。我想如果我們從更高層次的角度退一步看，感覺我們在年初給出的全年展望，對於 Deltek、DAT 和 Neptune 各種正在發生的事情，帶有一些保守因素。我們現在已經過了半年。全年指引小幅上調。聽起來這三項業務都有改善的跡象。

Can you just help us understand what, if any, conservatism is left in the guide? Did anything about how you're being conservative change regarding these three businesses? How do we just reconcile or understand full year guidance going up for the year, but also where you're still being prudent and conservative in your outlook for those three businesses that I mentioned?

你們能否幫我們理解一下，指引裡還剩下哪些（如果有的話）保守因素？你們在這三項業務上採取保守假設的方式，有沒有任何改變？我們該如何調和或理解：全年指引上調，但同時你們對我提到的這三項業務仍保持審慎與保守的展望？

Yes. I mean I'll take a crack at that and then ask Jason if he wants to add any color. So just to reiterate what's already been said, it was a lot of the increase is sort of in the bank. It's first half doing better than we thought.

是的。我的意思是，我先嘗試回答，然後再請 Jason 看他是否要補充一些說明。所以重申一下前面已經說過的，這次上調有很大一部分其實已經算是「到手」了。也就是上半年表現比我們原先預期更好。

And then there's the mechanics in the second half of turning organic and then sort of easier comps, especially in the third quarter at TEP. And then we have DAT doing a little bit better. So that's -- those are the good things. But there's still some range of outcome here that we want to remain prudent in the outlook so that we can finish the year with strength.

然後下半年在機制上會轉為有機成長，再加上比較基期變得更容易，尤其是 TEP 的第三季。另外 DAT 的表現也稍微好一些。所以這些都是正面因素。但結果仍然存在一定的區間，我們希望在展望上保持審慎，這樣我們才能以強勁的態勢完成今年。

Ken Wong, Oppenheimer.

Ken Wong，Oppenheimer。

I'll just ask one as well. You talked about DAT market improvement. Is that something baked into the guide in the back half? And then any thoughts on the freight broker liabilities ruling and how that might impact the business?

我也只問一題。你們提到 DAT 市場改善。這是否已經反映在下半年指引中？另外，對於貨運經紀人責任（freight broker liabilities）的裁決，你們有何看法，以及它可能如何影響業務？

Ken, yes, so we -- the way the DAT's business works, as you know, it's main on the carrier front, it's monthly subscriptions. So we've seen good progress in the first half. That will continue -- we think that will continue into the second half, not baking in a massive inflection, have just continued slight improvement. And how that works sort of snowballs and then that will roll over into '27? So playing more for '27 growth there, but certainly have some of that baked into the second half in [NS].

Ken，是的，所以我們——如你所知，DAT 的業務模式主要在承運人端是按月訂閱。因此我們在上半年看到不錯的進展。我們認為這會延續到下半年；我們沒有把一個巨大的拐點（inflection）納入假設，而是僅假設持續小幅改善。而這種改善的運作方式會有點像滾雪球，然後會延續到 2027 年？所以那邊更像是在為 2027 年的成長做布局，但我們確實也把其中一部分反映在下半年[NS] 的指引裡。

And on the [Montgomery Scots] broker liability case, a couple of things on that. The punchline is, we think this is a good thing for us in that. What happened -- first of all, what's happened is a bunch of state laws on this basically got harmonized at the federal level. So at least we know there's one set of rules in which to play by now, which is good.

至於 [Montgomery Scots] 的經紀人責任案件，這裡有幾點。結論是，我們認為這對我們是好事。發生的事情——首先，原本各州在這方面的一堆法律，基本上在聯邦層級被統一（harmonized）了。所以至少現在我們知道只需要遵循一套規則，這是好事。

And then what this really does is put more emphasis on the brokers to vet the carriers, which is precisely what Convoy and DAT does extremely well. This goes into the whole fraud situation from a year ago. And so this is -- we're part of the solution here, and we're encouraged that we get one set of rules to play with and now will be the part of a solution that everybody is focused on.

然後這件事真正帶來的影響，是讓經紀人更重視對承運人的審核（vet），而這正是 Convoy 和 DAT 非常擅長的事情。這也與一年前的整個詐欺問題相呼應。因此我們——我們是解決方案的一部分；我們也很受鼓舞，因為現在有一套規則可遵循，而我們將成為大家所關注解決方案的一部分。

This concludes our question-and-answer session. We will now return to Zack Moxcey for any closing remarks.

我們的問答環節到此結束。接下來我們將回到 Zack Moxcey，請他做結語。

Thanks, everyone, for joining us today. We look forward to speaking with you during our next earnings call.

感謝各位今天的參與。我們期待在下一次財報電話會議中與各位交流。

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

本次電話會議已結束。感謝各位出席今天的簡報。您現在可以斷線。