儒博科技 (ROP) 2026财年第三季度业绩电话会：内生增长达7%，上调业绩指引

Roper Technologies（NASDAQ: ROP）公布了强劲的2026财年第三季度业绩增长，上调了全年调整后每股稀释收益（EPS）预期，并强调其垂直软件产品组合中人工智能（AI）产品的早期商业化成果。

核心要点

• 第三季度营收同比增长8%至22亿美元，其中有机增长率为7%。两个软件业务板块的有机经常性收入增长了8%。
• EBITDA增长7%至8.64亿美元，EBITDA利润率为39.8%。调整后每股稀释收益增长15%至5.89美元。
• 管理层将全年调整后每股稀释收益指引上调并收窄至22.35-22.40美元，同时维持总营收增长8%以上及有机增长约6%的预期。
• 自由现金流为8.05亿美元，占营收的37%；尽管受债券利息支付时点以及与公司Ilumya业务ERP实施相关的收款延迟影响，自由现金流同比下降了4%。
• Roper部署了4.36亿美元，为iPipeline、Procare和Strata完成了三项补强收购。净债务与EBITDA的比率降至2.9倍，管理层估计未来12个月内拥有超过60亿美元的资本部署能力。
• AI商业化开始对新签订单产生贡献。管理层表示，CentralReach约75%的新签订单与AI专有或AI相关，而Aderant的这一比例约为25%。

关键财务数据

业务与运营表现

应用软件业务

应用软件业务营收增长7%，其中有机增长为6%。EBITDA利润率扩大30个基点至43.7%。

Aderant继续扩大其AI智能体平台，而Vertafore推出了更多工作流智能体并引入了基于价值的点数模式。Vertafore的AI点数结构设计为经常性合同，可在其Velocity平台上的当前及未来生成式解决方案中使用。

在私营部门解决方案和稳健的政府承包新签订单支撑下，Deltek再次取得了令人鼓舞的季度业绩。客户从永久许可转向云端方案的兴趣有所增加，但管理层继续保持观望姿态，因为更广泛的政府承包市场仍存在不确定性。

Strata于9月1日全面推出了其“财务决策智能”产品。这项基于AI的附加订阅服务可帮助医疗财务团队识别业绩驱动因素，并将分析与成本、利润率及资源配置工作流连接起来。

Roper还为Procare收购了Playground，为Strata收购了c. myers，分别增加了托儿服务管理和财务风险规划功能。

网络软件业务

网络软件业务营收增长8%，有机增长为5%。EBITDA利润率下降130个基点至52.4%，不过核心利润率与去年同期基本持平。

DAT通过网络扩张、提升ARPU（每用户平均收入）以及推广数字货运市场解决方案加速了增长。管理层形容货运供给侧趋势令人鼓舞，但表示在将市场定位为正在经历更广泛的复苏之前，仍需要更强劲的货运量。

Foundry的年度经常性收入创下历史新高，而SoftWriters受益于其AI订单录入解决方案的持续推广。Roper还为iPipeline收购了Origo，扩大了其在英国财富与养老金市场的业务版图。

技术赋能产品业务

技术赋能产品业务的总营收和有机营收均增长了10%。EBITDA利润率下降90个基点至34.3%。

NDI精准测量解决方案需求强劲，而Verathon在BFlex、GlideScope和BladderScan等产品上均实现增长。管理层表示，客户对Verathon的Cflex一次性膀胱镜反应积极。

Neptune在经历了强于预期的上半年后有所放缓。水表出货量接近预期，但在静态水表项目推迟后，产品结构转向价格较低、利润率较低的机械水表。管理层继续认为向静态和超声波水表转型的长期趋势保持完好。

管理层业绩指引

修正后的全年每股收益范围中值比此前指引增加了0.15美元，比公司最初的展望增加了0.95美元。

在获得监管部门批准的前提下，管理层预计第四季度将从Indicor出售其剩余流体控制业务中获得约6亿美元的税前收益。

风险与关注领域

• Neptune的项目时点和产品结构仍存在不确定性。管理层表示，进入2027年，该业务的潜在结果区间更宽。
• Deltek从永久许可向云订阅的转变支持了经常性收入，但相比永久许可销售可能会延缓收入确认。
• DAT的改善主要受货运供给侧趋势推动；管理层希望在宣布全面复苏之前看到更强劲的货运需求。
• AI的更广泛应用可能会受到客户预算周期、AI治理审查、数据安全防护和变更管理要求的制约而放缓。
• 由于Ilumya在ERP实施后需解决订单至收款环节的问题，其收款延迟可能会部分延伸至2027年第一季度。
• 最近完成的三项补强收购预计在第四季度可能对部分业绩产生轻微稀释作用，协同效应预计将在2027年显现。

分析师问答环节亮点

管理层表示，在Roper的21家软件业务中，有17或18家已在市场上推出了AI智能体解决方案。排除工作流已基本实现自动化或不属于常规软件运营的业务后，管理层表示，所有拥有实质性AI机遇的业务目前均已提供AI工具。

Roper阐释了三种主要的AI变现模式：包含使用点数的固定经常性订阅、直接按用量计费以及高价云端附加组件。预计订阅和按量计费模式将较早贡献收益，而与云迁移挂钩的变现可能需要更长时间。

关于AI基础设施成本，管理层表示目前相关产品能产生类似于软件的高毛利率。Roper优先考虑开发和部署速度，同时采用可跨多个AI模型分发工作负载的模型路由架构，随着时间的推移可能会降低成本。

在资本分配方面，管理层对2027年的收购环境保持谨慎乐观。Roper在本季度暂停了股票回购，以保持资产负债表的灵活性用于并购，但表示如果收购机会未能实现，回购仍是一项选择。

管理层将2027年的初步构想形容为整体具有建设性，但强调正式规划尚未开始。Deltek的增长势头和DAT的改善状况是积极因素，而Neptune仍是主要的不确定性领域，NDI的增长预计将趋于常态化。

业绩电话会议完整文字记录

---

完整财报电话会议逐字稿

管理层陈述

Operator

Good morning. The Roper Technologies Conference Call will now begin. Today's call is being recorded. [Operator Instructions] I would now like to turn the call over to Zack Moxcey, Vice President, Investor Relations. Please go ahead.

Zack Moxcey

Good morning, and thank you all for joining us as we discuss the third quarter 2026 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.

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.

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.

And now please turn to Page 3. Today, we will discuss our results primarily on adjusted non-GAAP basis. For the third quarter, the difference between our GAAP results and adjusted results consists of the following items: amortization of acquisition-related intangible assets, transaction-related expenses associated with completed acquisitions, financial impacts associated with our minority investment in Indicor and lastly, cash collected on Outgo 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 this presentation on our website.

And now if you please turn to Page 4, I'll hand the call over to Neil. After our prepared remarks, we'll take questions from our telephone participants. Neil?

Neil Hunn

Thank you, Zack, and thanks to everyone for joining us this morning. You can see our agenda on Page 4. So let's get into it. Next slide, please. As we turn to Page 5, I want to highlight 3 takeaways. First, we delivered solid third quarter results, including 7% organic revenue growth and increased debt guidance for the year. Second, AI momentum continues to build, and we're pleased with the progress in the quarter.

Our businesses are bringing new products to market, and we're increasingly encouraged by the breadth of that progress and what these products will do for our customers. And third, we're active again on the acquisition front, completing 3 bolt-ons with our long-standing capital allocation discipline.

Starting with our results. Total revenue grew 8%. Organic revenue grew 7%, EBITDA grew 7% and debts grew 15%. At a high level, we had a strong quarter and despite out of a few of our larger businesses. We had another encouraging quarter for Deltek, although we continue to take a wait-and-see approach. We continue to see nice supply side recovery at DAT and unfortunately, Neptune continues to face subdued meter market conditions. Based on these results, we're also increasing our full year debt guidance. It was a solid quarter, and Jason will walk you through the financial details in a moment.

Turning to AI. Since our last call, our businesses have introduced a broad range of products that apply AI to the specialized tasks and workflows our customers manage every day. Let me share a few examples. First, Vertafore launched agents for submission processing, benefit plans and docket research and has now added 2 agents that automate policy processing for independent agencies. Since the beginning of the year, Vertafore developed and launched 12 workflow automation agents with many, many more in development.

At Aderant, 7 AI agents are now in early access to help law firms build more accurately and collect cash faster. The appeals agent helps draft billing appeals, the collection agent prioritizes overdue accounts and the time agent improves billing descriptions. These are practical uses of AI to reduce administrative work and help law firms run more efficiently. Another example is PowerPlan, where a new intelligent utilization product tackles a significant industry challenge more than $180 billion of capital investments awaiting detailed asset classification across the U.S. electric utilities.

Our purpose-built AI capability analyzes project and accounting data and recommends how costs should be allocated to individual assets. By accelerating that process, it helps utility recover their capital investments and earn return sooner, bringing forward revenue, earnings and cash flow and the productivity gains are substantial.

Customers are completing this work in minutes that previous of days or weeks. This is another powerful example of our focus across Roper, putting AI to work to help our customers grow, perform better and operate more efficiently to work by Rafi and the team of PowerPlan. At Strata, Financial Decision Intelligence helps health care finance teams identify what needs attention and understand what is driving performance.

In a few minutes, I'm going to double-click on this specifically. CentralReach report draft AI is designed to improve provider productivity and economics. Clinical reauthorization reports typically take 9 to 12 hours per learner, much of which is nonbillable. By automating data gathering and preparing a draft for clinician input and review, the product frees up substantial capacity for billable care, creating opportunities for providers to serve more learners, supporting revenue growth and improving operational efficiency with their existing clinical teams.

And Foundry introduced new AI capabilities for Visual fix, while sub-launch quirk management products that bring AI and automation into church operations. What is encouraging is both the breadth of this activity and how specific these products are to the markets we serve. They address work our customers know well, work that takes time, requires specialized knowledge and matters how their organizations operate.

We continue to see a substantial opportunity to help customers do work more efficiently and effectively and we're excited to partner with our customers from across the enterprise to realize the promise of AI. Our confidence that opportunity is ground in the same advantages that have always made our vertical market software businesses strong. Deep domain expertise embedded workflows, unique data assets, trusted customer relationships and scale distribution.

Our businesses know their customers' workflows and where AI can save time and improve decisions. each business develops products close to its customers, while our AI accelerator team helps speed development and share reasonable approaches across the portfolio. We're still early in commercialization and are starting to see adoption and monetization of our AI solutions. Importantly, our products change how customers do work, so adoption will develop over time.

As customers use our products, that experience will help our businesses refine the products and expand our capability remit. There is more work ahead, but we love the progress. We have new products reaching customers across a broad range of markets and a meaningful opportunity to build from here. And third, we continue to execute our disciplined capital allocation strategy and are encouraged that we're able to execute 3 bolt-on acquisitions during the quarter.

Specifically, we deployed $436 million towards bolt-on acquisitions for iPipeline, Procare and Strata. The common theme is expanding the capabilities and reach of our existing businesses adding complementary products, expertise and customer relationships in markets we know well. Each acquisition builds on our existing domain expertise in distribution, creating more opportunities to serve customers and drive long-term organic growth.

Our approach remains consistent, evaluate each opportunity carefully and deploy capital to receive attractive long-term returns. So a solid quarter with another impactful step on our AI journey.

With that, let me turn the call over to Jason to walk through our financial results, capital allocation activity and balance sheet position. Jason?

Jason Conley

Thanks, Neil, and good morning, everyone. Please turn with me to Slide 6 to review our financial performance. Our third quarter was highlighted by solid organic revenue growth and strong debt growth. We delivered revenue of $2.2 billion, which is up 8% over prior year, underpinned by organic growth of 7%. Digging deeper here, organic recurring revenue grew 8% across our 2 software segments. Our enterprise software gross retention remains very high, consistently in the mid-90s. And further, core enterprise bookings were up mid-single digits on a TTM basis. EBITDA of $864 million was up 7% and converted to EBITDA margin of 39.8%.

Core EBITDA margin was down 30 basis points, with our software segment in aggregate, slightly expanding, while our TEP segment was down 90 basis points. Our Neptune business delivered more mechanical meters in the quarter, which has a dilutive impact on revenue dollars and margin. Over a longer arc, we expect a continued migration to ultrasonic meters, which drives higher lifetime value for customers and Neptune.

DEPS of $5.89 was up 15% over prior year and above our guided range of $5.75 to $5.80 on better margin performance. Free cash flow at $805 million represents 37% of revenue, but was down 4% versus prior year, given we had timing of interest coupon payments on bonds issued in 2025, and we had a temporary lag in collections at our Ilumya business as they move through an ERP implementation.

For the year, we expect free cash flow to be 30% or more of revenue. And with our share repurchase activity, free cash flow per share is expected to grow in the double digits area.

Now if you turn with me to Slide 7, I'll walk through our financial position and capital deployment. Enabled by seasonally strong free cash flow and proceeds from the divestiture of Indicor's instrumentation business, we reduced our net debt to EBITDA by half turn within the quarter to land at 2.9x, all while simultaneously deploying $436 million towards acquisitions.

We finished with over $300 million of cash and $1.3 billion drawn on our $3.5 billion revolver. In the quarter, Indicor announced the sale of the remaining flow controls businesses to [ Zyla. ] We expect around $600 million of pretax proceeds sometime in Q4 as the transaction moves or regulatory clearance. Also in the quarter, we acquired 3 tuck-in businesses for iPipeline, Procare and Strata. Consistent with our acquisition criteria, these tuck-ins play directly into the strategy of increasing organic growth where we have the highest right to win.

All 3 tack-ons are growing faster than their acquiring platforms and coupled with synergies are expected to generate very attractive returns. Looking ahead, we now have more than $6 billion of capacity for capital deployment over the next 12 months given strong free cash flow, which will be enhanced by proceeds from Intercore sale of its flow control businesses. In summary, we remain well positioned to deploy capital towards the highest risk-adjusted return opportunities through a disciplined and analytical process and approach.

Neil, with that, I'll turn the call back over to you.

Neil Hunn

Thanks, Jason. As we turn to Page 9, let's review our Application Software segment. Revenue grew 7% in total with organic growth of 6%. EBITDA margins were 43.7%, up 30 basis points year-over-year. Looking across the segment, AI momentum continues to build with more of our businesses bringing new solutions to market. We also continue to see solid progress on cloud conversions and with the encouraging signals at Deltek.

Turning to the business highlights. Aderant had another excellent quarter. The team continued to expand its AI agent platform, as previously mentioned, with capabilities across rates, compliance and general ledger forecasting. Vertafore delivered growth across its agency, MGA and carrier businesses and launched 6 new AI agents, along with value-based credit model. This value-based pricing model is a key enabler of product-led growth as customers will utilize credits towards any current and future genetic solutions on the Velocity platform.

As we discussed last quarter, Vertafore strategy starts with high-volume labor-intensive work, embedding agents directly into the systems customers already use. The latest launches expand the work those agents can do and build on the foundation to establish with Velocity. Deltek had another encouraging quarter with continued strength in its private sector solutions.

Government contracting bookings have been solid over the last couple of quarters, although the broader market remains uncertainty. Against that backdrop, customers are showing increased interest in moving from perpetual license to a cloud-based offerings. As we discussed earlier, we also saw a new AI solution launches CentralReach with a [ Port ] draft AI and PowerPlan with intelligent unitization.

Finally, we completed 2 bolt-on acquisitions. Procare, we added Playground which helps child care centers manage enrollment, building and daily operations. Playground has exceptional product market fit and is highly complementary to the existing Procare product stack. We're excited about the growth opportunity from pairing that product with Procare's distribution and payments infrastructure.

And in Strata, we added c. myers, which helps banks and credit unions plans for growth and manage financial risk. This quarter, I spent a full day with the Strata team and finance leaders from 9 of our health system customers at a codesigned workshop for Financial Decision Intelligence, or FDI, Strata's new AI capability for health care finance.

FDI became generally available on September 1. FDI helps finance teams diagnose what is changing in their business and where to intervene to improve performance. For example, when joint replacement margins declined despite rising procedure volumes, FDI automatically explores the underlying drivers from payer mix and reimbursement to labor an implant costs to identify the source of margin pressure and focus management's response. FDI brings together 3 advantages we see from across the portfolio.

First, proprietary data and ontologies. Every health system has its own accounting procedures and definitions. Two hospitals can both report their margin on joint replacements and mean very different things. Strata spent decades mapping thousands of provider general ledgers into a specific common structure supported by health care cost accounting down to the individual patient encounter.

That makes meaningful benchmarking possible and gives FDI its foundation. Its calculations use the same governed measures customers are relying on to close their books with consistent repeal results that are traceable to the source. You cannot prompt your way to that. It is earned, one customer implementation at a time over years.

Second, turning insight into action. SCA operates inside the Strata platform, delivering answers in seconds and connecting those insights directly to the workflows where customers manage costs, improve margins and allocate resources. And third, real customer trust. Health care finance leaders need confidence in the tools they use to make consequential decisions.

Strata has earned that confidence through long customer relationships. Trust is critical to AI adoption and is very hard to replicate. Throughout the workshop, the customer request was simple and consistent productivity, less time symboling and reconciling analysis and more time working with operating leaders to improve performance.

One CFO described analysis it took several people hours to being nearly instant and FDI doing a more comprehensive and grounded analysis. This frees finance teams have focused on improving profitability and supporting higher quality care. Strata is selling FDI as an add-on subscription creating opportunity to grow recurring revenue. This is just the beginning for FDI and I'm excited about the road map ahead.

I'm proud of the Strata team's development speed and close customer collaboration and pleased with its partnership with our Roper AI accelerator team. Deep vertical data and domain expertise within each business paired with central AI capabilities, that is the model we're scaling across Roper.

Now turning to our outlook. We expect to see application software organic growth in the fourth quarter to be similar to that of the third quarter in the mid-single-digit plus range.

Please turn us to Page 10. Total revenue on our Network Software segment grew 8% with organic growth of 5%. The EBITDA margins were 52.4%, down 130 basis points year-over-year. Core margins were essentially flat to the prior year.

Turning to the business highlights. DAT was strong in the quarter and accelerated growth supported by expansion of its network, higher average revenue per user and continued adoption of its digital freight marketplace solutions. We continue to see encouraging signs in the freight market with the supply side recovery progressing.

As we've discussed, we would still like to see a stronger load volume before calling a broader recovery. We also completed the Origo bolt-on acquisition for iPipeline. Origo connects U.K. pension providers, investment platforms and financial advisers, automating well transfers, enabling secure data exchange. It complements pipeline solutions for selling and administering financial products, expands its presence in the U.K. wealth and retirement market.

Foundry also continued to grow and reached a record level of ARR. SoftWriters grew as adoption of its AI order entry solution continued to expand. As I shared last quarter, following my visit to LA Script, this solution addresses a meaningful customer need reducing the time spent on manual order entry so pharmacy teams can focus more of their capacity on clinical review, customer service and patient care.

Order entry is a starting point, and the team now has additional AI products and beta. Sub splash is tracking in line with our year 1 target and continue to advance the offerings, including enhanced workflow automation. Turning to our outlook. We expect network soft organic growth in the fourth quarter to be a much higher than Q3 and be in mid-single-digit plus range is based on strength of DAT and sub/becoming organic in Q4.

Now please turn to Page 11, and let's review our TEP segment. Total revenue and organic revenue both grew 10% in the quarter with exceptional growth in MDI and strong performance of [ Araton. ] EBITDA margins were 34.3%, down 90 basis points year-over-year. Starting with NDI, the business had another excellent quarter, delivering exceptional growth driven by continued demand for its precision measurement solutions.

Verathon also delivered a very solid quarter with strength across its BFlex, GlideScope and BladderScan product families. In addition, we're encouraged by the positive customer response to the Cflex single-use cystoscope launch which further expands Verathon's single-use portfolio.

Neptune had a softer Q3 following a better-than-expected first half. We now see second half revenue slightly down versus our prior outlook for modest growth. Q3 mix skewed towards macanimeters putting pressure on both revenue and margins, that mix is likely to continue into Q4. We view this as temporary as the long-term secular demand for static meters remains fully intact.

Finally, I want to recognize the team that CIVCO, FMI, [ anionic ] and RF IDeas each delivered strong execution and growth in the quarter. Turning to our TEP outlook. We expect organic growth in the fourth quarter to be in the mid-single-digit range.

With that, please turn with us to Page 13. On this page, let's walk through our increased full year adjusted DEPS guidance as well as our Q4 outlook. For the full year, we're raising and narrowing adjusted debt guidance to a range of $22.35 to $22.40. That represents a $0.15 increase at the midpoint from our prior guide and a $0.95 increase from our original guidance.

Our full year revenue expectations continue to have total revenue growth north of 8% and organic revenue growth in the 6% area. For Q4, we're establishing our DEPS guidance -- adjusted DEPS guidance of $5.95 to $6 and expect our tax rate to be in the 20% to 21% ZIP code.

Now please turn to Page 14, and then we'll open it up to your questions. We concluded with the same 3 points with which we started. First, we delivered solid third quarter results. Revenue grew 8%, organic reg grew 7%. EBITDA grew 7% and adjusted debt grew 15%. We're raising our full year earnings outlook again, reflecting the strength of our year-to-date performance.

Second, AI Momentum continues to build across Roper. Our business has launched several new Gentex solutions this quarter, expanding the work we can help our customers accomplish. The Strata workshop brought that opportunity to life. Our ability to deliver that kind of value rest on advantages built over many years, deep domain expertise, proprietary data, embedded workflows, customer trust. Those advantages are unique to our businesses and central to how our businesses are developing and deploying AI. We're still early, but the opportunity is substantial.

As our software takes on more valuable customer work, we have the opportunity to expand our addressable markets and earn a greater share of the value we create. Finally, capital allocation discipline remains unchanged. This quarter, we deployed $436 million into bolt-on acquisitions for iPipeline, Procare, Strata, adding capabilities to establish businesses in markets we know well.

As you heard Jason mentioned, our leverage now sits at 2.9x, and we have over $6 billion of capital deployment capacity over the next 12 months. With that, we're pleased to be putting capital back to work through M&A and we looked at 2027 with cautious optimism about compelling opportunities to deploy capital and accelerate long-term cash flow per share compounding.

As always, our standard for deploying capital remains the same. We will pursue the opportunities that offer the most attractive, durable and long-term cash flow per sure compounding whether through acquisitions or opportunistic share repurchases. Looking ahead, our focus is straightforward. Help our customers accomplish more, expand the work our businesses can do for them and invest the cash we generate where we can compound it most effectively. That is how we will build an even stronger Roper and deliver durable long-term growth and cash flow per share.

As we wrap up, I'd also like to welcome Abhijit Dubey to our Board of Directors who is the CEO and Chief AI Officer at NTT Data. We're delighted to have him join us and look forward to his perspective and contributions as we continue to execute our strategy. broadly deployed across our product sect in our enterprise and create long-term value for our shareholders.

So with that, let's open it up to your questions.

Operator

[Operator Instructions] Your first question is from Joe Giordano from TD Cowen.

分析师问答

Joseph Giordano

Maybe one, ion a specific company, but I'm sure you saw like the journal article the other day on autism care hours, particularly at PE-owned facilities. And I'm just -- maybe it's for the wrong reasons that these all are going up in some of these places. But I'm just curious like what that means for the rollout of stuff like CentralReach? Like it seems like almost a requirement to have technology like that if you're going to push the hours to the extent that this seems to be happening.

Neil Hunn

Well, I think the -- I appreciate the question. So I think the thing that is great about CentralReach Solutions is that it's all about appropriate care that is auditable that's to the standards and there was an article or there's a flare-up on this about a year ago when we looked across the CentralReach customer base, and it was like the most compliant set of customers. I think this is in Minnesota.

I forget where it was a year or so ago on some fraud claims. And it was the automation with CentralReach was the cleanest set of providers in the jurisdiction. So same thing here. I mean it's just a good audit sort of backbone when you sort of use the CentralReach and our automated systems here.

Joseph Giordano

If I think about the fourth quarter guidance, I'm just curious like, I know it's early for '27, but like are these exit rates fair starting point for us to think about moving forward? Like I'm curious like something like Neptune, right, like what that's a project-centric business, it's hard for us to kind of contextualize. So like if that's -- what are the prospects like kind of into next year?

Neil Hunn

I would say for million sort of set up. First, it's super important that we're just starting basically next week, the planning process. So this is -- we have to go through all the process with all 29 companies understand the puts and takes and all the nuances. So we always have to caveat that. But we'd say it's generally a constructive setup.

The recurring sort of revenue sort of exit run rate encouraging certainly, the velocity of the product development on the AI side is encouraging. Now caveat there is we -- and now in the commercialization phase, and we've got to drive the bookings and adoption and consumption that's sort of where we are in that. I'm sure there'll be questions on that in today's call.

And then if you sort of go segment by segment, Deltek has had a couple of good quarters. We -- as I mentioned, we're still in a wait and see. We want to see that continue. And then the other thing about Deltek is it the business is definitively turning more recurring, more SaaS. So we have to overlap the perpetual in that business.

And then just a reminder, CentralReach, we get a full year organic. And Network, certainly, what we're seeing at DAT is encouraging, but it's supply-driven. So it'd be more -- it would be healthier if it was demand-driven, but it's definitely a better set of coming into next year versus coming into this year at DAT. And then remind you sub-laisfully organic or organic for the full year.

And then TEP, Neptune is really the to be determined. And it has a wider range of outcomes. It had a wider range of outcomes this year that we anticipated that we're experiencing. And so that's the that's a to-be term and we'll have to really sort of spend a lot of time with the management team and understand the industry trends heading into next year. And then NDI, which has just been awesome for us this year will be more normalized next year. So but the setup is generally constructive.

Operator

Your next question is from Brent Thill from Jefferies.

Ria Naidu

This is Ria on for Brent. On Deltek last quarter, you mentioned that you wanted to see consistent conversion of the GovCon pipeline before being able to call this an inflection from OBBB just create so this quarter give you that. And I recognize it's early to maybe just to elaborate on that last question. How are you thinking about Deltek for 2027?

Jason Conley

Yes. I think it was a very good quarter for Deltek. What we are seeing, thankfully, is that more customers are starting to move to the cloud. So the bookings were quite strong. That doesn't obviously materialize in revenue immediately like perpetual.

It's an uncertain market. I think what Deltek is doing is -- and we mentioned this last quarter has announced the end of support for the cost point on-prem solution in the first quarter of '28. So that's creating a catalyst for them to go back to the base and start to do that migration. So despite a sort of uncertain market, they're sort of making growth happen. So we'll see 1 more quarter before into next year, but momentum is good there.

Ria Naidu

Got it. And then just curious on any incremental updates on the M environment -- M&A environment since last quarter, I think you mentioned $436 million deployed towards bolt-ons. But I think I'm wondering more so on the larger deal side of things we're seeing. So any color you're able to provide there would be helpful.

Neil Hunn

Yes. I would say, as I mentioned in the prepared remarks, we look to next year with sort of cautious optimism. Janet and her team continues to be very busy. I would say the pipeline tilts more proprietary and proactive outreaches than large-scale processes are very bespoke.

We just had a couple of weeks ago, our second annual CEO conference where we had 20 or so target CEOs together for a day, 1.5 days going through Roper and getting in other businesses. So that was well attended and very high-quality targets at that business at the summit. And it's -- we're cautiously optimistic.

I mean all the pressures that have been on the private equity firms for now going on 3 years continue to exist, LP credit, [ TPI ] pressure and sort of just coming to grips with where public companies are valued. It takes there's a little bit of -- there's a lot of [indiscernible] psychology in that, and that psychology appears to be coming into reality.

Operator

Your next question is from Brian Peterson from Raymond James.

Brian Peterson

Congrats on...

Neil Hunn

We've lost him.

Operator

Your next question is from Terrell Tillman from Truist Securities.

Terrell Tillman

On Vertafore, I care a lot about the insured tech space. I'm curious, how is the -- you called out the ARR growth, how is that faring versus the overall app software growth? And then it's still the first question, it is a multiparter the first one. This idea of the value-based credit, will that just take some time for that revenue stream to ramp? And then I had a follow-up on Deltek.

Neil Hunn

Yes. I'll take the value base credit one, Jason can take the first one.

Jason Conley

Yes, Terry. So they're in line with the segment in terms of the ARR growth.

Neil Hunn

And on the value-based credits, just to remind everybody, this is a -- these contracts are being structured as recurring. So there -- you're buying a fixed amount. Those are -- have to be consumed in a specific period of time.

As a result, the revenue recognition treatment is recurring not reoccurring. So it will look and feel like subscription revenue. Now obviously, it's very important for sort of, if you will, for deployed teams, the customer support implementation teams to make sure there's wide-scale adoption and utilization.

So you get the upgrades, the uplift, the additional sort of credit purchases and the renewals. And that's where we are right now. It's where I think the Velocity platform was GA within the last 60 or 90 days. I mean the full GA release has been very recent and we're in the early adoption phase.

Terrell Tillman

Good to know on recurring versus reoccurring. I'm glad to ask that. In terms of Deltek, I don't know if this is kind of a change in tone here, but I forgot where you last gave an update on the SaaS versus on-prem for Deltek, but it sounds like maybe things are picking up. So is kind of the real takeaway that the license part of total revenue could be like an incremental headwind in '27, but then you're going to have right behind it this kind of acceleration in recurring.

Just anything more you can share on kind of what you're talking about here because it sounds like maybe it is picking up steam. .

Jason Conley

Yes. And I think the -- as you can appreciate, Terrell, the timing of getting to the cloud is a little harder because it's a different value prop. So there's -- that's sort of what we're managing through. But certainly, it's our biggest maintenance base at Roper. And so the timing of getting that to the cloud is certainly would outstrip perpetual. It's just kind of how that phases over the next few years is a question. So -- but we do feel good about the momentum as we sit here today.

Operator

Your next question is from Brian Peterson from Raymond James.

Brian Peterson

Can you hear me? Hopefully. I'm coming out okay.

Neil Hunn

Yes.

Brian Peterson

All right. Sorry about that. I just wanted to ask on the appetite for AI and the adoption curve. I'm curious what products are you seeing that ramp a little bit faster than expected? And also, just on the customer appetite and the packaging, has that been any challenges? Or is it more about change management? Just kind of curious what the customer appetite is at this point?

Neil Hunn

So the -- I'll try to be comprehensive in my answer. if I don't, we can certainly take this offline as well. So let me just follow a bouncing ball of this. So it was really a year ago, the models and the capability of AI we really could do agenetic work. So the first question is, can we think of the ideas, the customer magic, if you will, to sort of put into the products there? That question is absolutely, without question, yes.

I think it's 18 of the -- 17 or 18 of the 21 software businesses have Agentic solutions in the market. So that's the first question. So first, can we think of the idea? The second, can we build them? That's -- so that's yes, and that's latish we have the majority -- the vast majority of the company is having these products in the market. Now the question is, can we drive early adoption, convert sort of, if you will, the beta customers to early adopt the answer to that has been yes.

Just go back to the remarks I said in the call about Strata and in FDI or last quarter with soft writers with pharmacy automation, the answer to that has been yes. Now it's about, can we get -- we need to get a broad scale across -- into the go-to-market teams across all these businesses and get it broadly distributed into our customer base and drive adoption.

So the customer interest has been very, very positive. It probably the most positive I've seen in my career and tech because we're literally delivering a but that does magic things. So that's a good part. Now what I think all companies and we're not alone in this is now you're facing budget cycles AI committees, all the safeguards and governance of data. And so we're just going through the go-to-market processes.

The early indications are positive. But we said this year, was going to be largely a year commercialization learnings and adaptation and again, progressing well, but it's still to be proven. And then I think the first part of your question, where some of the better use reaction, certainly CentralReach, certainly Aderant, certainly SoftWriters and pharmacy automation.

The early adopters at Vertafore, the early adopters at Strata, I mean there's really not -- I'm just going through the list in my mind, there's not really been a really a flat launch. I mean it's -- again, it's early days, and we will not back thousand across all these products across all the companies, but we haven't had any object out-of-the-box failures if you will.

Jason Conley

I think one of the interesting changes with AI, obviously, is when you launch a product, you can iterate -- continue to iterate. So you'll have the media customer feedback. So we've seen sometimes when something wasn't quite right, but then we've been able to iterate quickly based on feedback from the customer. So that's been helpful.

Brian Peterson

I appreciate the color, guys. Maybe just one added element there. As we're thinking about use of kind of frontier models versus some of the lighter weight models, how are you guys thinking about that balance? And does that change any views on how we're thinking about oriented margins over the near to medium term?

Neil Hunn

The punchline is our gross margins on these authentic tools are very positive sort of at our software gross margins. Our our philosophy on this at the moment, and this is obviously subject to change, and this is well informed by our leadership team and our central accelerator a accelerator team is we are 100% optimizing speed of development and deployment.

And so we're working with a lot of models that are on the front here. We don't need the most Frontier model. Let's be very clear about that. We don't need 5, 6 Astra or sale 5 or whatever current releases. But where you use sort of the cheaper for downscale models from the frontier, but the hardest is that come with them, you can just get moving much, much faster on the product side.

Now over time, every one of the products -- 100% of the products have the model routers. And so we can sort of -- we have multiple sort of models underneath each product in case when you call on tropical and if they're not answering the phone and with the latency you need, you have media flip over tier model. That architecture allows us to switch to models either into small language or open source models over time to manage the costs. But right now, we're optimizing for speed of [indiscernible].

Operator

Your next question is from Faith Brunner from William Blair. .

Faith Brunner

Congrats on the quarter. Maybe to pull on a prior thread you talked about -- you're seeing that encouraging engagement around adoption and now the focus is scaling that. So as you can work through the early process of that. What have you learned? What challenges are still there? And how are you kind of adjusting your go-to-market strategy to address these new challenges? And how do we get to a more clear monetization path, if that makes sense?

And then I have a quick follow-up.

Neil Hunn

Yes. I mean this is -- keep asking this question because -- and we'll just share our learnings as we go. I mean, obviously, we've had to figure out how to price all of these, and there's not a one pricing size fits all. It's different per company and per -- I would say, per product, but per company based on where their customers are in the existing economic relationships we have with our customers.

So we generally -- we think we figured that out across the board. Then it's how do you sell this in addition to versus in lieu of everything else that we're selling because there still is strong demand for, if you will, the legacy software stack, so how we make sure this is on top of an addition too. And then it's like, like I said, like in health care, in particular, acute care hospitals have sort of an AI governance overlay.

And that was a new learning from the strata team, for instance, it's like it's going -- it has a traditional sales cycle and then another piece of it. So it's not a barrier. It's just -- maybe it slows things down a little bit. And those are the sorts of things that we're learning right now.

Faith Brunner

That makes sense. And then maybe as we get more cautiously optimistic about M&A, how are you thinking about -- how does AI change your whole thought process on what makes an attractive target, right? So is there different growth considerations, different competitive considerations that you now have to consider? And how are you guys working through that?

Neil Hunn

For sure. So the first thing, and we've been doing this since -- over the last couple years the chat moment, if you will. We need to certainly make sure do the diligence that the end customer of the software company we're looking at is a viable long-term customer, the customer is vital.

So look at what we did last year, it was autism therapy, so people providing therapy to youth and then people going to church. So the end customer is highly viable and will not be sort of disintermediated by AI, if you will. Then it comes down to what we see in our portfolio. Do you have sort of the defensive modes and offensive modes to be a winner in the AI world.

So do you have you have sort of the right amount of data, the data rights, proprietary nature of the data? Do you have network effects? Do you have sort of configuration lock-in? So -- et cetera, et cetera. I will go through the full framework here. So then you have the preconditions to be successful in AI. That's where we are now.

Fast forward a year or 2 or 3 from now, that becomes, I think, that a servecomes even easier because you'll take -- you have taken targets will have taken those advantages and will either have proven or not proven they're winning in the AI will be able to develop products, bring them to the market, monetize the products to drive additional customer value. And so the companies that we're in now are in the earlier stages of that.

Operator

Your next question is from George Kurosawa from Citi.

George Michael Kurosawa

Couple quick ones for me. It looks like buyback activity was paused in the quarter. I think it's something you'd kind of message you were doing as you're preparing for the M&A environment to potentially warm up. Just maybe how you're thinking about buybacks going forward?

Jason Conley

Yes. Thanks, George. I think it's the same as we've always talked about, which is the relative sort of risk-adjusted returns for us. And for us, obviously, the buyback is attractive, but the option value on the M&A, given all the dynamics we've talked about in that everybody really knows about is worth the deleverage and prepare our balance sheet for that event.

Now if we fast forward a couple of quarters from now, we still have that option available to us. We still have quite a bit on the authorization. So it's still a viable alternative if M&A markets don't cooperate.

George Michael Kurosawa

Okay. Great color. And you referenced on the free cash flow side, some impact from collections timing ERP implementation at Illumia. Maybe if you could quantify the magnitude of that and then ERP implementations have a habit of maybe extending beyond initial time lines. Is that is that are those issues now addressed? Or is there any potential that could extend into Q4?

Jason Conley

Yes. Certainly, a lot of focus by the team there, and I know they're going to do a good job. They're just powering through some order to cash challenges. They actually went live in January, but the big renewal season is in summer and fall.

So we expect to see some improvement in the fourth quarter. Some of that might spill into the first quarter of next year. But when you kind of take a step back for the year, we expect free cash flow per share to be in line with DEPS growth, maybe even a little bit better. We'll see how it plays out. So that should all sort of normalize as we close out the year.

Operator

Your next question is from Daniel Jester from BMO Capital Markets.

Daniel Jester

So I wanted to make sure I caught this right, but I think you said earlier in the call that a majority of your businesses now have an AI solution in the market. And I guess given the sort of urgency and the sort of moment in time that we're in and the past comments you've made about the ability to sort of speed your development I guess what's the time line to get to 100% in terms of your businesses and software having an AI tool? And are there any sort of commonalities in terms of limiting factors to get there?

Neil Hunn

Yes. I appreciate the question. So when you look at the portfolio, sort of there's basically Three companies that don't have an AI tool. One is not a true software company. It's a group purchasing organization. So they don't really have like an agentic software front that reports MHAs, reported in the network segment. They're using all sorts of AI inside the business, drive productivity. But that 1 set that aside.

Then you have 2 businesses that are in clinical diagnostics, DI and CliniSys. They have early AI tools, but this is a -- what laboratory software does and this remit, is this a nearly a fully solved technical problem. There's not like this is how you run laboratories and route samples to analyze and take results off analyzes and report that.

And so it's basically an already fully automated process. And so there's not a large genetic remit to do here. Now they're adding some AI tools around it, but it's not like these processes are already fully automated. So essentially a long-winded way of saying the companies that have AI remit, 100% of those companies have AI tools today.

Daniel Jester

That's super helpful. And then I guess I'll pull on the AI monetization thread as well for my follow-up. I guess as you think about next year I guess, how diverse across the portfolio? Do you think that the monetization will be?

Do you think that some parts of the business are going to have one type of way of monetization and other parts of the business, a very different one. I guess, maybe at an enterprise level, how should we be thinking about monetization? And maybe any color you can provide about the diversity at which -- which parts of the business maybe come sooner and which parts maybe will be coming a little bit later.

Neil Hunn

Happy to do it. All right. So on the diversity of pricing models and AI, it comes in 3 get flavors. The first flavor we already talked about on a call like this tokenized fixed budget use it over a period of time where you take it as recurring that appears to be the predominant the majority of what Vertafore, Aderant, CentralReach. I mean that's the predominant chunk -- maybe less CentralReach, but that's the predominant way that our companies are going to monetize early.

So it's a concession-based thing, but you buy a fixed amount of consultant user to lose it as you consume that you buy additional sort of chunks. There's -- the other model is straight consumption. It's a minority of what we're doing, but we have straight consumption. It's where our customers already buy from us largely on a consumption basis. So in pharmacy automation, a lot of the tools already paid on a per script basis.

So we monetize on a per script basis with AI tools there. Or at DAT, we monetize what we do at Convoy and DAT on a per-load basis, so a there will be on a per-load basis. The third way is going to -- and this is like, for example, a Deltek where their software remit is ERP for government contractors. And so how do you automate like the accounting close is one of the products that -- one of the product features they're delivering right now. That's going to be basically delivered in the cloud as an add-on.

So you're going to see, if you will, accelerate its move from ground to cloud at a higher price point once you get to the cloud. So those are the 3 models. To your point about which ones will show up first. I think it will sort of go in sort of the first 2 models. The consumption if you will, sorry, the subscription model where you have consumption embedded in that and the straight consumption. Those will be the ones that show up earlier and then the lifted shift to the cloud will take a little bit more time. I would anticipate things.

Operator

Your next question is from Ken Wong from Oppenheimer.

Hoi-Fung Wong

I just wanted to circle back to TEP growth. Decel from the double digits that you guys reported this quarter. How much of that is divestiture, how much of that, as you guys, I think, had called out some Neptune softness. Any color you can provide there?

Jason Conley

Yes. So second half, we said it was going to be high single digits and it being a little bit higher in the third quarter. And so that did play out. And so we've got mid-singles for the fourth. I would just say broadly, it has been the Neptune softness, and that's -- it's really a market sort of dynamic.

And then specifically for us, volume base is okay, but we actually had more mechanical meters has a lower ASP at lower margins. So that was sort of the driver of the second half, and we didn't have -- our divestiture -- we did divest the industrial businesses a few years ago, and we're getting the proceeds from that, but there's no sort of near-term divestiture activity in that segment.

Hoi-Fung Wong

Okay. Perfect. And then on the EPS guide or DEPS guide, any impact from the bolt-ons that you guys just picked up?

Jason Conley

No. I mean, the impact for us will be more next year. The current quarter, it's actually a little bit dilutive as we've got some synergies that we'll roll through next year. But it's actually -- for the fourth quarter, it's a little bit dilutive a few cases.

Operator

Your next question is from Josh Tilton from Wolfe Research.

Joshua Tilton

Maybe for my first question, I feel like there's been like a lot of hypotheticals and what is and how do we see the path of adoption play out around AI and just the monetization framework. And I understand why everyone is asking the questions. But unless I misunderstood, I thought you guys called out in the prepared remarks that you are seeing monetization from AI today.

So I'd kind of like to focus on that comment. Like where are you seeing the monetization today? How are you seeing it? And if we're already seeing it in some places today, like is that where we should expect it to be the most meaningful next year?

Neil Hunn

Yes. I appreciate the question, Josh. So yes, I mean, this is -- I think the question is more on the -- how do we scale this. And obviously, this is new just for us, but for the economy at large. And so we're all trying to figure that out. where we're seeing early AI sort of monetization is where the product stacks were first to market, so CentralReach reach. For instance, CentralReach, I'll have Jason have to make sure I sort of keep me honest here, but 75% of their bookings are either AI-related, AI-specific or AI-related bookings in that range, for instance. .

Aderant, the number is like 25% of their bookings are the ARAI related. The anticipation for what's now back behind sort of the -- they were, I don't know, a year ahead of the portfolio. Now the portfolio is catching up. 25% to 75% is a wide range, but it would be nice to see that as we sort of scale across a dozen of our companies, but it's too early to sort of say it's going to be 10%, 25%, 50% of bookings over time are going to be AI related.

Fast forward 3 or 4 years, like everything's AI, like this whole thing we say what's not AI. It's going to be indistinguishable. So I think we're in a moment in time in that regard.

Joshua Tilton

Makes sense. And then maybe just my follow-up. Neptune sales -- and again, I could be a little bit wrong here, but it sounds like Second half is a little bit less favorable. There's a mix dynamic. We believe in this view of moving towards the ultrasonic meters over time, better pricing. I understand all that.

I guess, is there -- was the mix dynamic unexpected? And I ask it in the sense that is there a correction that can be made into next year where we can go back to maybe a better-than-expected Neptune outcome after we saw what happened this quarter. We stepped back, we make an adjustment, and we go back to being a little bit better? Or do you expect kind of whatever you're seeing pressure the business today sort of last beyond this quarter, maybe into next year?

Jason Conley

Yes, I appreciate the question. I mean I think it's the static pipeline from Neptune customer by customer is strong. So there is definitely a market demand for it. It's just -- we just had pushes on the pipeline. So we do think coming into next year, it's going to be more favorable for us. Timing is going to be uncertain in terms of kind of the quarterly cadence. But yes, that secular demand is there. So we feel good about that.

Operator

Your next question is from Brett Linzey from Mizuho.

Unknown Analyst

Ryan on here for Brett. Yes, I wanted to just ask it a little bit more color on Neptune. Maybe just help us understand the market conditions the main puts and takes that you guys are watching and then maybe kind of how you expect the mix dynamic to play out over the coming quarters?

Neil Hunn

I'll take a 2 at this. So just to remind everybody, the macro, what's happening in the water meter -- the North American water meter end market, just to set that first, then we'll get into like a specific quarter and our best look at the outlook. So what's happening is during come especially in the Northeast or the water meters and people's statements, they cannot be sort of maintained or replaced.

So then post COVID, '22, '23, there is a surge, there's a catch-up. Now on the backside of this, last year -- this year, for sure, we're having to sort of comp over that surge. And so there's a -- in the industry, there are fewer meters that are being installed last year and this year versus the years before. You're kind of going from peak the trough. This isn't like a macroeconomic cycle, it's still like a COVID super cycle dynamic that the industry at large is digesting. We talked about that in the past, but just sort of level set you and the team and everybody on that.

So inside -- so we're just sort of on the -- we believe we're on the tail end of sort of that sort of reset getting into more a normalized sort of meter market. Now what's also happened inside of that is this mix from mechanical to static. Static meters is a better technology. They last longer. They measure -- they perfectly measure the flow of water over the life of the meter where mechanical degrades still over time. So it's a better value proposition to the customer.

That's why the market is generally going there. because it's higher value, it carries a higher price. And so you've had this positive mix that's happened in the industry, in particular, at Neptune, that helped offset some of the meter declines, macro meter declines in the market. In the quarter, as Jason just said, there are projects, large projects, the static meter projects that pushed, which had this -- the total number of meters that Neptune sold in the quarter was actually pretty close to the expectation, but it was a dramatic mix towards mechanical, which had both a revenue application and a margin implication.

And as you just said, that likely sort of has a chance probability that normalizes into next year, but that's sort of the setup of the mix dynamic, both through the macro and the mix dynamic. We believe over the long art. I'm not -- this is not a call for next year. It's like this is -- has been over our ownership like a mid-single-digit growth business. and we think those dynamics are intact over the longer.

Operator

This concludes our question-and-answer session. I will now hand the call back to Zack Moxcey for any closing remarks.

Zack Moxcey

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

Operator

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