管理層發言
Ladies and gentlemen, thank you for standing by. My name is Krista, and I will be your conference operator today. At this time, I would like to welcome everyone to the ServiceNow Second Quarter 2026 Earnings Conference Call. We will now turn the conference over to Darren Yip, Senior Vice President, Investor Relations and Market Insights. Darren, please go ahead.
Good afternoon, and thank you for joining ServiceNow's Second Quarter 2026 Earnings Conference Call. Joining me are Bill McDermott, our Chairman and Chief Executive Officer; Gina Mastantuono, our President and Chief Financial Officer; and Amit Zavery, President, Chief Product Officer and Chief Operating Officer. During today's call, we will review our second quarter results and discuss our guidance for the third quarter and full year 2026. Before we get started, we want to emphasize that the information discussed on this call, including our guidance, is based on information as of today and contains forward-looking statements that involve risks, uncertainties and assumptions. We undertake no duty or obligation to update such statements as a result of new information or future events. Please refer to today's earnings press release and our SEC filings, including our most recent 10-Q and 10-K for factors that may cause actual results to differ materially from our forward-looking statements. We'd also like to point out that we present non-GAAP measures in addition to and not as a substitute for financial measures calculated in accordance with GAAP. Unless otherwise noted, all financial measures and related growth rates we discuss today are non-GAAP except for revenues, remaining performance obligations, or RPO, current RPO and cash and investments. To see the reconciliation between these non-GAAP and GAAP measures, please refer to the earnings press release and investor presentation, which are both posted on our website at investors.servicenow.com. A replay of today's call will also be posted on our website. With that, I'll turn the call over to Bill.
Thank you very much, Darren, and thanks, everybody, for joining today's call. I looked at the transcript from an earnings call 7 years ago. Back then, we said ServiceNow would be the defining enterprise software company of the 21st century. I thought I'd give you a report out since then. We beat expectations in every quarterly report, expanded the profitability and free cash flow of the company, quintupled our total addressable market, accelerated six of our own unicorns to $1 billion or multibillion-dollar businesses, processed billions of workflows and trillions of transactions, grew the partner ecosystem globally, architected the most complete AI control tower for the enterprise, maintained our best-in-class renewal rate, increased our brand value, offered our customers deep enterprise context, unlimited choice and differentiated capabilities. And of course, we set a course to $32 billion in revenue in 2030 operating at the rule of 60 and beyond. So today, we're adding a stunning Q2 print to this track record. Subscription revenue growth was 23% in constant currency, 1.5 points above the high end of our guidance. Current RPO growth was 21.5% in constant currency, more than 2 points above our guidance. Operating margin was 29.5%, 3 points above our guidance. We had 123 deals greater than $1 million in net new ACV, up 40% year-over-year. ServiceNow AI ACV exceeded expectations again, surpassing $1 billion, keeping us on track to beat our target of $1.5 billion ACV by the end of 2026. We're feeling real good about it. What does it all mean? We are who we said we were. The path to value isn't just making AI. It's deploying AI securely across the enterprise. IDC forecast spending on AI software is going to grow 53% this year, 17% faster than AI hardware. Whichever chip wins, whichever lab wins, whichever price-per-token regime prevails, the enterprise needs one governed layer of record, full workflow and ServiceNow offers needed certainty in an uncertain stack. Our platform is optionality on all AI outcomes, not a bet on anyone. We're in the bull's eye of AI, cybersecurity, workflow orchestration, integration and automation. That's why we're growing fast — it's why we're only just getting started. We are who we said we were. Over the past several months, we've addressed a number of really good questions during meetings with investors. Today, I'd like to give every investor the professional courtesy of those exchanges. One question we regularly get is ServiceNow becoming a cybersecurity company. Here's the answer. ServiceNow already was a $1 billion-plus cybersecurity business. Today, our security and risk business is the fastest-growing of the top 10 cyber companies in the enterprise. I'll make it very clear. We now have a 10-figure cybersecurity business that's growing faster than all the other top cybersecurity companies. We're building the world's most integrated end-to-end security platform across cyber, risk and compliance, agentic incident response, exposure management, identity and access security, cyber-physical security and continuous vulnerability detection. The security stack sits on top of our ITSM and ITOM core infrastructure. So think about it: AI control tower plus Armis plus Vesa. Our customers want every AI enterprise to be visible, governed and secured in one command center. They don't want any blind spots. There are 2.2 billion agents entering the enterprise globally. That's 2.2 billion new identities, roughly a quarter of today's human population. Vesa maps access across human, machine and AI identities. We'll have 40 billion connected devices in the world in the next four years. Armis already tracks 7 billion of those devices in real time. Many customers, especially in the public sector, seek out Armis because government policies require state-of-the-art visibility. The attack surface — every ungoverned asset and identity — multiplies the blast radius. When you integrate awareness and identity with the actionability of the ServiceNow platform, you have a complete 360-degree capability to secure the enterprise. It's highly telling that the AI control tower is already gaining traction with our partner ecosystem. In fact, one global services firm is leveraging this portfolio to help their clients rapidly triage their cybersecurity activity. They're hunting down problems with ServiceNow and others are rapidly following suit. So once again, this is the eighth-largest cybersecurity business in the enterprise and the fastest growing, and we are just getting started. Another question we get is when will customer deployment of agentic AI mark an inflection point for ServiceNow's growth. Here's the answer. It already has. The percentage of renewal customers purchasing an agentic AI for the first time doubled quarter-over-quarter and year-over-year. So customers that weren't already on the AI journey are signing up fast. Most customers are now completely allergic to anything that looks like a project. They only want deterministic. ServiceNow only does deterministic. That's a big reason why customers with agentic AI in production have grown 9x over the last nine months. So here's a few examples. The Department of the Air Force is expanding its use of the ServiceNow AI platform. The deployment will unify IT operations and enterprise visibility. Experian is using ServiceNow to automate intelligence at scale. We will also integrate our platforms, embedding Experian data and decisioning into existing ServiceNow workflows in a five-year deal and many deals are going longer; you should see that in the current RPO. Maybank, Malaysia's largest bank, will leverage ServiceNow to establish a resilient operation center to fortify security and resilience. The U.S. federal government's largest IT contractors and agencies are now consolidating asset discovery and security response on ServiceNow. One agency is using Armis to move from asset-blind incident response to comprehensive threat hunting. Hitachi will standardize enterprise asset management across its global businesses on ServiceNow. Additionally, we're partnering with them to advance the Hitachi Intelligent Infrastructure monitoring solution. The city of Raleigh became the first local government to deploy ServiceNow's Level 1 AI specialists in production with no in-house AI engineering bench behind it. That's right. It's on its own. The city is moving to become fully autonomous, one ticket category at a time. Many enterprise customers are at various stages of the same journey. They are progressing from supervised to autonomous operations. A large global consumer goods company partnered with ServiceNow to launch its first agentic AI use case in ITSM to triage workflows. The customer rapidly deployed a repeatable governance-first blueprint for future agentic use cases. A leading global food and beverage company found itself with AI agents proliferating across five platforms simultaneously with no unified way to govern them. Their CTO has a mandate: no agent goes live without our clear governance, risk, value validation and observability. AI control tower went live this spring. They can now see in real time who was building AI and whether it had been approved; the value for them was immediate. It's worth highlighting IT operations management continues to be a source of strength for ServiceNow extending its outperformance in Q2, with attach rates to ITSM continuing to rise year-over-year. Our CMDB gives customers a trusted system of record for their infrastructure, applications, services and dependencies. ServiceNow delivers the intelligent layer and the infrastructure it runs on. We also get asked, is ServiceNow gaining traction in the CRM marketplace. Here's the answer. The market participants in the enterprise are very good companies. They're not going away. Having said that, we're doing very well. Already, we're a $2 billion ACV business. CRM ACV growth accelerated again on a year-over-year and quarter-over-quarter basis. Sales CRM average deal size doubled year-over-year. We're on track to execute over 2 billion service CRM cases this year. In addition, partners are increasingly positioning ServiceNow as an operational CRM platform, opening executive conversations across the C-suite. For example, in Q2, a partner closed a full front office replacement of a major CRM deployment in just two months. A leading North American automotive marketplace outgrew its legacy CPQ provider; it couldn't handle the volume or complexity of the business after five vendors failed to deliver. ServiceNow will help their sales teams achieve the speed and accuracy to operate at scale. An American software leader selected ServiceNow to modernize a highly customized quoting environment; our ability to connect CPQ with broader workflows, customer data and service operations was a key differentiator in the win. A leading North American telecom infrastructure provider selected us to unify commercial and field operations on a single platform. ServiceNow will scale volume 3x with no added headcount and cut repricing from weeks to hours. A regional financial institution consolidated its loan origination workflow onto ServiceNow's unified CRM platform replacing legacy point solutions. Here's the big one: voice is the next frontier, and we're winning it. We just proved something extraordinary. A large airline has gone all in, running their customer service voice calls on ServiceNow's voice AI CRM agents. This is live in production and handling 5 million annual voice calls in year one alone. The results speak for themselves. Customer satisfaction is off the charts. We also get asked about ServiceNow's role in the HR stack. Here's the answer. Employee Works is the strategic entry point for enterprise-wide employee experiences. This combines Moveworks conversational AI with ServiceNow workflows to create a single place to search, self-serve and take action across HR, IT and all workplace services. As Fortune 500 customers adopt Employee Works, we're seeing it pull through broader HR and employee experience opportunities, a similar pattern to what's happening in the CRM business. We get asked about the industry dynamics in enterprise AI. Here's the answer. Innovating for the enterprise, as you may have noticed, is pretty easy to talk about and hard to deliver. One of the reasons Knowledge 2026 was the biggest we ever hosted is because customers are desperate to make real progress with a proven innovator they can trust. Sessions on ITSM, ITOM and AI control tower were all oversubscribed because the influx of AI creates more activity for IT, not less. Customers demoed our latest product releases, including new AI specialists for IT, CRM, employee service and security. They learned about ServiceNow Action Fabric. This enables any agent, Claude, Copilot or homegrown model to tap directly into secure, governed enterprise actions headlessly. With so much of a data mess, they wanted to understand ServiceNow's context engine and autonomous data analytics which resolved data fragmentation across systems. With interest in ServiceNow rising, our innovation velocity is accelerating to meet the opportunity. For example, we have a big announcement coming very shortly: a business model evolution that expands our TAM with AI-native products. This new offering will be a conversational service desk experience: no tickets, and AI-coded automation. In a new product-led motion, we're targeting the Fortune 500,000, many of whom we expect to replace their legacy offerings and bypass startups, stop wasted time and cost. We already have several customers in beta and soon will be GA. This is the first of many net-new AI-native products we'll be bringing to market in the weeks to come. Our dynamic engineering team is on fire. Another area of real progress for customers is the expansion of our strategic partnerships. Deeper integration is ServiceNow AI Control Tower and Microsoft Agent 365, extended agentic AI governance from desktops to data centers with Project Arc and new agents secured by the NVIDIA Open Runtime and governed by ServiceNow AI Control Tower. Enhanced collaboration between ServiceNow's AI-native FTEs and industry-led Accenture FTEs is accelerating outcomes. There are too many to cover here. The momentum is everywhere. Here's a rapid-fire round for you. Are we worried about seat compression? Not at all. Our addressable user base is growing and 50% of our net new business is already non-seat-based. We keep seat-based pricing because customers prefer it for predictability, particularly now where a lot of pricing out there has been less predictable. Our customer going to build their own agent? I've yet to meet a customer who would even consider it. The best tech leaders know it will cost 5x to 10x to build an agent versus run one on ServiceNow. Is the government business weaker? It's never been stronger. National governments, regional governments, local governments, they're all expanding their ServiceNow deployments. There are just a few questions that I hope you found useful in answering for you over these few moments we get to spend together. For now, I'll answer the final question: why will ServiceNow get incremental share of wallet in the enterprise? I'll explain it as plainly as I can. There are many good companies in the enterprise today, and I do have respect for them all. Some have built very strong businesses over the years, others are new participants. They're good companies. The established companies aren't going anywhere, and I expect them to do very well based on the years of business logic and context in their platforms. The new companies are growing fast, and they'll be disruptive to some legacy point solutions and likely complementary to the bigger platforms. There are two major factors that should cement conviction in ServiceNow. One, if you were to assess all the players out there, which of them include a system of record, deep enterprise context, fully integrated governance and auditability, a proven global distribution channel, and a flexible pricing model that includes predictable consumption and outcome-based options? I'll save you the trouble. Only ServiceNow's completeness of vision checks all of those boxes. The second factor is pure market fit. Which of the platforms in today's enterprise was designed from the ground up to integrate to all of these players? Again, the only answer is ServiceNow. Risk at scale is why CEOs are losing sleep right now. They read all these threat headlines. They see all these platforms and token invoices and they don't know what to do. This is the gap. ServiceNow is the bridge. We're in the control business: one platform, one system of action — any cloud, any agent, any workflow, any model, governed, secured and accountable. This is the AI control tower enterprise leaders now know is mission-critical, enterprise execution with cybersecurity at the core of the platform and cybersecurity growing real fast. ServiceNow is the rules and rails of the enterprise; we're in command of the agentic enterprise from workflow to cyber. Our position is real, it is durable, it is compounding. This is the foundation for a re-rating of ServiceNow. Let's get it started today. We said defining enterprise software company in the 21st century. We are who we said we were. Thank you for your time today, for your interest in ServiceNow and for your enduring support. We'll never take it for granted. With that, I'll hand things over to our President and CFO, Gina Mastantuono. Gina, over to you.
Thank you, Bill. Q2 was an outstanding quarter that highlights ServiceNow's broad-based demand, strong execution and operating leverage. Once again, we beat the high end of our guidance range across every top line and profitability metric. AI net new ACV growth continues to outpace expectations. Our AI Control Tower is supercharging our security and risk business and ITOM is seeing strong demand tailwinds for the CMDB to serve as an essential governance and data foundation. Simply put, the momentum across the portfolio continues to build and we remain on track to deliver another exceptional year. Q2 subscription revenues were $3.877 billion, growing 23% year-over-year in constant currency and 150 basis points above the high end of our guidance. RPO ended the quarter at approximately $29 billion, representing 22% year-over-year constant currency growth with an increase in average customer contract duration. Current RPO was $13.2 billion, representing 21.5% year-over-year constant currency growth. That's a 200 basis point beat versus our guidance. Across our workflows, we saw widespread demand. Technology workflows had 50 deals over $1 million, including nine deals over $5 million. The combination of Armis and Vesa has had a strong pull effect on our core. ITSM was in 15 of our top 20 deals. ICON had an outstanding quarter, appearing in 18 of our top 20 deals with 14 deals over $1 million. Our security and risk solutions were in 16 of our top 20 deals, also with 24 deals over $1 million. CRM and industry workflows were in 16 of our top 20 deals with 15 over $1 million driven by sustained momentum in CPQ and sales and order management. Core business workflows had 12 deals in the top 20 with a blockbuster 24 deals over $1 million, driven by strong demand for Employee Works. Creative workflows had 18 deals in the top 20 with 14 over $1 million. From an industry perspective, business and consumer services led the way with net new ACV growing over 6x year-over-year. Education posted impressive growth, surpassing 125%, followed by telecom and media at nearly 40% year-over-year. Manufacturing also delivered strong growth in the quarter. Our renewal rate was a best-in-class 98% in Q2, underscoring the durability of our customer relationships. We ended the quarter with 658 customers generating over $5 million in ACV with 32 more customers crossing the $20 million threshold since last year. As customers scale, they are consolidating more on the ServiceNow AI platform, demonstrated by 18 of our top 20 deals including eight or more products. ServiceNow AI continued to outperform expectations in Q2 with ACV crossing over $1 billion and net new ACV growth accelerating sequentially, growing over 40% quarter-over-quarter. Deals including five or more ServiceNow AI products grew 5.5x year-over-year which drove a tripling of million-dollar-plus deals. In addition, the number of customers with agentic AI in production has increased 9x over the last nine months — a leading indicator for the future consumption opportunity. While still early, we're already tracking ahead of our target for AI to reach 30% of ACV by 2030. As Bill noted, our new AI-native SKUs are making our agentic solutions more accessible across the customer base, as evidenced by deal volume among first-time ServiceNow agentic AI buyers growing over 45% year-over-year. What's more, upgrades to our new Pro Plus SKUs are driving price uplift in line with the 20% to 30% framework that we laid out at our Financial Analyst Day. Employee Works, the AI front door for the enterprise workforce, continued to build strong momentum with deal volume growing over 150% quarter-over-quarter — another good example of our acquisitions amplifying the core. The robust attach rates of AI across the platform are also driving our data and analytics business. RaptorDB Pro deal volume grew 80% year-over-year again in Q2, and Workflow Data Fabric was in 17 of our top 20 deals. Turning to profitability: non-GAAP operating margin was 29.5%, 300 basis points above our guidance, driven by the revenue outperformance and timing of spend, primarily in marketing. Our free cash flow margin was 16%. Together, these results continue to demonstrate our ability to drive a strong balance of world-class growth, profitability and shareholder value. Moving to our outlook. For 2026, we are raising our subscription revenues by $15 million at the midpoint from $15.755 billion to $15.770 billion, representing 21% year-over-year growth on a constant currency basis. We expect subscription gross margin of 81%, reflecting more customers utilizing our hyperscaler partnerships as well as accelerating AI adoption. We expect operating margin of 31.5% and free cash flow margin of 35% and GAAP diluted weighted average outstanding shares of 1.04 billion. For Q3, we expect subscription revenues between $3.975 billion and $3.980 billion, representing 20% year-over-year growth on a constant currency basis. We expect current RPO growth of 20% on a constant currency basis. We expect operating margin of 31% and we expect 1.05 billion GAAP diluted weighted average outstanding shares for the quarter. In closing, Q2 was another outstanding quarter in an environment where most enterprises are still searching for AI's ROI; ServiceNow is the platform delivering it. In my conversations with CFOs, the question has become simple: where is the return? They've watched budgets burn on pilots that demo beautifully and never touch the P&L. ServiceNow doesn't just deploy AI; our AI Control Tower provides a single view to discover, govern and secure every AI system and tie every agent to a real ROI. AI that only advises is a cost; AI that completes the work is a return. On our platform, agents don't just reason — they take action. And because we have the contextual data, every action is anchored in the live governed reality of the enterprise — turning probabilistic AI into deterministic, reliable outcomes. That's how AI moves from an operating expense to operating leverage. Customers aren't paying us for tokens. They're paying for resolutions. That's why enterprises are choosing ServiceNow to convert AI ambition into measurable ROI. Finally, to our teams around the world, thank you. Your execution is why the world works with ServiceNow and why we are more confident than ever in the road ahead. With that, I'll open it up for Q&A.
分析師問答
This is for Bill or Amit, if he's on as well. I wanted to ask you about your Level 1 ITSM product cycle. I know you talked about it at Knowledge and went generally available with pilot customers. My question to you is, what are your observations over the last three months? What's working with that product cycle, what's the feedback in terms of how you can make it better. And Gina, you already have a couple of general AI data points on monetization. For this product cycle in broader monetization, any nuggets you can share with us on how much customers increase spending with ServiceNow when they go through the Level 1 automation price?
Gabriela, thank you for the question. The Level 1 ITSM product cycle has been very accelerated for us. The thing we're solving for, which we talked about at Knowledge, is to really have human-equivalent AI specialists, which can take on the full task end-to-end without any customer having to deal with individual pieces of technology, AI agents, orchestration or the reasoning. We take out all of that work for them and customers really benefit from the outcome. That is what's resonating with customers today. We have deployed Level 1 support AI specialists as well as 20 other AI specialists inside ServiceNow, and we're starting to do that with our customers. Today, 40-plus customers are using it already and getting a lot of good benefits. We're talking about closing down close to 80% to 85% of the service requests without having any human interaction. This is end-to-end service request completion, not just giving you the information but actioning on those requests as well, and that reduces the time. For example, some of these requests might take two days for humans to reach a result; we're doing that in 20 minutes. So the productivity gains for employees are very high and very meaningful because now they're getting the issues resolved and we're reducing the volume of requests humans have to deal with. They can now focus on something more complex instead of having to deal with all these volume requests. So this is a game changer. This is first in the industry, and we've seen huge success already; you should expect a lot more continuous evolution in this space. Customer traction is excellent so far; usage goes up considerably because you burn down a lot of analysis overhead as you are doing full end-to-end task resolution here, and that's really where the opportunity for us exists. It's also labor arbitrage. We also get to monetize the labor cost savings customers would otherwise have to pay employees. They can now do this in a much cheaper way so we can take away labor cost as well. There are a lot of great examples happening here and great progress so far.
I'll just add, Gabriela, to the monetization question because it's very important. I'll reiterate we crossed $1 billion in AI ACV in the quarter, well on our way to $1.5 billion. I also noted in my script that while early, we're already tracking ahead of our target for AI to reach 30% of ACV in 2030. These Level 1 autonomous agents, if you think about the complex tasks they're able to execute, really drive more and more usage as we think about the compounding effect of consumption. This is a great example of how we will reach our 30% target; we're tracking very well and we're really excited about what we're seeing.
On the Q2 upside, you mentioned the federal outperformance. Does that strength at all prove to be a leading indicator for a bigger bounce back there versus last year in terms of spend? How should we think about the seasonality of that segment? And maybe you can also frame how you're approaching that from a Q3 guidance perspective, if there's any prudence given duration or upfront impacts we should consider in parsing Q2 and Q3?
Thanks, Michael. Clearly, a strong Q2 in federal is good news in general. As we move into the big Q3 quarter for federal, it's definitely a good indicator. We called it out specifically because we did have a little bit more on-prem in Q2, which is purely timing. But I want to be clear that the beat was not all just on-prem. We had really strong net new ACV in the quarter as well, which allowed us to raise the full year guide. So federal business is strong as we move into Q3, the pipeline looks good, and we're very excited about what we're seeing.
Michael, I'll build on what Gina is saying by reinforcing the message I gave on cyber. Cybersecurity is going to be a massive tailwind for ServiceNow. The AI Control Tower for reinventing business is massive. Think about all of the assets, the incidents and the actions that have to take place to manage these large entities of government operations. Think about the Department of Defense as one example and the importance of saving people's lives — medical-material agencies in-flight, logistics, the devices, the coordination. The beauty of ServiceNow is, as Amit said a moment ago, it doesn't just think but it acts and it closes the loop. That is the full cycle of AI innovation and cybersecurity that we're building here at ServiceNow. As the attack surface becomes even more dangerous by the moment, we're up for the challenge. You're going to see us move up the leaderboard from #8 to contending for the top spot.
That's very clear. Congrats on the strong execution there.
Thanks Michael.
I have two questions combined into one. Your pricing is hybrid — token-based, assist-based — and competitors like Salesforce offer outcome-based models that could introduce pricing pressure. On the other hand, you own the data, you own the context and competitors need to get the data out, contextualize it and pay a lot for tokens to do that. So my question is: with many AI modules launching in the next few quarters or months, do you think on top of the value you can offer a price advantage versus competition? What happens to pricing of AI given all these differences between you and competitors?
Tal, we've been very clear in terms of keeping our pricing customer-friendly and thinking about how customers want to adopt our products and where they see value. We're never trying to introduce confusing models that are hard to manage or understand. The hybrid pricing structure we introduced, which is a combination of license and usage, directly applies to giving the customer predictability and flexibility. When you use more, that means you are getting good outcomes; otherwise you would not use more. Outcome is already tied to usage, so trying to define another metric is hard to understand and measure in enterprise contracts. That doesn't work. The way we address pricing requests resonates with customers. Regarding tokens, the key is outcome — the full solution versus just counting each individual call to an LLM. Customers are seeing the outcome from their usage, and they're also being predictable about what they will pay us. The second point: our context is what drives better results for customers. Competitors do not have our same depth of enterprise context; nor do LLMs. That's the secret sauce. When we run trillions of workflows on our platform, we collect intelligence which changes the game for our customers from a usage and results perspective. That's where differentiation comes in, and nobody else has that today.
To add on pricing: our pricing uplift for our Pro Plus SKUs continues to be above 30%. For our new AI-native SKUs that were just launched, those uplifts are in line with what we discussed at Financial Analyst Day, in the 20% to 30% range. Customers absolutely understand the value we're delivering to them.
Following up on the AI targets, which you're clearly exceeding: can you help us ballpark how much of the upside is due to adoption by new customers versus upselling existing customers into higher-tier packages? How should we think about what's driving the upside as you look at the roadmap?
Thanks for the question. The AI-native SKUs are making our agentic solutions more accessible to both new and existing customers who haven't tried AI before. Deal volume among first-time ServiceNow agentic AI buyers grew over 45% year-over-year, and uplift in pricing is 20% to 30%. So AI functionality is resonating across the board with existing customers and new logos. New customers are also leaning into AI first, as you'd expect.
I'll add that the use cases for AI are expanding significantly. Customers often start with a generative AI mindset and then move to resolution management, incident management and triaging case management. Once agentic use cases get started, we see a lot more interest to do additional work, which is why you see growth in AI adoption across the platform. This happens for existing customers as they identify workflows and find new use cases in other departments, creating cross-sell opportunities because we connect multiple departments and systems together. We become both a control tower and a full orchestration engine versus other vendors. The innovation cycle over the last 6 to 12 months has produced a large set of new differentiated capabilities for our customers.
On adoption specifically, we noted the number of customers with agentic AI in production has grown 9x over the last nine months. Adoption is progressing well on top of what we've discussed.
The 200 basis point beat on current RPO is impressive. We're getting questions on the mechanics of the full year guidance raise. Given the Q2 outperformance, the modest full year subscription revenue raise seems conservative. Is that just prudence or are there other dynamics we should be thinking about?
Thanks, Tyler. About half of the beat was a pull-forward of on-prem for a quarter, so that timing won't impact Q3 versus Q2. From a full year perspective, we passed along all of the net new ACV overachievement. I remain conservative in our guide, especially in this market environment. There's certainly upside, but we feel really good about the guide now. We raised to reflect the net new ACV overachievement and are being prudent for the back half as most would expect.
Bill, with all the talk about enterprise AI readiness and cyber driving increased awareness, investors worry buying cycles may elongate. Have you seen any changes or do you expect changes in deal cycles or decision-making? It seems ServiceNow is a consolidation point for IT spend, but any perspective would be helpful.
Thanks for the question, Matt. Think of it this way: frontier companies and hyperscalers are incredible, and the AI revolution requires activation inside the enterprise. We're in the bull's eye because frontier models need to be activated in enterprise workflows. That's the battleground, and whoever teams up with ServiceNow wins. We run massive workflows and transactions on the ServiceNow platform and built the CMDB early so we know the people, places and things. Over time we've integrated with systems of record and hyperscalers, so workflow automation and AI can operate without cumbersome workarounds. The attack surface is exploding and ungoverned assets multiply blast radius. We had a leadership position in security and operations and integrated well with other security companies. With Vesa and Armis, we're managing humans, agents and machines and going after critical infrastructure, people, networks and devices. The strategic relevance of ServiceNow is at the center of the AI revolution in the enterprise. What you're seeing is a company becoming both an enterprise software market leader and a cybersecurity market leader and the fastest growing in both. That's special. We've made it easy for customers to govern and control agentic AI. Seats are not going away; active seats are going up. Contract terms are elongating and pricing is delivering predictable outcomes. The AI control tower gives customers a kill switch for rogue AI when needed. Use the right model for the right job, manage tokens, and secure the stack — it's all critical. We're ready and our business is rocking. The new businesses we added last year complement this strategy and now it's all about execution.
Sales and marketing headcount was up quite a bit quarter-over-quarter and you've already added more sales headcount this year than all of last year. Is that timing, M&A, or is the company ramping distribution ahead of revised demand? How should we think about the sales distribution given the large headcount addition and how to think about that for the rest of the year?
Thanks, Samad. It's a combination of all three. Some of the increase is from M&A as we inherited some strong sales associates. Security and risk is ramping, so we're hiring ahead of that. We've always said we'll continue hiring quota-bearing, field sales reps and we will do that. You should expect to see more sales heads as we continue to invest; some of the increase this quarter is indeed from M&A.
Samad, Gina explained it well. We're a growth company and we will grow fast. We also committed to the capital markets that, after having acquired Moveworks, Vesa and Armis, we will complete this cycle of 2026 and start 2027 with the exact same headcount we had before the acquisitions. You'll see operating and free cash flow margins scale. You already saw a three-point operating margin improvement in the quarter, which exceeded expectations. We're serious about top-line growth, margin expansion and winning share in the AI revolution, especially cyber.
Gina, your gross margin guidance reflects more customers leveraging hyperscaler partnerships and accelerating AI adoption, which seem positive. Focusing on AI, how are you managing the tension between increased consumption mix and better token pricing? How should we think about upside versus downside risk to gross margin from here?
Thanks, Brad. Both increased hyperscaler usage and AI consumption can cause short-term pressure on gross margin. Notably, even though we had a little gross margin pressure in the quarter, we held operating margin flat. On the hyperscaler side, as we ramp faster than planned, cost per unit comes down and we believe mid-term there will be less pressure and better economics on the hyperscaler side. Regarding AI consumption, I mentioned at Financial Analyst Day that there could be some short-term pressure, but mid- to long-term we see opportunities to improve margins. All that said, operating margins continue to accrete and we remain disciplined on driving best-in-class top-line growth and margins, on our path to the rule of 60 by 2030.
To add on tokens and LLM usage: large language models are being commoditized and there's no reason to use the most expensive models for every task. We're deliberate about how we leverage emerging technologies — using open weights or domain-specific models where appropriate and frontier models where necessary. Capabilities can be swapped around and we optimize across the board to get the best outcomes while keeping costs low. Our engineering and research teams actively manage this and we work closely with many providers to optimize adoption. We're not worried about long-term margin impact because our pricing is based on the solution rather than individual tokens. That's how you should think about it long-term.
Bill, you've talked about cyber and high aspirations. Can you revisit your strategy and boundaries? You have Vesa and Armis — intriguing assets — but there are many areas you don't play in. How do you think about future opportunities in cyber? Relatedly, there's convergence in cyber and observability with new competitors coming from the cyber side. You've had some presence in observability; as these markets converge between workflow, cyber and observability, how do you think about revisiting strategy on observability?
Keith, we're building the world's most integrated end-to-end security platform. Think of us as the AI Control Tower and the agentic front door to Security 360. Armis gives us a major start in areas we didn't previously have full coverage for, and we already had a large security business. With our engineering commitment, we'll build many strong things. I discussed risk and compliance, identity and access security, cyber-physical security, continuous vulnerability detection and exposure management — these build on our CMDB and ITSM/ITOM strengths. We presented an organic roadmap today. If there are additional areas we wish to pursue, we'll consider them. We brought in leaders from acquired companies to run those businesses inside ServiceNow — for example, giving the CEOs of acquired firms leadership roles to preserve focus and domain expertise. The alignment of our platform and security-domain leadership gives me confidence in the product and go-to-market trajectory.
Keith, on security and cybersecurity we cover pre-breach with vulnerability and exposure management, including devices and assets. Then there's post-breach response where ServiceNow has been strong in managing incident life cycles. Now we have runtime security as well. We're bringing these pieces into one platform. There will always be peripheral areas in cybersecurity we won't cover, but where we have strength and opportunity — built on CMDB and ITOM — we create a compelling platform. On observability, we launched our AIOps specialist as part of our ITOM work, integrating observability signals into ITOM so customers can make decisions and take actions. We're not building observability outside the platform; it's part of ITOM and we integrate with other players to provide signals and then orchestrate actions.
To my friends in the security business, we remain partners. We run their solutions in our company and will deepen partnerships because the surface area is large and customers need the combined capabilities. The new information for shareholders is that we're fully engaged in this market and building our own integrated story on top of our foundation.
Bill, the commentary on the call is very favorable, but there have been indications from other tech vendors that organizations are shifting spend to hardware and AI, which can knock on IT spend. Have you seen any impact to ServiceNow sales cycles?
I have not seen a negative impact. This has become a C-suite story: running the business and optimizing processes. Our breadth and depth across workflow automation and enterprise processes gives us strategic relevance. Customers need a workflow automation layer that integrates with models, hyperscalers and systems of record. We're uniquely positioned with CMDB and our integrations. Many customers are consolidating and looking for governance and control as they adopt AI. The market reaction to AI has been positive for us. We're ready, executing and customers are turning to ServiceNow to manage this transformation.
You mentioned the AI Control Tower and controlling enterprise AI for customers. It's a huge opportunity and competitive. Could you talk about your right to win there? Is this more departmental or enterprise-wide adoption? Can you cite customers going through reviews and why they're choosing ServiceNow for that use case?
Adam, our heritage with the CMDB means we can discover and track any asset an enterprise runs, including AI agents and devices, across a heterogeneous landscape. We've always done open-ecosystem discovery, lifecycle tracking, cost management and observability. Building an AI Control Tower on that foundation allows us to discover anything running in the enterprise and manage it. That has resonated with customers: more than 500 customers are using Control Tower in the first six months since launch and adoption continues to grow. We keep adding capabilities such as the kill switch to stop problematic agents, identity governance with Vesa, and device coverage with Armis. The breadth and depth of these capabilities give us the right to win because they meet customer needs for control and peace of mind.
Running the business is what C-suite executives care about. They want to run processes efficiently — lead-to-cash, procure-to-pay, hire-to-retire — and ServiceNow is the platform to do that. To give examples: a Fortune 50 healthcare and retail distributor is capturing millions in annualized savings by consolidating vendor risk, governance and fraud detection across 1,000 suppliers on ServiceNow's AI Control Tower. Tech Mahindra signed a five-year deal to transform workflows across HR, IT and security and govern assets at mass scale with agentic AI. Maybank signed a five-year deal to establish a resilient operation center to fortify security and resilience and enable autonomous regional operations. NTT Data is using our governance processes to drive AI scale internally on a single platform, accelerating responsible innovation and monitoring in real time. Those are concrete customer examples.
Thank you all for joining. We look forward to talking to you next quarter.
Ladies and gentlemen, this does conclude today's conference call. Thank you all for your participation, and you may now disconnect.