ServiceNow Analyst Day 2026 | New $1.5B FY26 AI ACV Growth Target Unveiled At Las Vegas Forum

Welcome to ServiceNow's Financial Analyst Day 2026. Thank you for joining us today. Before we begin, I want to remind everyone that today's event will be webcast and recorded for future playback. Information pertaining to our forward-looking statements and a reconciliation of our GAAP and non-GAAP results are available on our investor relations website at investors.servicenow.com. As you can see, we have an exciting agenda for you all. Bill and Nick will kick us off and discuss our vision and opportunity. Amit and team will present the blueprint for agentic business, including deeper dives into the key growth areas that unlock AI transformation. We'll have a 10-minute break. Then Paul will go over our go-to-market strategy and lead a panel to showcase the tremendous value customers are getting from ServiceNow. Finally, Gina will close with a financial overview of the company's performance and outlook. So with that, let's get started.

Companies everywhere bought into AI, yet most still aren't seeing the return. Billions invested, but nothing's working together. Making it hard for people to get work done. That was not me. Correct, Nick. How did a ServiceNow AI platform break down the walls? Who are you talking to? The camera, Nick. Connect in any workflow, any AI, any data source. Through everything and everyone's finally new to get up in every corner of your company. We can resolve cases across department and deliver for customers. We can run HR workflows and improve employee experiences. And with the AI control, we can finally see and manage all our AI. ServiceNow is the one platform that lets you connect and control everything so you can put AI to work for your people. Take a seat, Nick, on a chair. Easiest if everybody just moves down one. No. Okay. Yeah. Okay.

Please welcome to the stage chairman and chief executive officer Bill McDermott and Vice Chairman Nick Sitson.

It's always great to come out to a video where the Nick character is a useless corporate bureaucrat. Um...

No, I'm not going to. Nice to see everybody. Anything you want to say before we dig into the content?

Sounds good. Well, why don't I bring up a few things one by one and ask you to comment on them. So, we'll go first to this. I think very few people here in this room, Bill, on either side, on our side or on their side, are interested in the past, but sometimes it's worth reminding everybody where the company's come from. So when you look at the trajectory of the company over the past several years, what comes to mind when you look at this graphic?

You mentioned an align board. I see our lead director Sue Boston is here. Paul Chamberlain is here from our board. Larry Quinnland is here from our board. So, it's great to see the board showing up. Yeah, I mean, you know, you got a lead director, Sue Boston, Larry Quinnland, Paul Chamberlain. These are individuals that have been with me through thick and thin. So too is our founder. We still have our founder on the board, which is awesome. And our board is a really great board, very committed to the company, very inspired by what we're building and how we're executing. And hopefully you feel that the entry point you're getting in at today is just like never going to happen again. So let's shift gears. The portfolio has obviously evolved substantially in the years that you mentioned. This is the representation that customers will see at Knowledge. When you look at this AI control tower for business reinvention, what are the important things for investors to understand?

So, one of the things that Colin and his team have done brilliantly acknowledge is the customer's voice is really out there. And I know you're personally inspired by several of those stories. Any that you would pull out just for this crowd as a preview?

Love that. Let me do a couple more just to set the stage. So you mentioned we're nice. Yes. And we've taken this posture in an environment where it seems like there's more coming into the enterprise than ever before. You always say trust is the ultimate human currency. What does that mean in practical terms?

Let's see if we can do two more in two minutes. So you've talked a lot about the growing opportunity. Yeah. It's not controversial to say there's some skepticism about is the opportunity in fact bigger. I don't want you to litigate the entire thing, but when you look at the prize in front of us, what is your takeaway about the trajectory ServiceNow is on?

Maybe we'll have a robot moderate this conversation next year.

No, I don't think so. So, let's leave it here. So, I don't want you to steal Gina's thunder, but you did review this slide, so she knows we're going to show it. It's not about rhetoric, it's about results. We've heard that before. What's the promise we're making about the results trajectory of the company over the next?

He'll be back for questions. That's a great way to start.

Please welcome to the stage, president, chief product officer and chief operating officer, Amit Z.

Great, great vision by Bill. I'm here to talk to you about how we are going to execute against it. It's great to be back for my second financial analyst day since joining ServiceNow. During FAD last year, I shared our vision and plans for enterprise agentic platform and I'm really excited to show you what we have accomplished so far. The last 12 months for us have been marked by innovation and growth. Our autonomous workforce of AI specialists are already delivering impact with customers like FedEx. Employee Works launched just two months after the acquisition of Move Works beat Q1 expectations by 5x. One of our data analytics products exceeded $100 million in ACV in its first full year of availability and data analytics is on track to be a billion-dollar business for us. The next evolution of our CRM products expanded into omni-channel intake sales order management and CPQ which will help our CRM business cross $2 billion in ACV and our security and risk business crossed $1 billion and expanded further into our AI identity governance and OT cyber security helping to differentiate our core. Our AI products are nearing $1 billion ACV and the momentum only continues and the AI control tower became a market-defining solution. So all this innovation and elite level execution enabled us to beat and raise our guidance in all the quarterly results last year. We also stayed true to our principles by continuing to expand our open ecosystem and the platform by partnering with partners across the tech stack and all industries. And we've taken also an AI native approach to transforming every corner of our own business as well as building AI into every product. And we delivered a new conversational experience across our workflows, launched monthly releases and auto automatic upgrades as well as powered more innovation by using AI tools and AI native approaches across all of our portfolio. But let's talk about today and fast forward to what we're going to be delivering as we go forward. Every analyst including all of you probably are covering, as you cover software, is asking if AI is going to replace the platform or AI needs the platform. The recent headlines are answering that question better than I could. At Meta, for example, an AI agent exposed sensitive data which is an enormous security incident with no external attacker. The own AI agent was the failure mode. An AI agent at Pocket OS hit a credential error and deleted the entire production database and all the backups of customer data in just 9 seconds. And the industry has been trying to put band-aids on all these issues using agents and spawning more and more agents. But none of these standalone AI products can solve the core fundamental issues because none of them can govern the system as a whole. And there's an important distinction here to think through. Autonomous work is not just powered by one new innovation. It's really two essential capabilities coming together, working together to drive real outcomes. This probabilistic AI which is what generates the answers and deterministic execution is what runs the enterprises and every enterprise needs both of those. And that is what it means for AI to be an operating system with enterprise context. And thanks to our CMDB and context engine, we know why a particular decision was made. We can encode the real-time relationship between every asset, person, service, dependency, and policy. I know many will say nowadays that 'I'll just build it myself. I'll stitch together the frontier models, take some open source models as well because I've been told that the next best SaaS application is just a prompt away.' But that idea falls apart when you test against three things. First, time to value is drastically underestimated. What should save a customer time and effort only ends up costing them more. Second, the total cost of ownership is very complicated. They compare a build versus buy decision with the cost of an API typically and an engineer salary. But that misses the point. You have to add model selection and the constant updates which happen with the new new releases from the model companies. Prompt engineering, security and compliance validation, managing new requirements while also not breaking the hundreds of systems which you already connected to which run the business itself. And that's just the start. The third fail test for build versus buy is the security and governance gap. Autonomous AI agents that take action inside enterprise system without a harness will create infinite risk surface. One compliance error from a homegrown agent could cost millions for a business. And we've been also looking at research around this. Customers building their own LLM based solution will typically spend five to 10 times more than using ServiceNow depending on the complexity of the business they have and the different systems they have to connect and all of the solution will take longer to build and it's usually less secure. So the ROI and the TCO is not even proven if you want to go down this path. And the ultimate trap of build versus buy is this. You know, building an app is definitely not the same as running a business. And that is why we made it possible for customers to both build and buy full solutions on our governed platform. At the core of the platform are four integrated pillars that create a complete sense, decide, act, secure loop for AI which is built on a modern AI native stack and this gives customers visibility, context, intelligence, automation, automated actions, governance all on one platform. Gartner projects that 40% of agentic AI projects will fail by 2027. But not because AI is incapable, just because AI isn't governed. And that's what we are here to fix. Our platform is about more than helping customers not just fail. It's about powering the success and growth. And we have proven that we can do this at scale. Today, ServiceNow runs 100 billion workflows and 7 trillion transactions annually, growing at 25% year-over-year. And that scale creates a flywheel. Every action on our platform deepens our operational context and enriches our CMDB and context that makes our AI better. Our new UI has higher repeat usage and we will only grow more with the launch of auto. Our unified AI experience you will hear more about later. So basically actions which creates outcomes, outcomes which create new actions. The flywheel continues to accelerate and every step makes our ITAM, ITM, SPM, ITSM and all of our core products more valuable than ever which is triggering more customer conversations and continue to expand our IT business as well. And humans usually need agents and agents definitely need guardrails. You get both through the platform we have delivered and this is a durable advantage we built first in IT and now expanding even further into many new domains. And customers are also feeling this compounding value in real time. I give you an example like CVS has taken hundreds of millions of actions now on ServiceNow supporting over 170,000 colleagues of theirs across 9,000 stores. Robin Hood is deflecting 70% of employee requests before human intervention, saving over 2,000 hours a month. Trident Care achieved 96% scheduling automation with our AI powered CRM, improving care for millions of patients while also increasing revenue. So the question becomes, what does it take to enable and govern autonomous work for the enterprise? Today my team will show you how it's done through the four pillars of sense, act, decide, and secure. But first let's start with the AI platform that everything is built upon. John, over to you.

Thank you. As Amit said, we're going to spend the next hour or so talking about five sections. I'm going to start with the ServiceNow AI platform and we have re-imagined and reinvented our platform and you're going to see a lot today. You're going to see autonomous workers. You're going to see the power of AI control tower. You're going to see new AI native experiences, conversational experiences. And we have reinvented our entire platform for the agent era. Now to be clear, we're still a system of action. We are the workflow platform and things are evolving. We're seeing patterns in the market that I'm going to address as we go through this presentation. But what I want to talk about first is that we have reimagined the platform from the bottom up. And what that means is that we've recreated the AI stack and it allows us to do things like innovate at the speed of innovation or at the speed of AI which is very very important because the expectations of our customers is changing rapidly. They want to see these innovations come very very quickly. They want new experiences very very quickly. Now the things that you're going to see around multimodality and voice and new conversational experiences, those are all part of the native platform. They're not bolted on. And the first area that I want to dive into that Amit talked about is this idea of context. And context is extremely important to any agentic system. It drives the AI. But first, I want to play a video of Boris who is the creator of Claude Code at Anthropic. And Boris is going to talk about how important context is to an AI system and why the ServiceNow platform is uniquely positioned for context. Could we roll the video, please?

Thank you Boris for joining me. Really excited to have a chance to have a conversation about all the work you've been doing. How's we've been also partnering between ServiceNow and Anthropic.

Yeah. I'm so excited to be here. There's so much to talk about and the world for enterprise customers is very fragmented.

That's right. I mean average customer will have probably 300 different systems underneath.

And in different world versions. Yeah. I mean the business is just so complicated. There's all these different tools. There's all this different process and data and so you need something like ServiceNow to organize and bring scan to all but all the dependencies all the data is so complicated and you make it so simple customers can just work with one platform and then they don't have to consider all this complexity you solve it for them. Today we run close to what is it, 100 billion workflows on our system and around 7 trillion transactions. So we are collecting so much information on why somebody's doing something, what decisions were made. So when I use a large language model and tools like yours and then combine that with the context engine and then the understanding of the whole enterprise workflow, it changes the outcome very much for the enterprises. If I'm doing something and I don't have the context, I'm not going to do a great job.

If I'm just told go do the thing, but if you don't have enough context and with a model, it's just exactly the same thing. You want to give the model a task. You want to give it a tool to bring in the context that it needs and it's just going to do such a great job at it. And ServiceNow is a really great way to bring in that context that it needs to do the job.

So as Boris said, context is the driver of AI and we have a context engine inside the AI platform at ServiceNow. But what is context? Well, context is not data. Context is not a decision. It's not an outcome. Context is actually the history behind an outcome. It is the decisions and what made up those decisions that are important to context and then important to the AI system. That's what our context engine delivers. Now, it doesn't just stop there though because there are new outcomes and new decisions that are created every day inside of our system. And so the context engine gets better over time and in turn the entire agentic system at ServiceNow gets better. Amit talked a little bit about this earlier. And someone in the industry would have you believe that every single process inside of the enterprise should be an LLM call. There's no need for structured workflows anymore. It's all about the LLM. Well, that's not very smart. It's not very efficient. And it is not the way that we want to drive our platform forward. Now, you're going to see a lot on agentic AI and generative AI today because it's extremely powerful. It's extremely flexible. So things like our AI and our AI agents and the things that you're going to see today are based off of that. However, all the context in the world doesn't fix the predictability of AI. Meaning the power comes from the idea that you could ask the same question twice and get two different answers. On the right hand side is structured workflows. We've been doing this for decades. These are things like flows and approvals and catalog items and they are what drive the enterprise today. And the trick here is not to offer one or the other. You need to offer both. And what we're going to do in the reimagined platform is to harmonize those two things together so that agentic systems and activities call into structured workflows and structured workflows can call into AI agents and we bring them together in a way that nobody else can do to provide the most efficient effective outcomes. Now, as I was saying up front, things are changing around us very, very rapidly and our customers want broad access to our system and we're seeing others in the industry exposing their system of records through APIs and through MCP to allow for reading and writing of their system, essentially becoming a database, a system of records. And while that's very exciting, that's not what we're going to do. We're going to expose the system of action and we're going to do that by opening up this layer to the clouds and the OpenAI and the Geminis of the world. What is the system of action? Well, that's where the true power of our platform comes from. That is our workflows, our flows, our processes, our skills, the context engine, playbooks, all of those things. That is the system of action that is ServiceNow. But we're not just going to expose that. We're going to monetize it. And we're going to do that by introducing something called the action fabric. And there are a few things you should take away from this. It is any protocol, any tool. So you can use your tool of choice and you can talk to the action fabric headlessly and kick off the autonomous work that is so powerful within our platform and it's governed. So all of the business rules and everything that's happening you can call into it and those actions are taken and workflows are triggered. But we're not going to stop there because we are now going to monetize that process. And as Amit said, we have a monetization model today for Generative AI. It's called Now Assist. It's a billion dollar business. It'll be a billion and a half dollars by the end of the year. And what we're going to do is plug directly into that system. So now that any time that an outside human being, machine, AI agent, third party agent calls into the action fabric, we're going to burn assist. The flywheel is going to spin faster. And this is a tremendous TAM opportunity for our company because now anything and anyone and anybody can call into the action fabric and take advantage of what ServiceNow is known for, automation. And what's going to end up happening is we're going to have this universal action layer where all of these systems are calling directly into our action fabric and spinning our flywheel even faster than it does today. Now there are other things that are going on in the market that we want to address with our platform. One of them is autonomous agents. These long running agents work on your desktop and they help you do things. They write code. They will handle your schedule. They will monitor your email. They're great. They're assistants. And we wanted to build one of those. And we did. And the first stop was talking to CISOs and security and they said, 'Absolutely not. You're not going to be able to deploy those things. They're unsafe. They're not governable. We don't want them.' So we started Project Arc with Nvidia. Our partners and friends from Nvidia, we got together and we said, 'Well, how do we fix this?' And what we ended up doing is using one of their technologies to secure these agents in essentially a sandbox mode. What that did was give us the ability to tell these agents what they can do, what they can't do. 'Please don't delete my entire inbox.' What systems they have access to and it gave us the control to allow a CISO to say yes. But we are the automation company. So we wanted to expose our agent and many many others. I think today we just signed a partnership with Anthropic for co-work to talk directly to the action fabric. And what that does is it allows your assistant now not just to manage your calendar or plan a trip for you, but it can also ask for time off and kick off these headless workflows that are going on in the background. So I can do things like take time off and change my HR or my benefits and request a new laptop all from these agents that are running on the desktop. Now, there's going to be a lot of these agents. You might have four or five running on a desktop. You might have tens of thousands of desktops across your enterprise. So the last thing we wanted to do with Project Arc was plug it all back into the AI control tower. And each and every one of our agents is talking directly to AI control tower, telling it exactly what actions it's taking, what systems it's trying to get to, and allowing somebody in the enterprise now to look holistically across the enterprise, not only at our agents that are running on the desktop, but at all agents, giving you that holistic view. Now, the last piece of this puzzle for the action fabric is build anywhere. And again, our customers, the ecosystem, our partners are saying, 'Look, we want to use the tool of our choice, but we love your platform. We want to build net new applications. We want to build primitives that run in your action fabric.' And in turn, what that does is it makes us a system of action. It gives us a tremendous consumption opportunity across the board and it drives the flywheel in ways that we couldn't have imagined before action fabric. And so what we want to do today is show you how easy it is to use your tool of choice and build net new applications, net new workflows on our platform. And to do that, Jithan's going to show us an awesome demo. Thank you, John.

Super excited to be here. Some of the incredible innovations John just spoke about, I really want to show you how it all comes together on a demo. AI and no-code has fundamentally changed how applications are built and how agents get imagined into existence as autonomous specialists in an app. I'm a developer. My HR team, that's the job I'm going to do today in front of you. My HR team asked me for a brand new addition to our employee benefits app. So typically I have few tool choices. I could be using Cursor, Codex, Gemini. Today I'm in the mood for Claude Code. Here I am getting on to the next screen which is Claude Code. And what I'm about to do is in the cloud I'm going to ask to help me build a pet insurance module for the employee benefits app. And the first thing I would do is in simple natural language, I'm going to say 'Add pet insurance enrollment to the extra care benefits app' which I just spoke about. Tap. And what you're seeing right now is Claude is starting to get to work. You are seeing something called Fluent. It's our ServiceNow's AI ready platform language. We have open-sourced our build agent skills and platform knowledge directly to agents like Claude Code via our SDK. That means the whole thing, the UI, metadata, access controls, business rules, all of it. And the last screen you saw was the app is almost built by Claude. Now I'm going to actually jump in and show you how it actually lives in our own native IDE which is actually ServiceNow Studio. Now let's jump over right here. You can see on the top of the list the benefits app. Let's go ahead and click into it. And the moment any application which lands on our platform like Amit was talking about and John was talking about, it elevates the app to an AI native application. Automatically when an application is built on our platform, our platform will actually recommend a set of agents who will live and breathe every day with the humans in the loop day-to-day 24/7. You can see our application has already recommended that we create an enrollment agent, right? Let's take a look at the agent. And in that, you can see the enrollment agent is all set up. It can qualify, understand and manage the end to end enrollment with minimal human intervention. Insights like this are possible only with the CMDB, the context engine, and the action graph which John spoke about earlier. And you can see I'm already on my way to actually create an agent. And every assist which is delivered in the app is a runtime assist for the user. Imagine every application having three to four agents living in it. Every interaction is a monetization moment every day 24/7. And now we'll jump into the actual building of the agent. Let's go ahead and build it. You can see with the click now, Claude Code with ServiceNow Studio and the SDK skills, the agent is already getting built and once the build is complete it's ready. It has its own roles, instructions, autonomy. It can actually sense, decide and act every day within the application, but we are not shipping it just blind. Let's make sure we run the right automated testing and scanning so that it's ready for production. And overall, now it's actually running through the scan process. It's going to come up with a readiness score.

Platform is built. It has run through its own evals. Come up with a score. Great score 90%. Now I am ready to actually submit for deployment. Now what you're seeing is app engine management center. This is the product or the platform where every application goes through to make sure it has the right security controls and the access controls everything built in. And by default any app which runs through the any agent or an asset AI asset which runs through it automatically gets registered as a part of the AI control to let's do this is where we do one more last review of the application. Everything looks good and let me make sure I have the right readiness which comes through with the release note and all of it. You can see I'm ready to now deploy the application right here. Now, this is the actual app which is live ready for you to prompt. Now, I'm going to switch my hat a little bit. I'm an employee and I would like to actually understand what are my benefits from my pet insurance point of view. Right here, I'm going to actually prompt and type in. You can see it's pulling in massive amount of user awareness and context at the back end. It looks very simple but in order to qualify and make sure this agent delivers the right accurate information is actually bringing in CMDB context engine every one of those process mining capability you can think of the last point I have a 2-year-old Frankie which I want to make sure I can actually get the insurance setup right there you go it's now going towards making sure I have my Frankie getting access to the coverage what it needs to be beautiful, isn't it? So what you just saw is the actual agent is live in action. Every app built on our platform will get auto autonomous agent embedded within. And that's how we are fundamentally changing the way AI apps and agents are built within our platform. I'll say it one more time. It's like John shared earlier. It's about you can build in anywhere any tools and any choice what you have run and govern and service now with the enterprise grade controls and security every app ships with an autonomous agent embedded and running the AI assist meter at scale. So that's the end of my demo and I would like to now bring in Gar who's going to talk about the sense part of the overall value problem. Thank you.

Thank you, John and Jim. So, you just saw why the Service Now AI platform is the system of action for autonomous work. And now I'm going to show you and talk to you about how we bring the sense and decide pillars to life on that same platform. We do that by helping our customers achieve four things. First, connect all their data wherever it lives. The second control it with enterprise grade governance. Third, contextualize it with enterprise wide intelligence. And then converge that contextualized intelligence right into the flow of work. And our customers are responding strongly to the strategy, making it one of the fastest growing product businesses in Service Now history. Workflow data fabric now has over 4,000 customers driving more than 3 billion monthly data transactions and we've recently added consumption based pricing and more than 700 customers have already consumed more than half a billion credits. And then there's Raptor DB where our new premium pro skew has seen explosive growth. We've gone from zero to 100 million ACV in five quarters with an ASP well north of a half a million dollars. Okay. So let me step you through the four C's of our strategy. First connect workflow data fabric is fundamentally architected for the agentic era. See most data fabrics are primarily built for decision support for insights. Workflow data fabric on the other hand is built for insight and action with read and write capabilities and it supports all types of data wherever it lives. And that's another crucial distinction. You see, we embrace the system of record and the data platform choices our customers have already made. You can, but you don't have to move the data into service. No, you know, I know there are others who are playing for data gravity, but for us, what really really matters is knowledge gravity. And virtual data fabric already offers 250 plus connectors. And we're expanding our reach even further with 100 plus zero copy connectors. So customers can access data wherever it resides. No replication required with full support for MCP clients. So our AI agents can work with any MCP enabled source. And with auto for workflow data fabric, our customers can describe what they want in plain English and let AI build the new integration. But look, connected data is not the same as AI ready data. Today, teams spend more time finding, cleaning, and preparing data rather than using it. It's slowing down AI adoption. It's limiting AI accuracy. And it's leading to some pretty frustrated data analysts. And I can't imagine the AI agents are too thrilled about it either. So to ensure AI decision accuracy, you need data that is fully visible and governed throughout its entire life cycle. You need tight control of your data. And this truly needs to be non-negotiable. So that's why we're introducing the Service Now data catalog, which delivers native metadata management, data lineage, privacy, and ultimately trust. So humans and AI agents alike can now instantly discover and use curated data products, safe in the knowledge that these data products have their enterprises seal of approval. That's great, but getting your data AI ready isn't a one-time activity. Enterprises have to keep it AI ready. And so to ensure the ongoing AI readiness of data, we'll be taking our data control capabilities even further with a complete AI driven autonomous data governance solution. And we'll be delivering data quality, observability, enrichment, and policy management all unified inside the Service Now AI platform. We'll deliver this through a combination of Service Now products and partner products in our massive workflow data network which now spans data quality, observability, MDM, security and integration partner products because just like we do with the data links, we want to embrace and extend what our customers already have and that's how our strategy is fundamentally different. We going to take that a step further by introducing partner passport so customers can procure and consume select partner products using service now consumption credits.

On to contextualize with more than 100 billion workflows a year running on our platform. No one is better able to understand our customers business context than us. And as John mentioned, we bottle that magic up into something we're calling the context engine. So you know the platform of platforms as Bill referred to now has a living graph of graphs. A graph that brings together our knowledge action access asset and decision graphs all anchored on our powerful CMDB. Built right into this context engine is our market leading analytical semantic layer and we've used that as a foundation for a new product that we're announcing called autonomous data analytics. And so think conversational analytics to guide the you know what happened what will happen what should I do type of decisions that need to be made by both humans and AI and it's fully autonomous. So think embedded, think always on. AI analysts working tyrannously on your behalf, surfacing insights, interpreting enterprise wide data in context, spotting outliers and issues, providing recommendations, even taking action. And soon we'll be packaging this capability into autonomous data apps, easy button solutions to bring the power of this insight to action capability right into our technology, customer and employee workflow areas. And so as an example of such a data app, customers will be able to combine product usage, support, contract, engagement data from Service Now and let's say Snowflake or data bricks, and then identify churn risk to then trigger autonomous customer retention actions in service now immediately.

Finally, converge. Today, enterprises typically analyze in one system, act in another. We decided to unify both at the database level with Raptor DB. We have Raptor DB standard which is freely available to everyone and a premium version called Raptor DB Pro that delivers even greater scalability and performance through advanced database features. Now we're adding two important new capabilities to Raptor DB Pro based on feedback from our large customers as well as those in regulated industries. So the first is live archive which is a cost-effective archival solution for service now that then also allows you to seamlessly query across hot and cold data and second life connect which allows you to point your existing BI tool against Raptor DB Pro for real-time analytics with no ETL or data movement. Together we feel these expand Raptor DB Pro's addressable market in our install base by 10fold. So there you have it. Core foundational capabilities to power and deeply differentiate service now's autonomous AI strategy by bringing data and intelligence right into the flow of work. Next, I'll hand it to Bhavin Sha to cover employee experience.

Thank you, Gora. I see a few familiar faces. I know some of us have connected over the years. I'm Bhavin and I'm responsible for service now's employee experience products and AI front door. I want to cover the third part of this chart here where we're talking about employee experience and the acting upon different systems on behalf of employees. Essentially what we're doing is building on the service now data fabric to drive this employee experience and drive values on top of that for customers. When Move Works was acquired, Bill and I got together and we felt that we were uniquely positioned to take ownership of the AI front door for the entire workforce. The reason for that was by combining the Move Works employee experience with the Service Now workflows and data fabric, we were creating a powerful front door for work across every system and every employee. Now, we've been busy. My kids call it integration maxing at home, but we've rolled out Move Works to every service now employee. We've launched the front door called Service Now Employee Works and we've integrated that front door into our new commercial model in just 4 months. So, lots of activity, lots of work going on there. We're moving fast because there's actually a gap in the market. Consumers are validating this in every conversation I'm having and in every deal now that we're winning. In the past 2 and a half months alone, we've seen a 10x pipeline build for our go to market efforts. That means that demand is here and we're now so ready to capture it given the two companies and what we both afford and can produce together. I'd say this, if there was an M&A award, I think Bill and I deserve it because both of our teams are on fire right now.

The AI front door is also moving fast and the frontier is moving fast as well, not just with the models, but with enterprises too. I'm sure you guys know this talking to customers, talking to different organizations. To build an effective experience across all employees, you need the ability to execute. This is really critical and hard to do. We started off with everyone being sort of captivated by AI smarts and generative capabilities. But then we transitioned in 2024 into what I'd call co-pilot chaos. This is when every sort of functional AI was being introduced, every platform, every offering with no real large-scale enterprise impact. We saw various studies come out. People were wondering where's the real value going to head? So today, what business leaders are looking for is what we characterize as enterprise AI. That means it works across the business, end to end, not just middle to middle, east to west, not just north to south. So the differentiation here is that service now and us together have been able to execute so fast because we can bring these capabilities end to end east to west to these enterprise customers. The thing to understand is that for an enterprise customer a user isn't just a user, they're actually an employee and this is where our differentiation goes even deeper. Employees have to navigate a complex journey spanning peers, managers, executives, countless systems, and unique business contexts. This is what it takes when you're running a large organization. These employees have a need that expands beyond just the end user that a lot of the offerings think of people as. We have a deep understanding of the workflows, the context, and the action. We're the only platform that meets these demands of the enterprise with the employees in mind to get work done. This is really resonating in the meetings we're having in the conversations and the deals we're winning.

They understand that we can see this through the lens of an employee throughout their year, throughout their month, throughout their week. In order to execute this end to end obviously requires another formula. The fully probabilistic yolo model of personal AI isn't sufficient. We've tried that and we've talked about it earlier. There are parts of the business that are not open for negotiation. They demand reliability. Think about a payroll adjustment, HR investigation and escalation. These are all things that have to be done a certain way based on the company's policies, based on the company's culture, based on how the company was created over the years. Not everything will be identified immediately and frankly some processes according to our customers will never be identified because they always want to make sure there's a human in the loop. That need for a human in the loop is something that service now does a really good job of unifying across for these organizations. My personal conviction of joining service now comes down to this. Personal AI gives you outputs but employee works in this new product that we've now rolled out really delivers outcomes end to end. Service now harnesses this and executes these plans in a way that gets work done for a company. Let me give you an example. Sending you to this conference becomes less labor intensive for your company. That is what real enterprise ROI is. Businesses want AI that can deliver outcomes similar to human labor. To do that, you have to go end to end. You have to go across all of these systems very effectively. In a world where everyone's token maxing and that's seen as a flex, enterprise AI is actually maximizing token efficiency. That's on everyone's mind as you know. What customers are really seeking is how can companies do this and this is ultimately how they will pocket the benefits. Now this brings us to a unified experience. I spend every day working on thinking about this, the whole team is rallied around this. With Employee Works we're able to deliver a seamless experience across the patchwork enterprise. We deeply understand the company and its employees. We bring together different platforms and workflows and that's what transformation for a real business comes down to. That's how we deliver real outcomes. But there's more. We have an AI marketplace that allows you to build deep into the Service Now platform, but also along with thousands of pre-built agents for popular apps like Workday, SAP, and more. This lowers the cost of ownership. The old model of expensive implementation, specialized developers, long timelines is gone. The new model is vibe coding, customized experience in minutes. This is how AI scales. Not by adding more tools, but by empowering more people to build on a governed platform. Let me give you an example. CVS Health, Fortune 10 company, 220,000 active users, 2 million conversations, a quarter million fewer calls and chats to IT and store service centers. This is real money to the bottom line. 40% year-over-year live agent chat reduction. Another example, Honeywell, they're an industrial giant. 80% of the inbound requests and work, the deflection, is being handled by the AI. The human mediated workflows are actually happening 60x faster because the AI is doing the intermediation. They're seeing equivalent reduction in labor costs. In closing, I've been selling to this market for about eight years to the Service Now install base. One thing that we've seen each time we would roll this out was that when you roll this out to all employees, there's actually an impact on the number of workflows and automations that get consumed because you lower the friction to get help, to find what you need, to do an action. That causes usage to climb. On top of that, what makes us really excited is that Service Now has 25 million active users on their employee center and this becomes the install base on which we're going to be building the future of employee work. It's going to be a really great year and we have a lot of other exciting product announcements and releases coming up for the rest of this year. I just want to give you a quick preview into what you'll hear more tomorrow with regard to Service Now Auto. You'll see this and learn more about it at the keynote that we'll have in the morning. Service Now Auto is really the combined intelligence of Move Works and Analysis coming together in a new unified experience. You'll be seeing this in action shortly, but also I want you to understand that we'll be handing this off to Pat to tell you more about autonomous IT and from there we can show you some more. All right. Well, thank you very much.

Good afternoon, folks. Thank you for coming to this. I know we're hoping to educate and share some information here, but hopefully I can add to that. Bob mentioned, I'm going to talk a bit about autonomous IT. I'm going to take a bit of a general statement up at the beginning. We'll talk a little about just workforce automation and then I'm going to dive more specifically into how we're applying that to the world of IT because that's kind of relevant to us because it's still the biggest part of our TAM right now. At least the biggest part of our current revenue base. This has been an automation company really since the get-go. I was one of the founders of the company. I worked with Fred Ley in the early days. That was always his mental model. He wanted to build a general case workforce automation platform. That's what's been driving value for our customers for 20 years. You take a process, you put it on service now, it's more efficient. That efficiency and productivity is what people are paying us for. That's the fundamental deal here. We make you more productive, you pay us. There has been a big change in the tools we have available to solve that problem though. AI gives us a new tool in the toolbox to apply. We do think though that we have a bit of a different take on how to apply that to the world of workflows than the traditional enterprise. I'll start by saying that we absolutely believe there's a value in what I'll call horizontal AI. Bobin just talked to you a lot about our Move Works Employee Works product. We bought a company here because we think it's a real value for our customers and for us but fundamentally horizontal AI is about interacting with you as a requestor of services. It's your unified place to ask for things, to get information, to kick things off, to check on statuses. It makes you more productive as an employee. Behind the scenes though you still have a variety of what I'll call vertical processes. If the thing you actually requested is a pallet of steel for a factory in Milwaukee, there is still a purchasing process which goes through before that pallet of steel actually shows up in Milwaukee. Those vertical processes are where the actual black letter savings are for AI because that's where people have a job. I'm a purchasing specialist. I'm a sorcerer. I work cases for a living. That's where the big value is for our customer base. It's in the verticals.

If you can solve the verticals, it lets you get out of the game of reporting things and asking for things and interacting with things and into the game of actually solving a problem, actually resolving something, executing a business process. So, you're not reporting that your email is broken. They actually fixed your email for you. You're not asking for a pallet of steel. You may ask for it, but some automation will actually make sure the right steel from the right vendor shows up at the right factory on the right day, and you can build a car. Fundamentally that's a business process. This however is hard. It's hard because most of these processes today are human mediated and many of them will probably remain human mediated into the foreseeable future. Human beings take part in these processes and a lot of them span multiple systems as well. There's various steps, there's state changes, there's technology shifts in there. Our industry has tried to work around this complexity by frankly throwing bodies at it. We will throw FTEEs at your project and we will try to wire up your purchasing process with some little shim here and some duct tape here and a little bit of bailing wire here and we probably get it working. But we've done one of your many business processes at a big investment of FTEEs. We fundamentally believe there's a better way to do this. We want to get out of the game of one-off bespoke process automation and into the game of giving our customers autonomous workers. This is a new paradigm. You saw Bill mention it. You're going to see all over the stages here. We are rolling out autonomous workers first inside our IT departments but also inside other workflows and ultimately beyond service. The idea behind an autonomous worker is very straightforward. It's just like a human being. You assign it work. It approves things. It produces the same audit rules. It follows the same logs. It lives in your user record. You don't have to do business process re-engineering to do one of these autonomous workers. It gets us out of the game of saving five minutes for a customer here and 10 minutes here and 20 minutes here and into the game of going to a customer and saying, 'Hey, you've got a 100 people doing that job. I can help you do it with 50 or maybe even 80, whatever the number is. It's less than 100. Let's talk.' That's the real value behind the automation in the age of AI.

If you look at where we're going more specifically, we have got a path to zero touch. You will see about 20 of these in the hopper, but four of these I want to talk to you about today. Level one service desk specialists. I'll dig into this one a little bit more in one slide, so hold your questions. Junior IT operators, asset analysts, and product managers for PMO. The idea is these are all jobs that our customers have that we feel like we can do some subset of that work with automation. So we can go to our customers and offer them that value. The first one we have which is this is live. This is not hypothetical. I've got six live pilot customers now. I've got about 50 customers in the hopper who want to get live with this very quickly. We've actually had to turn people away because the over subscription is so high here. But these are autonomous IT specialists. You put them in your user record. If you've already bought service now, you assign them to a team just like you on board a new human being. And they do stuff. They answer questions. They take basic actions. They solve cases. They do so in lieu of a human being, but they follow the same rules a human being would. This is important for our customers because it will help them get value and it will help them get more efficient and it will help them frankly save headcount. Fundamentally, that is the game they are in. It's important for us because that's how we grow as a company is we offer that value to our customers. And fundamentally it is important to all of us here because this is the productivity that AI is bringing beyond just the technology industry and into society as a whole. This is the promise of AI. We will help make the overall economy more productive. We're going to start with it, but it's not the end of it. With that said, enough of me talking about this. Let's bring Amy up. She can show it to you.

All right. Great. I'm already here. Thank you, Pat. That's fantastic. I'm incredibly excited to be here and share the product with you and share how what Bobin and Pat talked about really come to life. We're talking about both individual employee productivity which then transitions to the productivity of a team which then goes into the productivity of entire workforce when you look at autonomous workers. With that I'll show you how this all works. I'm going to start off here with employee works. This is our new unified front door to work. Imagine I'm an ITSM manager at Electri and I've got a lot on my mind as I start my day. First off, I just moved from California to Washington. I've been getting some physical therapy for my knee and I don't know if I'm going to get covered for that in my new state and if my current plan will do that for me. I don't know who to ask. It's the kind of question that might stump us all at work. Do we go to our insurance provider, our benefits, our doctor? We don't know. So today I'm going to go in and ask Otto what to do. I'm going to type in this prompt and see if I get coverage in Washington. Otto immediately gets to work. It's checking my plan, verifying that this life event qualifies me to change my benefits, and it's summarizing the answer for me. But not just that, Otto tells me what I need to do to kick off the process. I have not changed my address yet. So I'm going to go ahead and put in my new address and enter that in. Now Otto gets to work, updating my address across various systems at work, like Workday, which just saved me a huge headache. It's also telling me everything I need to know to change my benefits based on the policy, based on the context of who I am, what's available to me. I can assess it's a minimal change in coverage. This looks like it makes a ton of sense for me. So I'll go ahead and enroll in this plan. Otto, go ahead and finish the process for me. Everything's done. In about a minute, I went from a complex benefits question to a completed solution, which is amazing.

So, it's not just about me. I'm also thinking about my team at work and what they need. I'm going to start a new conversation and I want to see if there's anything I need to do to unblock my team, if there's anything pending my approval. Otto can go through all the different requests and things sitting in multitudes of systems and pull that together for me at a glance looking across any tickets that might be open or things that need my review. I see that summarized for me here and also prioritized, which is really helpful. I can see that Alex is asking for a hardware refresh and this is actually stalling his productivity because it's gone for quite a while without my action. Yikes. I'm going to dig into that and ask a little bit more before I approve this new computer. I want to understand what's going on here. Otto can bring up his request and then I can go in and actually click into that request and have it surfaced right here. This is a really cool part of our new AI experience. We can bring the information to you. You no longer need to hunt, navigate, go other places. I can see everything I need to know. Alex's machine is clearly out of date. Everything's maxed out on it. I'm going to go ahead and approve this for Alex. Fantastic. Always feels really good to unblock the team. Now, from there, there's another pending conversation, something that I've been asking about recently that has an update that I can jump back into. In this case, I have a couple of my engineers that provide VIP exec coverage and they have a shift that I need to give them specific access for. One of these team members, Jordan, I have to go in and actually review the access. But before I do that, I want to make sure that I can also revoke that access when their shift is over. I'm going to ask a couple questions. When does the shift start? When will this access be revoked? Can we do that automatically? Otto goes to work one more time. The first support engineer is provisioned automatically. But the second one Otto can go through and provision and also make sure that their access will be revoked at the appropriate time, which is super cool. I'll go ahead and approve it. Everything looks great. I just did stuff for my own personal productivity as well as the productivity of my team, which is fantastic. It's a totally new employee experience, which is so exciting.

Next up, although that was a lot about employee productivity, we also know that sometimes despite everything we do, a team can get swamped. There's too many inbound requests. I'm going to go over to my service ops work desk and see how my team's performing overall and go into my service ops dashboard. Unfortunately, even though I did all that great work for my team, there's still some bad news. Our backlog is up. CSAT is dropping down. We're still overwhelmed by the volume of work coming in. Like Pat talked about, that's where an L1 IT service desk AI specialist can come in. I'm still learning about what this can do for my team. I'll ask Otto what I need to know about it. Otto again goes to work summarizing everything that this IT agent can do for my team. It's reviewing specialized capabilities. It's also projecting how many incidents this could solve for my team. It brings up an entire profile right here on how well this AI specialist will perform including the eval score which gives me high confidence that this will perform at a level that I need for my team. I can look through and see everything set up including the skills that this AI agent will use and also escalation path so in case there's a really complex issue it can get routed to a human. This looks pretty fantastic and based on what it's going to do for the CSAT of my team, there's no question I'm going to activate this AI specialist. All right. And just like that, it was added to my team. That's incredibly easy for any manager who's feeling swamped or that they don't have enough resources. They can activate these AI agents and add them to their team. No admin, no configuration, no deployment, just a few clicks and it's there. We'll fast forward in time and I come back to that same dashboard and I see great, everything's tracking on time now. We've got our AI specialist on the team. Things are progressing well. CSAT is back up. But I also want to audit how this AI specialist is performing. I'll go in and ask Otto to give you an analysis on this AI specialist. It goes through, looks at all the past activity, looks at the metrics, seat scores for this individual, and it pulls together again an awesome comprehensive briefing for me. I can see that the specialist is handling 52% of all requests, CSAT of 4.6. This is fantastic, but I also want to audit exactly what it's doing in a particular instance. I'll click into one, and I can see a full record here of exactly every step that that AI specialist took. So I have no doubts about how it's doing this work, how it's resolving these incidents and the effectiveness it has not only for my team but for those that it's helping. Great. Everything's working really well. This combined team of both humans and AI specialists working together. This is truly the future of work. Not just AI that assists, but AI that acts and resolves delivering real business outcomes. Very excited to share that with you. Next up, John Ball will be joining us to talk about our innovation in CRM. Thank you.

Good to see everyone again this year. I'm going to be covering the fourth key step to unlocking AI transformation, act, and I'm going to do it through autonomous CRM. Let's get straight to it and start by recapping just how far and fast we've come in CRM. In 2023, I was up here and telling you that we have become the fastest CRM player ever in the history of the industry to get to a billion dollars in CAC. And now, just three years later, we're going to blow through that, double it, and blow through two billion. We've massively expanded our functional footprint to deliver awesome experiences across the entire customer life cycle from lead and opportunity management to configure price quote and order management all the way back to where we started in customer service and field service. We're doing this at scale, managing over a billion cases per year, over a billion order and work order tasks. At peak times in CPQ, we're configuring 100 times a second every single second. We're recognized as a leader in CRM by analysts like Gartner, Forrester, and IDC. We have industry specific IP that speeds time to value across multiple verticals. This growth and success is driven by our deep understanding of how to solve real challenges in delivering great customer experiences. I'll give you a hint, folks. It's not about tracking interactions in the database. In service, you have to provide great omni channel intake of requests and you have to make resolving those requests easy and efficient. In sales, you have to go beyond just tracking leads and opportunities. You have to make it fast and easy for sales reps to configure, price, and quote those opportunities. In all of that, you need workflow, powerful workflow. You need the ability to model the products and services a company sells as well as the types of requests, orders, and changes their customers are entitled to because without workflows and without the ability to model this declaratively, you're just writing a bunch of custom code and a vibe coded app on top of a shaky foundation doesn't resolve the request. It just makes disappointment happen faster. So whether it's handling a warranty claim, disputing a Visa transaction, or ordering a new telecom service, all of these examples require powerful deterministic workflow at the core. Now, what AI does change is how customers, sales reps, and customer service reps interact with these systems to get the job done. With conversational AI, you can talk and chat with the system using natural language. That's cool, but it kind of reminds me of some great Elvis lyrics. A little less conversation, a little more action, please.

Because you don't reach out to a contact center or a customer service center to have a conversation. You reach out because you want action taken to resolve your request. Understanding this point is crucial because the vast majority of customer service requests are not how-tos. The web and YouTube solve that. I'll use a simple example to illustrate my point. Say you want to change an existing order. This seems simple and straightforward. But to deliver this, you need to understand the intent request order change. Then you need to understand all the specifics. Is it a change of the delivery date or of the quantity or of the actual product being ordered? Conversational AI is great at capturing all those intents, but then workflow is required. If it's a change in quantity, do we have enough inventory? Or if it's a change in the product being ordered, you've got to rerun the CPQ process to check for compatibility and then generate a new quote. That's CRM workflow logic. That can't be solved with AI alone. Here's my most fundamental point. You have to get it right every single time. You're certainly not going to run refunds, disputes, orders, or anything else mission critical in a stochastic process that might hallucinate. In sales CRM, CPQ is a great example where AI can really turbocharge productivity. Imagine sending a draft quote to a prospect just minutes after a Zoom call with that prospect that is tailored to the specific requirements they described in the Zoom call. That's now possible with AI powered CPQ as long as you have headless and speed of thought CPQ engine that enforces all the compatibility rules, bundling, discount policies, etc. Luckily, we do. This is not theoretical. We're live in production with sales, service, and CPQ use cases with amazing customers like you see here, driving better customer experiences and millions of dollars of savings. Several of these customers are presenting here at Knowledge. So you can hear their stories firsthand, not from me, from them. From Nvidia, who reduced time to quote from 5 days to 5 minutes. It's just an amazing stat, 5 days to 5 minutes. Or Rossman, a large European retailer who deployed our Agentic CRM for retail store support, saving a massive amount of time and allowing store associates to focus on the customer, which is their real job. Now, there's no better way to understand this all than through a demo. Please welcome Chris Shots, CEO and founder of Logic.ai, who now runs all of sales CRM for us. Chris, take it away.

Pretty complex generator. It's got dozens of options, dozens of different pricing rules. It's a pretty complex piece of equipment and he wants to change his order. What he's going to do is call into a leak tree generators and interact with one of our CRM voice agents and see if he can get his order changed. Let's see how Marcus does. Hello, this is Jennifer from Electric Generators. How may I help you today? I placed an order a month ago and I think I need to make a change. Sure. What is the name on the order? Marcus Webb. Thank you. I see a few orders. Is it the most recent order placed on March 15th? Yep, that's it. I see order 1034 for a 10 kW generator. We are about to start the manufacturing process. What can I help you with on this order? Okay, so here you can see Marcus has interacted with our voice agent. She's asked him a couple questions about his order and behind the scenes, the AI agent is querying our order management system and finding possible orders that might be for Marcus that he placed a month ago. She then clarifies the correct order and then finds it. Let's see what Marcus wants to change on this. I'd like to change the order. I'm not home often. I want to add an automatic transfer switch and remove the manual transfer switch. I'd also like to add a maintenance plan so it's always ready. Okay, sounds good. Let me see what available options we have for that unit. It looks like we have the automated transfer switch and we have additional options for the maintenance plan. For a 10 kilowatt machine, we recommend the gold plan. This plan covers all the recommended preventative maintenance and includes a quarterly check by a technician in person. Shall I go ahead and update your order? Sure, go ahead. Okay, so what's happening here is the AI agent is interpreting what Marcus is saying in the large language model and then mapping it to options in our configure price quote application. So this is what John was just talking about with this concept of speed of thought and 100% accuracy. We've got an agent that's querying the option availability in CPQ and then making sure that all of these options fit together real time. In this case, Marcus needs an automated transfer switch and a maintenance plan and the agent is making sure that all that'll work for his order. Hi Marcus, your order has been modified. Great. I also just realized that I need to add a 220 volt circuit. Sure. One minute while I check machine compatibility. Now, this is a pretty typical case for a complex, expensive order like this where there's a lot of back and forth dialogue with a customer, especially when you're doing order edits and order changes in manufacturing systems. Not only do we have the voice agent interacting with CPQ to make sure all the options fit together, the performance is really important. We put a lot of work into our solving engine that runs inside of our CPQ app to make sure that to John's point, we can get answers that are speed of thought so that Marcus has a good interaction with the AI agent real time. Yes, it looks like we do offer a 220 volt circuit for this machine. Let me put you on a brief hold while I finish making the configuration changes. So now that we know exactly what Marcus wants, the AI agent can take all those options and then run the manufacturing bill of material rules because this is an engineered piece of equipment. The bill of material is dynamic based on the options. She also needs to run all the pricing rules and then perhaps most importantly figure out when they can actually produce it and manufacture it. She needs to give that information to Marcus before she can close out the order. Hi, Marcus. Your new order total is $13,526, an increase of $2,000. Delivery will now be May 15th. Would you like me to email you the full proposal now? Sure, that sounds good. The email I have on file is [email protected]. Is that correct? Yes, that's correct. So now that we have the product configuration, the bill of material correct, the sales bill of material that goes with this machine, now the agent is taking all that information, putting it into a customer facing document that she can then send to Marcus real time so we can confirm billing address, shipping address, payment terms, and things like that. Great. I just sent you the updated proposal. Let me know when you receive it. So now Marcus has the email on his phone. Okay, I received it. It looks good. Can we make the changes? Sure. One minute while I make the order changes we discussed. Okay, you're all set, Marcus. You should receive a confirmation once order management is done processing in the next few hours. Is there anything else I can help you with? No, thanks. Thank you for doing business with Electric Generators. Have a great day. So now you can see Marcus in a couple minutes was able to interact with a voice agent and actually make some pretty complicated order management changes using the agent interacting with our configure price quote and order management applications real time. Thank you.

Great job as always Chris. So hopefully for those in the audience that demo helped you understand how we're combining AI data and workflows to really change the game in CRM. We're delivering better customer experiences while improving our customers cost structure, which is really something that every company on the planet wants. That's a huge tailwind for our business. And with that, it's time to pass to my colleague John. John, be coming up here. Take it away, sir. There's the clicker. Thank you so much.

Thanks so much, JB and Chris. As JB mentioned, my name is John ID and I'm responsible for our risk and security products. What I'm here to do is to walk you through the fifth layer of a five layer cake that Amit and my colleagues have stepped you all through. I'd like to start by making a bit of a bold claim. Service Now as a security leader. We spent the last five years already establishing significant preeminence in the governance risk and compliance market growing that business tremendously, nearly four times greater growth over the last 5 years than the market itself is growing. My primary goal of the next 10 minutes is to essentially provide you with the proof points that back the claim that I'm making on the slide. Let's start by grounding you on the foundations. As Gina and Jean shared at the end of Q3 last year, our security and risk business, a bit of an unsung hero within our portfolio, surpassed a billion dollars of ACV at the end of Q3. We grew our business security and risk organically 40% in 2025 versus 2024.

What has this growth been powered by? Two foundational anchors that have enabled us to achieve the results that we've achieved. First, our IT asset data gravity. All right. For 20 years, we've been the preeminent provider of IT asset insights and the workflows built on those assets that enable compounding in terms of the asset gravity that we have for a typical enterprise. You combine that with the east to west coverage that we have across so many buyer persona and user persona: back office, human resources, supply chain, source to pay. I promise I won't go through all of them, but essentially 10 primary persona that we address across the enterprise. And if you believe like I do that all enterprise data is ultimately security and risk data, you could see how that provides us with both the right and the responsibility to achieve the preeminence that we've already achieved.

But what we realize is as we're building the next generation architecture for this platform shift that we're going through, the agentic AI platform shift, we needed to add to the IT asset data gravity that we already have. So over the course of the last two quarters, as you probably have heard, we've been a bit busy inorganically. So one of the first things we did was acquire Armis. In one sentence, what is Armis to a CISO? Armis is a cyber asset graph. It enables us to take our IT asset dominance and provide a comparable view of that same data to the CISO while adding incremental attributes and asset types that we previously did not cover exceedingly well. Code unbelievably powerful in this age of AI, OT, IoT, medical devices, but all of that is our cyber asset graph. You combine that with what we did in March by closing on our transaction of Vasa, and at its bare essence Vasa is an access graph. It's a way for a typical enterprise to gain insight into who and what has access to what, and then you build extremely high value applications on top of that, agentic or not, that is extracting insight from that data plane. And of course the third dimension in the multi-dimensional core that ServiceNow risk and security is our knowledge graph, essentially the enterprise context that can take the exact same set of assets and access in Deutsche Bank and that will generate a different outcome from the exact same set of assets and access in Allianz because the context of those organizations are different: different policies, procedures, rules, regulatory frameworks, people, etc.

So it's this combination. If you want to leave with anything from this presentation, our core in ServiceNow risk and security is powered by cyber assets, the things, access to those assets by humans and other things, and enterprise context. All of this furnished across all of the platform building blocks that my colleagues have been talking about, and this combination is already meeting at the customer. We're not the only bright folks that have this insight. As an example, a global international financial services leader, whom I might add has a representative in this room, is using ServiceNow risk and security today, has combined that with Armis for capturing connected asset information across the building management systems, and able to extract potential vulnerabilities from those BMSs and essentially automatically remediate prior to any incident occurring. This same customer also uses Vasa to enable visibility and intelligence by both humans, systems, and agents across the 50 or 60 AWS services that power a subset of the cloud estate. And it's a comparable story that we see across an international consumer packaged goods leader whose story I won't step through for brevity.

What are the growth drivers both current and future that are powering this business? There's four that I'd love to leave you with. As we become a full-blown workflow to use the ServiceNow platform, I expect my colleagues in finance to begin to share insights about this business over time. Here's three leading indicators that I want you to remember. One, how is this business doing organically? How are we growing usage as a key leading indicator to then growing ACV? That's one. Two, as many of you saw in our April 9th announcement around our AI native packaging, we did something super interesting. We enabled the entirety of our customer and our partner base with the rights to AI control tower, the delivery mechanism for all of this IP. So one of the additional leading indicators that you should watch for is our effectiveness in building net new IP and driving our existing IP to market as an attach to AI control tower. The third, I've been in the cyber industry directly and indirectly for about 26 years. Was Brad? Remember him from 25 years ago? And one of the dreams of the CISO ecosystem through this time is collective defense. Collective defense. The attackers are actually collaborating on the deep dark web. As you all know, the economic and the architectural prerequisites enabling the defenders to actually collaborate have been, shall we say, lacking. I believe that the machine speed by which agentic workloads can identify code defects, chain them into vulnerabilities, and do malfeasance with that requires a next generation architecture. And the action fabric that SIG talked about earlier is the enabler for collective defense in the enterprise, all of, and I'll describe why in a bit more detail in a minute.

Um, this is the architecture by which everything comes together. The data plane I described earlier powered by assets, access, and knowledge, manifested through the context engine and delivered through the action fabric to both ServiceNow risk and security workloads and third party workloads. So imagine a circumstance where even though we're now a fully fledged provider of exposure management solutions post Armis, a customer is already using CrowdStrike for exposure management, they're using Microsoft for endpoint, but they want incremental value arising from that three-axis core that I described powered by ServiceNow. We can serve that up in action fabric, and CrowdStrike agents, Microsoft agents can take advantage of that derived insight for both decisioning and for action. That's the next generation architecture that we're making, all of which spins meters for ServiceNow as we're delivering value to our customers. So to wrap, our enterprise data intelligence layer that's unparalleled in its coverage of buyer and user persona in a typical enterprise requires us, provides us with a responsibility to enter this market in a material way in the way that we have, and to become a participant in the primary table by which the next generation architecture gets built. Um, as this platform shift becomes increasingly mainstream. Action fabric is the collaboration layer, the enabling architectural building block that enables all custom risk and security workloads, third party ISV workloads to collaborate with ServiceNow workloads to reduce security, increase security outcomes, increase risk outcomes, and generate business outcomes as a result. And the last thing I'd like to wrap with: remember the role that zero trust played as an architectural northstar of a cloud world. As workloads moved unmasked outside of the firewall, a comparable next generation architecture is needed for the agentic world, and I will posit that the notion of permanent access privileges in systems is going the way of the dodo. It just doesn't make any sense. And so we're going to play a role in making that vision a reality by taking the IP that we have, the IP that we've acquired, and the people that we've assembled to contribute to this zero privilege architecture becoming a reality. So to provide you with some insight into why I believe this claim continues to have credence, I'm going to have Nenshad Vitaliwala, our AI platform product leader, demonstrate this through a simple demo scenario. Hey,

Thanks very much, John. You've seen a glimpse of what the AI control tower can do, but today we're going to dive into the secure risk and compliance capabilities. So, first off, let's look at CVS Health. They serve over 185 million people a year. And AI powers all of their operations with hundreds of models, agents, tools, and prompts. The greatest unmanaged exposure for CVS's security team is the speed at which AI is increasing their attack surface. Today, I'm going to be the security admin in CVS's AI center of excellence, focused on securing our AI investments. The AI control tower gives me a single view of our AI security posture. Now, the moment I log in, I'm going to go ahead and check out the governance page, look at security, and what I see is that AI asset security score has dropped overnight to 28%. Now, I just need to figure out why. And sure enough, a new alert gets flagged up here that the Aetna benefits AI agent has anomalous privileges. This agent helps millions of members understand their coverage in plain natural language and it's at risk of leaking PII. Member names, addresses, phone numbers, they can all be inadvertently linked and leaked to other members. So, let's go ahead and start the remediation process.

I click remediate and immediately Armis' early warning flags agent vulnerabilities and active exploitations targeting healthcare. And while this agent was built with good intentions, Veronus detects it's gained elevated permissions and can share PII with other agents, requiring immediate action to stop a potential data leak. In addition, our asset intelligence here on the right shows how this information can be shared between other systems and agents. So going back to this overprivileged state of the agent, I'm going to approve the recommended access permission adjustment and multiple things are going to happen simultaneously as a result. Let's go ahead and do that. Yes, please reduce my agent access. So, we're going to temporarily disable the Aetna benefits agent. We're going to work with Vasa to remove permissions. We're going to update the AI control tower inventory. And then the exposure record is automatically created in the unified security exposure management application created for my team to review when debriefing. Now, this agent's elevated permissions and its connection to other critical assets were caught because it was under continuous monitoring. So, I asked myself, what other risk, vulnerability, or active threat may be out there that we don't know about? Let's go take a look.

So, I can see here that I've already gotten my score up to 78%, which is exciting. But I can go ahead and use the power of Armis' shadow AI detection and find that there are three other AI agents running in CVS Health's business units. This is shadow AI without any visibility or governance. Each one could be the next compromised or overprivileged agent. So I'm going to change these to managed. An AI control tower opens up four new use cases in USM: one for the original agent and three for the additional shadow AI agents that our asset discovery capabilities identified. I mark them and once they're managed, now continuous security governance and controls monitoring happens across every managed agent. So here's what I just showed. We caught an agent with elevated permissions serving 37 million members and we contained it with a full audit trail in a single operation. Number two, we uncovered shadow AI agents that no one previously knew even existed. And number three, we created incidents for all of these findings with audit ready documentation. In this representative example, I am firmly the human in the loop as CVS's current policy for agentic execution requires. But the AI control tower can also operate fully autonomously from continuous monitoring to anomaly identification to remediation. And with that, we're going to hand it back over to Amit to take us home in the product section. Thank you.

All right, thank you Nenshad, thank you John and to all the speakers. That was just amazing. I think I'm sure you can see how proud we all are of what we have built and a strong innovation roadmap. The world-leading analysts also make it clear that they agree. In the last few years, analyst recognition has grown from six categories to leading in 39 categories, and that number continues to grow. And recognition isn't just limited to our traditional products; across the new categories, especially agentic AI, ServiceNow is now consistently recognized as a leader. And our future roadmap is also accelerating. That's partially because AI is not just empowering our customers to build and work faster. We also are leveraging the same AI technology to speed our innovation as well. What you see behind me is the breadth and depth of what's coming in 2026 alone. Across every area, we are shipping AI native capabilities and autonomous workflows to power our customers' agentic enterprises. And underneath all of this is a platform that continues to get stronger and more powerful every day. And the one key piece of our roadmap strategy is acquiring specific features and functions that accelerate how we deliver value to our customers. I know there are a lot of questions about our M&A and the investments we recently made. So let's talk about it directly.

Bill also alluded to earlier that we're making strategic decisions to acquire both tech and talent that strengthen our platform and also further differentiate our core, and we already have proof that that is already working. You heard that. But then in terms of the big bets we have made from Bhavin, Amy, and John Ball in terms of how successful our products have grown and how well we are getting differentiated in the market because of those acquisitions getting integrated into our product portfolio. For us, these acquisitions are not just about buying growth. They're also about delivering the critical pieces to power our customer agentic capabilities and agentic enterprises that fit right into our platform, making it stronger and more relevant.

We have also taken the same AI native mindset that had reimagined our products and our roadmap to evolve our commercial model as well. As many of you know, seat-based pricing does not reflect the value of AI. So we have moved beyond it. AI is now embedded in every tier of our products from foundation to advanced to prime. Another shift is how we meter value. The unit of value is the work that gets done in real time. And we now have hybrid as well as consumptive meters across all of our entire portfolio today. And this gives customers something predictable like a subscription commitment and something which is flexible so it scales with their actual adoption. And we're using different meters across the whole portfolio. Now AI assist, for example, for AI and data related products, human and non-human identities for Vasa, assets for ITAM and Armis. And over the last year, hybrid and consumptive meters accounted for more than 50% of our net new business, and that number will only grow.

Innovation in our commercial model and our products also go hand in hand. For example, we're building an AI native service management product designed for the mid-market. It will be entirely consumption-based and conversation first. This will be one of the first of many ways we're taking the strengths of a platform to entirely new customers through new channels as well. And with a shift towards autonomous workforce, we're also going beyond traditional software budgets and tapping into the labor market. Customers can now hire, manage the performance, and retire digital equivalents of human workers for a fraction of the cost. And these autonomous workers can be added or removed on demand and are constantly expanding their skill sets as they learn more and more as part of our platform. And this is the hybrid workforce of the future which we are now making available today. Our AI specialists capture at least 6 and a half times more value while saving the customers over 80% compared to the cost of human fulfillers which it replaces.

And beyond the numbers and beyond the market recognition, at the end of the day, customers are the biggest proof points of our success. The enterprises that run the world have trusted us with their most critical operations at one of the highest renewal rates in the industry today. So we have covered a lot of innovation in the session today. So I want to come back to our four pillars of everything we do: sense, decide, act, and secure. And of course the AI platform that everything is built upon. Here's what we know to be true. We are in the midst of the most significant enterprise transformation ever. Every company in the world will leverage autonomous work. The only question is how and who helps them get there safely. We have the platform, the architecture, and the track record. We have the customer relationships, the partner ecosystem, and the talent. We know how to execute it at speed and scale. And we have a commercial model as you heard before to capture the true value of AI. And we have done this before. We know what it takes to lead a market through a generational shift, not just participate in one. Thank you for joining us today. I hope you are as excited as I am for what's ahead.

So, it might be a bit of a controversial start, but the problem isn't just in the technology. That's not the hard part. The problem is in changing the organization. Um, and when I talk about the organization, we're talking about the people. We're talking about the governance, and we're talking about the processes. So all of these parts really is where companies struggle because they have to get all of those pieces in line and then the technology actually comes quite naturally and quite easily. Um, you can always find very smart people to build something, to find a workaround to make it work. We've had those problems in our deployments and we worked through them. That was okay. Um, but really the harder part was getting all the stakeholders on board, working out what processes you wanted to keep, what data you wanted to keep, and how it needed to operate. So I'd say that those are the harder things. The technology actually you guys have made it quite easy for us. The technology is there.

The stakes are just significantly higher. I mean, to put that in context, and you just said it, right? So, we operate in 220 countries. We move about 18 million packages daily. And we manage that through multiples of tens of millions of workflows. It's very different if you're playing a one person band or a two person band. You can sort of get the music right. But now if you're playing an orchestra, you have to get that right across 50, 100. Take that at the scale of an enterprise, for every decision, every workflow, every task to be seamlessly coordinated, to be seamlessly executed, takes a lot of precision and to have the right data available at the point of insight and action that it's needed, that data you can trust, and then try and bring the entire organization along to the point that Oliver just made around talent and change management. It's an entirely different operating mindset for the enterprise. You can't do that with a one or two person experiment on the side. You can't do that when you know that on the receiving end of your execution is life that is waiting for that parcel to get delivered on time. I mean, healthcare is our largest segment. Aerospace is a pretty big vertical that we serve. These are high-stake businesses. You can't introduce risk in an environment. You have to be able to trust that anything that you're doing, whether it's AI or otherwise, you can introduce it responsibly across the breadth and the depth of those workflows. That's the hard part.

So I think it's always been a partnership. Um, I think we met sort of four or five years ago. Um, and I think the partnership started really there with customer success, and then we were invited to participate in the lighthouse program. Um, and that was really where we, and Hitachi Energy specifically, got to help design and build some of the components that we're now seeing today. Um, and really that allowed us to help you shape it. Um, it didn't come without its road bumps. Um, and there were definitely some technological challenges that we had to overcome, working with the customer success teams, working with the product teams to actually get these things fixed and then deployed at scale, because as Victor just said, this isn't just about one or two people. We deployed it to 60,000 people overnight. Um, and when we did that deployment last year, we saw the real benefits instantly. We saw a 25% reduction at our service desk in people contacting it. Um, we saw a 10 times increase in self-service on the service desk that just was not there before. Um, so we saw marked changes. Um, but actually being able to work with you and partner with you to help develop it and change it. Um, and one of the biggest areas that we really, I think, helped with is on the whole value creation. Um, and this is something that we're seeing in elements of control tower. Um, but really then the value estimation and generation because it's something that I was always getting challenged by my CFO on: 'Well, how are you proving that the investment that we're making in the time and the licenses is actually returning a positive value for me?' Um, so being able now to see that and to quantify, and you have all the data in the system, you can see what is being done faster, what's being done slower, what's even not being done at all. Um, and that now you can see it a lot more easily. We went through the spreadsheets, the PowerBI, the manual ways of calculating, and now we have tools and products to do it. Um, so I think that part of the partnership, and I do treat it like a partnership, has always been there and it's really helped us become successful and see those sort of results that we've had.

Sure, Paul. As we just outlined, FedEx operates one of the most complex supply chains in the world. And we've been doing that for about 50 years. And the one thing that we've come to acutely understand is the amount of fragmentation that exists in supply chains. And that amounts to about 1.8 trillion of value that's leaked annually because of that fragmentation. So what we want to do now through FedEx Data Works is make sure that we bring solutions to market that allow for that inefficiency to be trapped. We want to make sure that we build orchestration solutions that fill the void and connect all the elements of this fragmentation inside the supply chain. So starting with source to pay announcement that we just made, we want to make sure that the insights that we generate, the first party data that FedEx network generates, which is about two petabytes of data, we want to release those into signals that benefit our customers so that they can go from a reactive to more a predictive posture in the interventions that they want to drive inside their environment. Source is one example. We want to then extend that to building workflow solutions that allow our customers to more seamlessly orchestrate supply chains inside their enterprise, and that's where the partnership with ServiceNow and FedEx gets pretty exciting.

You know, just because I can doesn't mean I should. I mean, for me, it boils down to that. I set the stakes high and, you know, it doesn't mean that we will not build stuff. You have to take a more nuanced approach. We're very keen on making sure that the core of our value chain where we want to have the differentiation, where we want to go and help our customers do more deeply, are areas where we will want to own IP and we will want to build. But I don't want to be known as the best HR system company in the world, the best IT system company in the world, or the best finance system company in the world. I will shamelessly take that from folks that have much more experience than FedEx will ever have in this enterprise and bring PRA best practices inside and build the digital backbone that you can help bring with speed and channel my energy instead into the core parts of our value chain. So for me it boils down to: just because I can build, I go back to one or two person band, doesn't mean I should. There are ways we will build and there are various areas where we will partner and increase our speed to market.

Of course. And look, I mean Hitachi is an engineering company. We're not short of people who want to build stuff. Which is a great thing. But I think as Vic said, it's like just because you can build it doesn't mean that you should build it. And I completely agree with that sentiment. Take the best bits from everywhere else. We're not a software sort of predominant house. We build Hitachi Energy. We build huge transformers. We build SK switch gear. We build trains. There's loads of stuff that Hitachi builds. Let's keep building that. Let's keep building what we're good at. Let's buy what we haven't, what we don't build ourselves. And then integrate it together. So I think it's maybe slightly less of a build versus buy comment and actually more of a build versus orchestrate, and how you assemble this into your digital backbone or your digital core, whatever your company sort of refers to it as. But how do you bring the pieces of the jigsaw together into something that then works for your organization and helps drive your organization forwards? And you can buy things from ServiceNow, from Salesforce, from Microsoft, from Google. I pick whoever it is, but you've got to bring the best parts of that together for your enterprise. And every enterprise is different. Your needs are different to my needs. I'm not shipping hundreds of thousands of packages around a day. But then we do have mission critical infrastructure that operates in a different way. So we have different needs. So take the best parts that other people make and build it into your best solution.

Please welcome to the stage, President and Chief Financial Officer Gina Masteno.

Thank you, Paul. What incredible customers. Um, and hello everyone. I've been really waiting backstage for a while to get out here and I can't tell you how excited I am. So, um, not only am I excited to be here, but I'm really excited not only about where ServiceNow is today, but about the scale of the opportunity that we see in front of us. Listen, I know there's questions swirling out there in the market, and that's healthy. Great companies should be able to answer hard questions such as: will the growth of the foundational models come at the expense of budgets for the incumbents? Will seat compression shrink revenue? Do software stocks become only margin stories? These are questions I get all the time. In ServiceNow's case, the answer to each of these questions is an emphatic no. We are a unique platform company. We're the AI control tower for business reinvention. And as you've heard throughout today's presentation, ServiceNow is not a traditional SaaS company. We're the orchestration layer AI agents run on, not software they replace. We're the AI operating system for the enterprise. We are truly in a category of one where AI makes us both more competitive and more profitable simultaneously. We have a clear path to grow the top line and drive continued margin expansion to deliver durable shareholder value. We'll look at this next, but first I want you to walk away remembering three things. One, our structural advantage is ours and ours alone. AI only reinforces it. The AI super cycle is a revenue tailwind for ServiceNow. Two, ServiceNow is the AI platform enterprises are already buying. It's not a bet on future potential, but a flywheel that's already spinning, where the best-in-class workflow and governance layer where enterprise AI value occurs. And three, margin expansion and AI growth are not at odds. They're the same story. AI-driven internal efficiencies fund the innovation and bring focus to our growth investments.

Let's dig deeper into the growth opportunity. Our core business is strong. It comes down to execution. As Bill talked earlier, this is a company that executes. In 2025, we grew 20% year-over-year to nearly 13 billion in subscription revenue. We're looking at a 5-year CAGR of 24%. And we added more than two billion in revenue in 2025 alone, which is more than the entirety of our subscription base in 2017. Now, assist ACV crossed 600 million last year, more than doubling year-over-year. That momentum carried into Q1 with ACV crossing $750 million. Now, assist isn't separate or distinct from our core workflows. They are our core workflows. This kind of organic innovation is a powerful growth catalyst for it. Much like how AI fueled customer demand in the upgrade cycle from standard to pro, our better together story continues to strengthen. In 2025, 91% of our net new ACV came from deals with five or more products, up from 86 the year before. This includes a 7.5x increase in now assist deals with five or more products. Customers are not experimenting with one AI solution in one corner of their business. They're deploying AI across the enterprise. Customers are going all-in in our core technology workflows where we saw over 50% growth in deals with five or more tech products.

Also, we have three key growth accelerants. Security and risk is the next growth vector for technology workflows. In 2025, net new ACV grew 40% year-over-year. And at less than 20% penetration from the base, there's plenty of room for expansion. With AI control tower, customers govern AI across the enterprise. Its ACV has quadrupled since launch. Armis and Vasa further extend the TAM. CRM represents a massive market opportunity crossing 1.8 billion in ACV in 2025. Sales CRM is leading the way with ACV more than doubling year-over-year. We win because of our single platform across the entire customer life cycle connected natively to service and operations. Data and analytics is a multiplier which more than doubled net new ACV year-over-year. As you heard Gar Gar say earlier, Raptor DB has already surpassed 100 million in ACV in just its first year.

Every AI agent deployed and every custom workflow built pulls demand for data connectivity and performance. All of these growth vectors also have underlying tailwinds created by the proliferation of AI data and assets. More agents deployed means more governance, more data connectivity, more platform usage. That's why half of our net new ACV has already shifted to non-seat-based pricing models as they catch those tailwinds. Those underlying units including assets, infrastructure, platform usage are seeing significant growth. We don't count seats here. We count dollars with the strength of AI adoption. We're also seeing a growing mix shift towards consumption. That's why we're democratizing access to AI with our new AI native packaging. Every new SKU has a bundle of tiered capabilities across the core product, AI, workflow, data fabric, MoveWorks, and our AI control tower. As our customers purchase those higher value bundles, we expect to see an average price lift of 20 to 30%. This new packaging also unlocks customers' AI consumption journey earlier. Now at every level, consumption becomes an incremental growth driver as enterprises scale usage of assist, connectors, and the underlying assets being governed. Then as customers look for more advanced capabilities, they upgrade to prime for our most premier offering. AI consumption is already showing up. Existing now assist customers who renewed in 2025 expanded their ACV by an average of over 3x. It's not just about assist packs. Every cross-sell, every subscription purchase is adding assist and is part of the consumption story.

As I'm walking you through my growth story, you may be asking where will the budget come from to pay for these incremental consumption costs? AI spend is expanding budgets for ServiceNow. Customer conversations we're having today are focused on reducing labor costs to fund ServiceNow's advanced AI capabilities. Let's play this out. If you had a team of 20 support analysts today, the team would cost over a million dollars annually. About 90% of that is labor. 2% is ServiceNow. Now what happens when you move into an agentic AI world? As enterprises look for efficiency, they'll naturally target their largest cost center, labor. ServiceNow's autonomous AI agents can resolve 75% of the team's work, reducing the necessary headcount to just five. The customer wins. Their total cost to get that work done drops 65% and resolutions happen in a fraction of the time. At the same time, those 15 freed up seats convert into 6.5x more in AI agent consumption. Just like Amit talked about earlier, even after accounting for license reduction, total ServiceNow spend grows over 5x. Lower cost for the customer, better experience, faster outcomes. That's the definition of a compelling value proposition and it's driving the shift we're seeing today. AI consumption compounds at workflow as workflows get more complex. Generative

AI laid the foundation with each task consuming 1 to 10 assists. Agentic AI deepens the value curve by completing multi-step tasks, consuming significantly more tokens. Autonomous AI specialists represent the next step change, purpose-built to perform specific job functions end to end, with the two layers combined consuming over 15 times the assists of generative AI. As AI solves more complex workflows, usage rates climb and so does the value we capture. This is just the beginning. Our autonomous workers will cover all corners of the enterprise as the platform scales. Docyign is a customer exemplifying this journey from Gen AI to their first agentic use case and now the zero-touch service desk, their first step towards an autonomous workforce. Using ServiceNow, Docyign has a target of autonomously handling 90% of all IT tickets so human agents can focus on the most critical work. They expect to save millions, and the opportunity is massive. Docyign is realizing their vision of true workflow transformation and creating a playbook that they can replicate across their entire business.

This is just one great customer example. You've already heard directly from FedEx and Hitachi earlier about their incredible AI journeys with ServiceNow. What does it mean for ServiceNow at a larger scale? Let's look at IT incident management. Just one use case within ITSM, we see over 100 million incidents per month on the ServiceNow platform today. If 75% of those incidents can be processed by an autonomous workforce, this translates into a $3.5 billion ACV opportunity net of any seat licenses that go away. Multiply that across other ITM use cases and then across the entirety of our platform. We power 100 billion workflows and 7 trillion transactions annually. And you can see the incredible opportunity in front of us. How long does it take for a customer to realize autonomous outcomes at scale? With all new technologies in the enterprise, it takes time. But as you heard from Paul and others, we're finding every possible way to help our customers accelerate that journey. Let's look at an example. When a customer purchases agentic AI capabilities, they receive a generous assist allocation, meaning assist overage tends to be relatively more limited in the first couple of years. What starts as a strategic deployment across ITSM, CSM, HRSD becomes a foundation for enterprise-wide AI transformation. By year three, they're taking on more complex agentic use cases, inflecting consumption. Year-over-year AI ACV compounds, driven by the deepening adoption across the enterprise as the customer naturally graduates up the value chain to autonomous workflows. The result is a fundamentally different revenue model with fewer seats but far greater value. We anticipate by year five, this AI customer would be spending 4.5 times the initial assist entitlement.

These journeys have already begun. As Bill teased at earnings, we're raising our 2026 AI ACV target from $1 billion to $1.5 billion. The demand we're seeing is real. These are not features bolted on to existing products. They are built in. Their solutions have strong adoption and measurable outcomes already. When AI attaches to existing workflows, it doesn't just add revenue today. It makes the platform stickier and expands the ACV opportunity for each and every customer. This is a flywheel that's already spinning, not one just being built. And looking further out, by 2030, we expect 30% of our ACV from ServiceNow AI. When the unit economics work, when trust builds, when complexity scales, this is what happens. And I really just love that map.

Okay, that should give you a sense of our growth story and trajectory. Now, let's turn to profitability. AI is structurally expanding ServiceNow's margins. I'm going to repeat that. AI is structurally expanding ServiceNow's margins. I'm often asked whether AI inference costs will compress our gross margins. That framing doesn't apply to us. AI reasoning is less than 10% of our cost to serve. If inference costs rise, the margin impact remains modest. Customers aren't paying us for tokens. They're paying for a resolved outcome. Reasoning is one input. Workflow orchestration, governance, context, cross-system action — that's where the other 90% of the value and cost fit. We're differentiated, and our pricing reflects the full platform: the CMDB, the workflow engine, the governance layer, the business service map, 20-plus years of operational context. That competitive positioning is what sets us apart from the standalone AI providers and why our gross margin profile holds. This allows ServiceNow AI to continue to ramp with subscription gross margins remaining above 80%. While our move to hyperscalers is impacting gross margins in the short term, the ROI on that strategy has paid off as net new ACV from public cloud partners nearly tripled year-over-year. We're also not just selling AI solutions. We're using them ourselves. AI is driving meaningful year-over-year gains in output from our fully ramped reps. We're growing the top line while getting more efficient with every sales dollar invested. We're also seeing an acceleration in the incremental savings from agentic AI flattening the hiring curve. The $200 million in savings in 2026, that's on top of $100 million that we saw in 2025, for a total of $300 million in expected annualized cost savings from agentic AI flowing to the bottom line in 2026. AI agents are doing 90% of the monotonous work. ServiceNow's own support and service operations have been rebuilt on our agentic AI. This margin expansion is structural, not cyclical. We are the proof of concept. Every customer is being shown what ServiceNow has already built at enterprise scale.

Turning to our long-term targets. I know you're all waiting for this all day. We keep you here for this part, right? Okay. Phil gave you a little preview, but I've got a little surprise. So hold wait, please. In 2021, we established a long-term target to achieve $15 billion in subscription revenue in 2026. Many were skeptical then. I see a few of you in the room. Fast forward to today, we're on track to beat that target by half a billion dollars organically. I know the number is higher than $15.5 billion. Organically, we're beating that by half a billion dollars. Not many executive teams can say that about their long-term targets. I know you all know that, too. Our momentum puts us on pace to double that target in 2030. That's $30 billion-plus in subscription revenue, and it's not a blue sky scenario. It's what a durable platform growth story delivers. As you heard from Amit, though, we haven't been standing still. We've accelerated organic innovation to catch the tailwinds from emerging opportunities made possible by AI. We've expanded the TAM with recent acquisitions. And while we're not asking you to underwrite this upside today, we see a strong path to it. A higher 20% CAGR from our current 2026 guidance and $32 billion of subscription revenue in 2030. Pretty impressed that I got to say that number, right? And he didn't take it. But this is not blue sky. There's a roadmap of real defensible growth engines, and they are not heroic assumptions. Security supercharged by our demand for AI control tower and the new TAM unlocked by our acquisitions of Armis and VZA. Data becomes even more critical as enterprises deepen their AI investments. You heard that from customers today. AI has upended the CRM market, and we are taking share together. These three vectors compound at over 25% growth year-over-year through 2030. Most importantly, cutting across all of it is AI, agentic workflows, autonomous workers, unlocking value in ways that simply did not exist two years ago. And I would note this does prudently bake in a deceleration in some of our more mature products in this environment. It's a show-me story. I get that. We get that. That's why we're ensuring your topline growth option with a commitment to continued strong profitability. Combining topline growth and profitability at a level few companies achieve at any size, let alone ours, puts us on a trajectory of achieving the goal of 60-plus by 2030. This is what we're building. A business that delivers accelerating value for customers and shareholders simultaneously, year after year. That means focusing on GAAP profitability as well. Two years ago, we committed to getting stock-based comp below 15% of revenue by 2026. We did it in 2025. We also told you that sub-10% is the longer-term destination. Today, we're telling you when: 2029.

Let me put a finer point on how we're returning capital to shareholders. We doubled our share repurchases in 2025. In Q1 alone, the $2 billion ASR represented nearly double the shares we repurchased in 2025 all in one quarter. The result: we expect to be dilution net neutral for 2026. We still have $4.2 billion in authorization remaining. So we have plenty of capacity to keep managing dilution going forward. We're always evaluating the best use of capital to maximize shareholder value. Our framework is clear. First, we reinvest in organic growth at high incremental returns. Products like AI control tower, Now Assist, workflow data fabric — these strategic internal investments deliver. Second, we deploy capital into acquisitions of technology and talent with a focus on tuck-ins that open new TAMs or meaningfully accelerate our product roadmap. Third, we are committed to returning capital to shareholders, balanced against the significant growth opportunities we see ahead. We will continue to be thoughtful about that trade-off. With the sizable free cash flow generation that will come with our significant margin expansion, we will also have tremendous flexibility as we think about capital allocation in the future. So let's end where we started. I know there's questions swirling in the software industry, and it's easier for some people to put us in a box with others. The fact is we are in a category of one. ServiceNow is the orchestration and governance platform that AI requires more of, not less. The AI super cycle is a revenue tailwind for ServiceNow. We showed you the math today. It works. The margin expansion is structural, and we are living proof. Our own AI transformation is the strongest customer reference we have. We bring growth and profitability. Two engines of shareholder value, both powered by AI. That's how you get to the rule of 60-plus at more than $30 billion in revenue. Thank you all for joining us today. And now we're going to welcome back Bill, Amit, and Paul.

Okay, go here. Yeah, we going over here.

Okay. Uh, thanks everybody. Kirk Matern from Evercore ISI. Thanks for a great presentation. Both the technology depth and the long-term vision. It was great to see. I think my question is somewhat maybe for everybody on the panel. But I think one of the debates going forward is going to be at the orchestration or at the agent control plane. I think every big enterprise, the LMS all understand that a lot of the value might accrete to that layer. It's also super early days in terms of agent deployment for most big companies. So, as investors, what should we watch for? What are the milestones that we can see from you all to know that you're hitting on that strategy? Because right now, I think everybody's sort of in a land grab, and I think everybody's asking the question. I think most people understand you have the right to win there, but what are the metrics, the KPIs we should be watching for to understand your strategy playing out?

Yeah. No, thanks for the question. So I think the way I see it, there will be a lot of pieces being orchestrated by different parts of the technology providers, but the idea of an autonomous worker takes away this requirement to do individual work by each of the vendors. So what we're doing with autonomous worker is taking away the effort you have to put in to create individual agents, manage them yourself, figure out the orchestration, the reasoning, and all that, which is not really very conducive to a long-term way to manage a business. So the metric I would look at is how many people are now going to start adopting autonomous worker and how we see that kind of proliferation of AI specialist inside enterprises so that they can now get away from dealing with individual pieces themselves but getting the full value of a solution. So what we're seeing now with AI specialist, for example what we've introduced, is this 20 AI specialist, especially the L1 support specialist. We're seeing a lot of customers don't want to do that building and managing and taking care of the spare part themselves. They want to elevate that and get a full solution. So I think that will be a trend over the next few years versus what is happening now. Same thing happened with cloud. If you remember, everybody used to buy LAMP stack, they would want to build themselves, then eventually they realized keeping and maintaining that stuff, dealing with changes in the technology, is not easy, and they ended up going to hyperscaler, somebody who provided the full stack, and then you can build an application with it and you don't have to deal with all the different pieces. And I think the same thing will happen with us now, and that's the metric I'm watching for, and I'm seeing that already play out with a lot of customer conversations I'm having.

Uh, thank you. Obviously very impressive to see all the progress you guys have been making and the targets coming out. I think seeing that ACV target of 30% coming from AI out just a few years is really amazing. But I guess on the other side of that is kind of implicitly it would suggest the non-AI components are going to be growing much slower. Back of the envelope, I did was 10% or less CAGR during that same time period. I think I've had an argument to investors about how that's the wrong way of looking at this. But I'd love to hear from you all as you've received that question of like, but what about the rest of the business and that seems to be growing slowly. Is that a worrisome sign? Help back me up on why that's maybe the wrong way to ask the question.

I love that one. So AI is the core. You're exactly right. Customers want to buy products with AI built in. And that's why we introduced our AI native pricing and packaging. It's why even if people don't want to go full in automatically all the way up to Pro Plus, which is now Prime, they can start with Foundation. They can taste it. They can start working because who really wants to buy any software today that doesn't have AI embedded in it? So it's 100% the wrong way to be thinking about it. And remember, when we first launched Pro, right? No one said, well, the core standard is declining and Pro is doing so well. They said, oh my goodness, Pro adoption is fantastic. This is wonderful. And you get 25% price uplift. So many more. That's exactly what's happening today. Only now it's 30% on top of Pro. And now we're embedding AI into even the foundational pieces. So not everyone has to go full stack right away, but they can really start utilizing the AI, understanding the benefits. We firmly believe that once they start seeing the benefits in a small scale, they're going to much more rapidly proliferate and grow with us. And that's when the consumption wheel continues to fly. But you're already seeing pretty incredible growth in $750 million in Q1 up to 5. This is remarkable growth and we are just getting started. And so it's wrong to think about, well, if AI is doing so well, it must mean the core is not. It means AI is pulling the core, and that's what we continue to see. We're driving all of our customers to be wanting to consume more AI. That is the benefit of the ServiceNow platform. It's going to help our customers when we really see the value accrue over the next one, two, three, five years.

Hey guys, Alex Zuken with Wolf Research. Wonderful presentation. Obviously, you would expect nothing less. A couple of competitors out there saying, 'Hey, we're taking some share from ServiceNow.' You have two targets out there, and I would say none of your competitors have the security angle, and it seems like the variability of the upside is partially driven by executing on this new, I think Bill you called it the largest new TAM of cyber crime. So maybe just talk about how do you see the competitive narrative and landscape evolving and how does the security component of your portfolio drive that maybe higher target that you laid out there?

I'll just add one thing. The other part which is also important is the data part. So the stack we have built, having an AI platform which as you saw today is very deep, has very critical functionality required to run any kind of AI systems out there, including all the workflows we've been talking about. Now having a data platform which can bring all the information together to do the context engine work we talked about and be able to make decisions very quickly and predict a better outcome really drives our workflows and outcomes much better than anybody else can do today. And you bring security on top of that where you ensure no nefarious things will happen in your platform, really changes the game when you go and talk to the customers. They don't have to bolt on all the things separately. They don't have to bring all the signals from separate areas and manage all that stuff in a different environment. So when we bring security and data on the same platform and then you have workflows, the agentic workflows, the autonomous worker, it's really, I don't think there's anything out there in the market today. There's no competitor who can do that. They can talk about pieces of it. So if you hear the noise out there, they're talking about pieces of technology, maybe mid-market or somewhere else in the enterprise space. There's nobody like us. And then what I talked about earlier about bringing sustained technology now to mid-market, that also takes that business away from them. So you'll see a lot more of our capabilities being delivered, AI native, completely consumption-driven, and conversational experience now delivered for AI native for mid-market with the native mindset, and bringing ITSM and other capabilities to mid-market we didn't do before. That gives another opportunity we have not talked about earlier.

Thank you guys for taking the question. Keith Weiss from Morgan Stanley. Thank you for the great presentation. Two questions. One real quick question on timing. Gina, you mentioned the pro skew and we saw a really nice adoption of the pro skew. Can we expect now a sister prime, or not quite sure what we're calling it now, to ramp similarly to what we saw from the pro skew? Number one. And number two, more strategically, you guys are bringing agents to your customers, you're bringing the autonomous worker, you're bringing the AI specialist, but you also have to open up your platform for your customers to build their own agents or let other people bring their agents. So when we're thinking about the 5x uplift from autonomous worker, what's the uplift when you're just opening up your platform and letting other people build agents on top of your platform like they're going to request?

I'll take the first one and I'll help. So on the first one, actually we're conservatively building in similar ramp with our AI that we had with pro. I actually think, and so far it's actually been slightly accelerated to what we saw with pro, but our numbers that I showed are building in the conservative estimate of similar penetration glideway as we saw with pro.

Yeah, on the agent stuff, I mean I think there is the architectural evolution which is happening where they are going to be agents interfacing into applications. They're not just going to be users or humans just interfacing. So having an ability for us to provide that access in a governed manner with monetization is the right way going forward. So I do expect more and more of that kind of use cases emerging. And what we've done is we've been very careful about how we expose it. So we have an MCP server. We allow agent-to-agent capabilities as well as communication, but we heard about action fabric. The idea is that we will wrap this thing with a set of APIs, governed but also monetizable. So we measure every assist using the assist kind of metric, measure any access, and we meter that, and people can burn down. So it makes our assist more fungible, not just able to do Now Assist kind of use cases but also now accessing data, but we can get to manage it and make sure that you register for it, you have what kind of requirements, the SLA, the security. So we're bringing all that stuff into this thing. And one of the announcements we did is with Anthropic to be able to also do things with their co-work, but it's also in the governed and measured capability so that it doesn't allow people to just get access to it without any kind of permissions.

Uh, thank you for taking the questions. Arjent Vatio with William Blair. Actually, I wanted to follow up on Keith's question about Action Fabric and I'm curious if you worry that customers might push back that you're essentially introducing a gate for their data. And I'm curious how much of it is theirs versus yours. Or does this all not matter at all because the ROI is going to be high enough that customers are going to come out winning on top of this. And then one question for Gina. I'm curious when the pricing model evolution might take place such that most of your revenue, not net new ACV, is consumption or non-seat-based. How do you see that playing out throughout 2030? Thank you.

Yeah. I'll address that. I think we have been talking to customers about how they want to access our environment and what ways they want, one what the volume they would need, what is the different kind of metering we would require for that, and what would that metric cost would look like. And so far it's been pretty well understood by them. They realized they used to access things but there was no guarantee of SLA, there was no way to know who was accessing and what security issues you would run into. So when you provide a much more governed platform, they seem to be very comfortable with it so far. And when we've been introducing this concept, we've talked to many customers already and there's never been an issue in terms of worry about having this kind of metric been delivered. And given that it gives a fungibility with the Now Assist, it also makes it much easier for them. I think it's not a separate thing which we are introducing or creating a new kind of metric which would be confusing otherwise.

On your question with respect to non-seat-based, when we think net new ACV would be more. Listen, I was really thinking that you all have been pretty impressed that we're already at 50% so far. And by the way, it's been increasing pretty rapidly over the past couple of years as Now Assist has been driving a flywheel. And so, we haven't given timelines for what we expect that to look like long term. I do expect the 50% will continue to increase. I don't think we'll ever be 100%. I think some of our business will always be seat-based. And if you think about the new AI native pricing and packaging, just by virtue of the initial subscription you're getting a large chunk of assists in there. So there is consumption already baked into that initial seat, and so consumption will continue to be a bigger portion as we go forward.

Exactly.

Thanks very much, Brad Zelnick, Deutsche Bank. Really appreciate a very compelling presentation today. Bill, I wanted to follow up on your comments about now being the right time to go further down market. Why now? And why might it be different? Because in the past, it seems medium-sized enterprises didn't really value the platform the way that large enterprises did. And I'd be curious as well what your view is on AI readiness in kind of medium enterprise and relatedly what does this mean for partner leverage? Does this create an opportunity for a whole new set of partners? Thank you.

I think it's great Bill, and I think during my presentation I talked about autonomous built into the product. So Ahmed and his team are really innovating on self-implementation there, but then also leveraging AI on our services implementation standard which will be launching tomorrow and really compressing that time to value. We see a massive opportunity on the AI readiness part. What I would say is, these mid-market companies, which I think is where you're going, Brad, in that segment of the business, I think the innovation that we've done over the past couple of years really enables them to be AI ready with their data. So things like Raptor DU Pro, workflow, data fabric, all the things that the teams have innovated on that we didn't have even two years ago. Now you can go in, you can plug that in, you can get the data ready because we know AI success is going to be all about the data and grounding these models inside of that data. So the new innovation capabilities combined with what Bill's talking about on the self-implementation but also the autonomous implementation work we think is right. If I could just finish with one further comment there.

Hi, good afternoon. Gabriella Bortis from Goldman Sachs back. I wanted to ask Paul a follow-up to some of the case studies that you showed earlier like the level one ITSM case study. So it strikes me that the outputs for some of these case studies are really beautiful but when we actually think about what a complex enterprise environment looks like it can actually be pretty heterogeneous in practice and you're coming out with this 100 day guarantee on ROI. So maybe I think Paul for you and Amit maybe, how are you able to bridge that gap? Tell us a little bit more about what the AFDES are doing in practice because it seems like there's a lot of technical milestones to go from garbage in, garbage out to something that looks like the case studies that you're showing us here today. Thank you.

No, thank you. It's a great question. The way we think about it is really twofold. One is I talked about a little bit during the presentation, but our four deployed engineers truly led by John in the front row here are truly elite. They work with customers on really the high value, I'll call them high value workflows. So things like massive usage outages at a huge bank and the cost per outage is incredible. The volume's not that high but super critical important to that bank. Then we think about how do we actually identify the existing workflows that they have. In some cases, customers want to redesign those entirely. You may have a process like an incident management process if we're talking about ITSM that you've just had for years and you now want to redesign it, optimize it and then identify it. So that's the high volume part and that actually the high volume part drives a lot of the assist consumption and drives that asymmetric scale from a value standpoint that I talked about during the presentation. The high volume use cases are where that comes from. Now the great news is the product team has innovated across the autonomous workforce now which you saw a little bit of today. You'll see a lot more over the next two days. So we now have autonomous workers that you kind of plug in based on the products that you own and they work side by side the human workers. So autonomous workers actually help us drive the high value high volume use cases much faster while the FTEES look at the high value use cases. So we're attacking it from both angles and the receptivity from the customers has just been incredible.

I think the FD model is only for very selected. What we do with FD is to identify very high value use cases as Paul was mentioning but also kind of help customers imagine the business process for those complex use cases and understanding where the integrations are required, what changes you need to make, and then really the product takes over and then plans have been the what you implement and the FDS usually move on. So it's not a continuous FD engagement like many other vendors have typically.

Great. Hi Samana from Jeff. Thank you for spending all this time with us today. It's great to hear from as always. If I could maybe dig into the 30% AI ACV, that implies roughly like $9 billion in 2030 plus or minus. If you think about unpacking that, how much of that is from the portfolio as it exists today versus what you think you have on the roadmap? And I guess the related question in that is how much does the change in the pricing packaging influence that and does that require any changes for the existing install base when they come up for renewal? I know it's a several part question but I'm trying to learn from Alex.

So, a couple things. First and foremost, I want to be clear on the new pricing model. So Prime is basically the same pricing as Now Assist Pro Plus, right? So we're not increasing pricing in our most premier offering. The customers who are all in on Now Assist are not going to expect to see increased pricing. Where the increased pricing comes is when people in lower tiers and lower levels continue to adopt and grow their AI usage by bundling products or by going from standard to foundation, foundation to advanced. And so what I say is that obviously the pricing model is baked into that 30%. But it's not a huge differential from where we are today. It will mean that it's really about penetration. How many more customers are driving upwards to those higher level pricing packages and what I said earlier is that we're expecting a similar penetration trend as we saw from standard to pro which I think is actually quite conservative as you go forward.

I think you kind of answered it already, but just for customers that renew, it sounds like if they're already on Pro or Plus it would be the same pricing, migration no matter what, but it doesn't require any changes. That's all I was kind of curious about.

No. I think this came across, but just in case, there are no options at ServiceNow that are not AI. It's just degrees of the offering and how advanced it becomes with the autonomous Prime, but everything is AI enabled. So there's only one ServiceNow and it's an AI ServiceNow. So I think you probably already knew that, but just in case customers have asked that question, we only have AI.

We are only counting that incremental. We are being very consistent with how we've always been and how we've always treated it. There's AI in everything, but base subscriptions were not counting in that AI revenue. That's why it's only 30% and not 100%.

Hey, it's Tyler Radkkey from City. Thanks for doing this. Continuing on the theme of multi-part questions. I'll try to keep it to two. But Bill for you just on M&A given that's been a huge topic. I guess first clarification, there's no M&A in those 30 and $32 billion targets, but philosophically how do you think about what you have today? Do you need to do similar size or larger deals compared to Armis and Vasa and whatnot? And then Gina, I think this is the first time we've seen two different kind of scenarios for the revenue. As we think about 2030, what why is there two? Can you help us understand the differences between the upside and the base case?

You're lucky I didn't give you the real real upside numbers. That would have really confused you.

I'll let Bill start.

Agreed. And on the question on the 30 to 32, given the uncertainty in the market, we felt that it was prudent to give a range and not just one number. We feel highly confident in that 32, but also are taking into account just market sentiment at the moment and wanted to assure you that there's a real strong glide path and we presented it to you 25% plus CAGR on our growth engines really driving to that 32. But if you're a little hesitant in this market, you want to tie your number to 30, I'm okay with that, too.

And that will actually conclude our Q&A session today and our webcast. We invite all of you to join our executives at Chica for a drinks reception after that and you can ask them any additional questions there. Thank you.

Thank you everybody.

Thank you.