ServiceNow, Inc. ($NOW)

[Audio Gap] I've been at ServiceNow for the last 18 months. As you mentioned, I run the product engineering, operations and product strategy for ServiceNow. Previously, I was at Google Cloud for 6 years. We're on all the platform teams, including a lot of the strategic products and portfolio there. And previous to that, I was at 24 years at Oracle ran various different groups. Before I left, I run the cloud team as well as all the middleware and the platform team over there as well, which was substantial part of Oracle's revenue. Very excited to be at ServiceNow. A lot of things going on, great portfolio, great product plans going forward as well as a great customer base. So really good time to be at ServiceNow.

Yes. Yes. I think we should start off with what was probably the biggest topic for me when I started to the partners and customers out at the Knowledge Conference, and that's AI Control Tower. A lot of traction with it. our checks suggest that customers are deeply concerned about their AI for all. They don't know what these agents are doing, shadow IT, not to mention the cost side of it as well. They need something to bring this all together. You can talk about what's differentiated kill switches, all those pieces. But I think 1 thing before we get into it, that's interesting is that's -- this is a very contested space. Everybody wants that governance and control customers are choosing you guys to do that. So can you describe why it's so differentiated and why you guys are the right platform for customers?

Yes. I think CAI control tower, last year when people started talking about agentic and even when I was at Google, the amount of AI work going on, and I was building a lot of the AI technologies. Is it was clear that customers are going to face a lot of issues managing the AI capabilities, the usage of AI discovering a lot of things which is going inside the enterprise and really making sure secured and governed. So last year, we launched a control tower, One of the reasons behind that is whenever we're talking about agentic to those customers, they were really getting worried about security, governance, compliance, auditability as well as just visibility broadly. And ServiceNow has this product called CMDB as a core foundation of our product portfolio, which where we control -- we manage and monitor all the assets inside an enterprise, any IT asset, hardware, software. For us to do that -- and extending that to AI was very natural. -- agents are software. We are discovering a lot of other things for every IT systems out there anyway. And so we added the agents as a way for extending what we did in CMDB very quickly and easily and then giving them so much monitoring and measurement as well as the cost management associated with that. And that opened up a lot of conversations with our customers. And the reason it's kind of valuable for us as well as why we are differentiated. One is I said, we already had the foundation for managing and monitoring a lot of the systems and AI was very natural. Second is, I think we were connected to so many different systems out there. I mean, ServiceNow is kind of enterprise OS, the way you think about it, connecting various different systems East to West across different systems already, very heterogeneous, very open ecosystem. So for us to be more of a neutral third party where we can now not just track our AI agents, but also discover things coming from any different systems out there, hyperscalers, all the large language or frontier model companies, any AI systems or system of record. And providing a full governance layer on top of that, just like we did for any asset, became -- it's much more differentiated because of that. And the other part, which we have been doing very well is we recently did an acquisition of a company called, we which has a patented technology call access graph. An access graft for nonhuman identity is becoming a very, very critical element for monitoring and managing AI agents because their use cases as well as identity changes so frequently, and you have to do that monitoring real time. They do around 30 billion different permissioning and parameters for nonhuman identities. So we brought that into a control tower, and that suddenly changed the game for a lot of our customers because they can feel more confident about AI systems not taking wrong actions or doing wrong things and giving them the visibility. So tons of stuff going on in our product portfolio -- and that has been really well adopted and differentiated, and we continue to invest and provide that capability with the eye control tower now.

And maybe to double-click a little bit in there. You have this product that's answering a lot of customer needs. Customers are in different parts of their AI journeys. At what point do they kind of hit that they're like, oh, we need to do something that's like AI control is? Is it a cost function. They're like, of a sudden we're spending more than we expect or security the main function where they're like, hey, if we want to push this and put to production, we need to have a way to make sure that it doesn't go role.

Yes. I think it usually starts with when more usage of AI such happening. Like if they want specially a genetic where you have all these business processes and systems connecting various different agents together. And there could be a lot of shadow AI projects happening. And then you have no idea what's going on and who's spending what or who is using what -- and last year, as we said, when the agentic use cases were starting to emerge, a lot of the worry customers had around was just control and visibility. And -- but AI control tower conversation started was just peace of mind. -- where you don't have this sprawl of AI without IT knowing what's going on, who's using what, how much they're spending. So it really accelerates the AI adoption by having this as a central control plane and heterogeneous because a lot of the providers today, they do pieces of the technology for their own visibility. They might connect into a few other things, but the end-to-end discovery mechanism and the whole monitoring and visibility mechanism. It doesn't exist in the market other than AI control tower. So I think it just removes that barrier for adoption of AI because now IT team members as well as CIOs and CISOs. So I feel a little more confident they can get going with AI without just random stuff happening inside the enterprise. And with Enterprise anything -- I mean, you've seen a lot of examples. I mean you saw the recent pocket OS and others, where if you don't have visibility and control, things can go completely wrong, and you need to have some guardrails, and this is what it allows them to do.

Yes.Yes, that is a gating factor they can deploy more. if we go to something you already touched on, which is acquisition, Moveworks, Armis, Veza, One thing that I noticed is I think investors have been kind of trying to digest what it means, why they're strategic. When I talk to the customers and partners, they seem really clear about what it brings together why it's such a high value. There's been a lot of companies to do acquisitions. This was 1 of the ones where they seem to be paying attention a lot and kind of see the vision that you have. So what do you think that they're getting that maybe investors kind of haven't wrap their mind around. .

No, I think we've been very disciplined and thoughtful about what we acquire. And over the last -- as I mentioned, with the AI growth, there's opportunity to really innovate faster, have -- get a lot more domain expertise in some of the areas. And one of the biggest areas we've seen a need for and is around security and governance, right? So the acquisition like Veza and Armis fit into this idea of how do you really monitor and manage the proliferation of AI systems but also devices. Armis provides the ability to do OT, IoT as well as medical devices, vulnerability management and exposure management. Today, our business and security space is $1 billion plus, which happened last quarter -- Q3 of last year. And that was an organic business we built over a few years, and it continues to grow very, very fast. What we saw from customers was that you already manage all the IT assets. Why are you not able to now do also security and asset and vulnerability management for non-IT such like OT, IoT and specifically medical devices. So Armis fit in very well with that kind of mindset, right? We -- allows us to extend to any kind of devices out there. and also extends what we have in CMDB and augment that with millions, billions of devices into that system as well. So that, as a growth with physical AI and other things come up, we see a lot of value for ServiceNow to provide that exposure management, vulnerability management on top of post-breach what we were doing already. Veza is a little much smaller than Armis is, but this idea of patented access graph technology, again, augments everything we do for nonhuman identities, just like we use for human identity. The identity governance specifically becomes very critical. So our security business and the product portfolio road map got accelerated by this acquisition. We're bringing a lot of good domain experts in this area who are well-versed and have been doing this for quite a few years and made our story much more powerful to customers as they think about AI adoption as well. We continue to do the integration with third-party systems, right? So this is not like it's replacing a lot of things we were doing before -- not doing before, like we integrated into a cloud strike for endpoint, we integrate with Palo Alto and network-related stuff and get a lot of signals, but the whole exposure and vulnerability management with the idea post breach. And today, CISO is our second largest buying center for ServiceNow already, right? IT, CIOs were #1 and CISOs are #2. So it was natural for us to keep on expanding into these new areas, no doubt increase our TAM, but also accelerate our road map and bring a lot of this expertise into our product. And it has resonated very well with our customers. Our pipeline since the acquisition has grown for security portfolio very, very well. And we continue to see a lot of good traction with the product being integrated. Similarly, the Moveworks acquisition was very much led with this idea of conversation first, AI native, experience layer on top of ServiceNow. We had a very -- we have a great set of technologies today for doing actioning, fulfilling a task. On the request, we wanted a very AI native experience, and Moveworks brings that. So that's now integrated into our employee EC Pro product we had, the whole employee center and bringing this thing together as employee works. In the first quarter, we announced, I think, in the Q1 results, we did more as a combined product than what Moveworks has done by independent last year. So we saw acceleration because of the road map being integrated, able to deliver a lot of new AI native capabilities on our product portfolio, and it's becoming our user experience across all of the offerings we have today. right? So we've been very thoughtful about how we bring all this stuff together to really differentiate, accelerate the road map as well as continue to innovate in all those areas.

Yes. One of the debates out there obviously has been if AI is going to generate all this value, where is that value going to accrue? And where is most of that getting to get captured? I think you guys have been pretty explicit that the is not necessarily the. You guys have the context engine, the graph oraf anchor the CMDB layered with Veza, Armis is included in there, too. You guys have about $100 billion annual workflow, million excuse me, 7 trillion transactions and you guys put it pretty simply at your Analyst Day, there's no competitor that can do that, right? Help us understand in layman what the technical reasons are that others can't do that. What do you provide for your platform that enables that differentiation?

Yes. I think you, remember, I mean, we have been doing this work for our customers for 20-plus years. The amount of data we collect, as we said, 100 billion workflows we run, plus 7 trillion transactions happen on a platform, which is growing at 20-plus percent, 25%. So the Metadata we create, the context we generate around why a decision was made in the business process. Why an exception was done, who made those exceptions. Those are not available in any documents, which an LLM can read. These are all information we're collecting as you run these transactions. That's in a context engine, which is distributed across multiple areas, and Metadata is very hard for people to understand what to do with it, if it's not done by ServiceNow as a part of our workflow, right? So we build that the CMDB, as I said, is really kind of the enterprise database today, right, the IT system of recked and used by all the Fortune 500 and pretty large number of Fortune 2000 companies. So that has become the kind of the underlying ways for large enterprises to operate on ServiceNow. All the systems today, they operate the operational part of it in terms of executing the workflows, business processes around for IT employees for security, what we do for customer service, FSM and other things like that. is pretty broad and deep for many, many years. That information allows us to now take some advantage of AI technologies like LLM, which is probably around 8% to 10% of IP in a stack, but the rest we build and operate over this information we have collected plus the efficiency we have gained. And everything doesn't require need to be an LLM driven, right? There's a lot of deterministic things we do, which requires a much more understanding of what the business process is and not have to say is spending money on tokens, but you can still operate that very efficiently. So we bring this deterministic mindset to a very probabilistic systems like LLM and still give you a guaranteed outcome because you can't make mistakes in enterprise systems. If you get the wrong answer, you instead of onboarding an employee, you deboard an employee or you have the wrong financial results, it's not going to be very valuable to you as an enterprise. So the context we have gotten, the way we run the systems, the 20 years of experience plus having up stack being very AI native now, right? So we're getting the same advantages as AI companies have in terms of technology, but then you're wrapping and providing a lot of insights into these things. Plus, we're also working in a very open ecosystem, right? Customers start in many different ways. And you can't just say that there's only one way to operate today a business. So having that opportunity to integrate into a different system providers, like a system of records, integrate into any kind of hyperscaler, be able to also run -- or use any large language model, any data provider, any tool to build your workflows custom on top of our ServiceNow platform and have the whole governance and security layer. One of the biggest issues you run into when you're building the systems out there or running a business application or business process is the security and the governance. I would say most of the time, when we build our products, we have to be compliant. We spend probably 50% to 60% of our cost of engineering and R&D is around that whole security, compliance and things like that, which are very hard to replicate from scratch. It takes years, years of hardening. The life cycle maintenance around that is pretty long, and customers need a place to call when things go wrong as well. So that's all the things we provide out of the box, working together with our customers. proven and operating at scale. And that's really what -- and the partner ecosystem associated with that, right? And the go-to-market muscle we have for years and years of building that relationship with our customers. So those things all add up to really be able to provide that comfort for our customers as well as differentiated and innovative products long term.

Yes, that 50% to 60% statistics, pretty intesting...

Yes. People forget that people think build is the only thing you have to do -- and build is itself is 20%. The rest is all around everything you have to do around it. .

Yes, that sort of takes to be in the enterprise, right? Yes. boring, but painful but important to if we go up 1 level with this AI topic, Bill broke the news on the Q1 call that the now assist target is being raised from $1 billion to $1.5 billion, 50%. I think in Q1, you guys were already around 750. You got growth in customer spending $1 million plus on now as is grown 130% year-over-year. That's all real acceleration. And importantly, I think in kind of emphasized the methodology hasn't changed, even as if you guys kind of repackage some of the pricing. What's given you confidence in that revised figure as you move away from this kind of sidecar AI concept to help fuel that growth and maybe even kind of go beyond that?

No, no, I think we've been monitoring, of course, our pipeline as well as a customer interest and adoption associated with that and the renewal as well. So as Bill mentioned in Q1, we'll be going from $1 billion ACV to a $1.5 billion target this year. Our plan was $1 billion. We feel very confident in $1.5 billion ACV on that one. It's because I think, one, with the genetic use cases and some of these things go in production and people using it, we're seeing the volume of usage go up very fast. We're seeing a lot of interest to do multiple use cases than what they started with a employed -- with enterprises. So that gives us the ability to say, you know what, we will end up using up whatever entitlements the customers had previously. And when they renew, they will renew at a higher longer -- broader entitlements or they might buy us is path early enough, right? So that is one vector, which we see happening. Second thing, as we think about some of the things we're adding around our product portfolio. Our product portfolio has grown drastically, right? So we've added things like AI Control Tower, we're doing with employee works in terms of amount of requests coming into a system, which will end up burning down a lot of and the request volume will continue to go up. what we're doing with security portfolio is pulling a lot of this conversation for monitoring and managing a lot of AI systems as well. So suddenly, we are in more conversations about AI than maybe previously we were -- so I expect our renewal sizes to be higher as well as we're seeing expansion into a lot of new cross-sell areas, which we didn't have. We simplified the packaging as well. So we're bringing AI across all of our product portfolio, and that will drive a lot more conversations with our customers going forward as well. So I think in general, just the demand is there. We're seeing a lot of adoption. We're seeing a lot of interest. And then I think the innovation cycle continues to grow, and we see a lot more capabilities now continue to be introduced into our portfolio, which will help our customers with their AI use cases as well.

Yes. And touching on the some of the pricing and repackaging that you're doing. I know it's early days and it's going to get phased over time. But what does the customer reception been to that? Do they understand this new schema? Does it help them with that adoption friction?

Yes. For those who are not aware, what we did, we had a product we have called Pro Plus, which is the now assist the premium SKU, which was AI-driven, which we've been in the market for a couple of years, and that has done very well, the pricing structure we had, which was a hybrid pricing structure licensing base, but combination of seed, but also consumption, right? And that has been in the market for some time, and we've seen customers understand that hybrid pricing structure, gives them predictability as well as flexibility. They're not going to get sticker shock or any kind of billing shock, but also they know what they're using and that they're using it and seeing value, they will renew at a hard so they have predictability in that regard. But we also account for revenue upfront as well as we use that just like any subscription. So the model has worked very well. So what we decided was based on what we're seeing from our customers to bring that same kind of structure to all of our SKUs, not just to the premium SKU. And the idea is that every SKU will have some AI capabilities tiered by different functionality level. The top tier always remains most differentiated. The bottom tier will have some basic capabilities, but customers can start with AI as they choose to, and they can use a consumption model to see value and use it while that we still have a licensing kind of mechanism. And given how successful we've been with -- now Assist and the Pro plus kind of combination, we are now introducing that across the product portfolio. and also removing this nickel and diming where we structure where you have to buy every feature separately. We're bringing it together into much more structured across ServiceNow product portfolio, giving them the AI control tower capabilities, the experience layer, data connectivity in this kind of a tiered structure. And every customer I've spoken to so far really like it because this gives them the ability to take AI if they choose to in all the SKUs, but they do it by consumption, so they only use it if they choose to, and they will have associated entitlement and simplifies our go-to-market -- our sales teams can really pitch this very simply and easily. And customers can move up the tiers as they choose to based on functionality they like as well as the adoption they see value from. So no, I think it's very well liked so far. The hybrid model has worked already. So we're not experimenting with new pricing structure. It's just now adding the same pricing structure, which has worked into all of our SKUs and simplifying our go-to-market as well as customer conversation. So we're very excited about how this has evolved. We announced it at knowledge, which was just a few weeks ago, which is our large customer event. And everybody seems to be feeling comfortable with the way we are thinking about this and how we will go. And there's no force migration. They can do this at renewal. -- and the Pro Plus and the Prime, which is the new version, the top version of this new structure is equivalent. So there's no pricing change for those customers. The rest will go through the license changes whenever they renew if they choose to -- so customers like that message. And so far, we feel good about where we are.

If we touch on the less positive aspects of AI, I think it's created a lot of confusion out there. I think you guys said in your last earnings call that customers don't know what to do, they're somewhat confused. We kind of hear the same thing. I've heard more people than usual talk about terminal value risk. So it's getting out into the ether a little bit and probably creating a little bit of reluctance, right? So when we think about that what does that look like? How does that impacting decision-making from your perspective? And maybe the other important piece is, is this an issue that's going to persist for 3, 4, 5 quarters? Is it something that will be resolved in a few quarters? How should we frame it that way?

I guess maybe if I understand your question. One, there's definitely AI -- there's a volume of message in AI in the market is large. The customers don't know in many cases, what is real, what is not real, what to go and where to do things to work with and how to get going, right? That's why I think our goal always has been about talking to customers about how to get started and where to get started, right? Some of the -- most of the times, the conversation is the use cases, what makes sense. And we are very prescriptive now than maybe a year ago about which use cases for a particular customer makes sense. We provide them POC capabilities. We provide them the help engineering capabilities as well as needed in a short amount of time, but get them the right use cases identified so they can get going. And that has removed a lot of the barriers for us at least. Having said that, I think customers also are cautious about security and risk and the whole visibility as we talked about earlier. So that's why I think the investment we're making with the air control tower again residents, even if they don't might want to use anything from AI from us, they can still use the eye control tower to have the central control plane. So it's we have a conversation even if they're thinking about something else from an AI perspective. Similarly, the security around devices are all becoming very real because there is a proliferation of physical and other things coming up. So we now certainly have many, than probably what we had before, right? If you look at where we were at 2 years ago to where we are now, the volume of conversations we can have with the customer, we can start from so many different angles. The employee works is a very good example as well. any kind of conversational interface you want to provide to customers to use things across their enterprise, find information, ask for task or getting work done, again, we provide an option. Same thing with what we're doing with AI-driven omni-channel intake with CSM or FSM. So I think that is really opening up a lot more senior level conversations for ServiceNow. IT always loved us. and they continue to love and invest in what we're doing going beyond with autonomous AI and other things like that. Same thing with CISO. So suddenly, you have a lot of supporting structures inside the company for ServiceNow. And AI, irrespective whether you use AI or not from us, there's always a product available to you, right? And the AI part, I mean, again, we'll see in many years. I don't think you use large language models or frontier models for everything. -- because very costly over time probably or some things you don't even make sense to use. We give you this flexibility and the ability to really do the right thing to run and operate a business because eventually you want outcome, cost savings and a lot more efficiency. And that's why we are driving towards. So the idea of autonomous AI and AI specialists is that, that I'm giving you an agent autonomous worker who will replace a human worker at a much cheaper alternative but guaranteed outcomes Instead of just having this thing do something, which might not be really productive and they have to redo every time, right? So we're trying to improve that situation with outcome-driven mindset at least.

Let's follow that thought of the LMs out there. We've seen a little bit of a leapfrogging right. One day, ChatGPT is going to take over everything then it's Gemini and its cloud, who knows what will be next. What is your view on that? I mean do you expect that cycle of the leapfrogging to continue? And then from the customer perspective, are they dedicated themselves to 1 or the other? Or they just want whatever gets the job done the best.

Yes. I think it's still early days. I mean you're right. I mean there is a huge amount of leapfrogging as well as, I would say, equivalence of commoditization. So there's a lot of -- between the large language models. And we work with all of them and we work with a lot of open source models as well. We try to be more effective about what will give us the right result, irrespective of which one is the model. I think customers same thing, right? Underneath, as we said, we also have a large language model in our stack. Customers get to choose, but a lot of time, they just take the default. Most of them don't care. A lot of them wanted to get a guaranted outcome as well as secured environment. As long as we can guarantee that, we can choose and change the models underneath the covers, they would not even know or care most of the time. There's some who will because they might have standardized on 1 or something like that or they might have a license or committed agreement, which they might want to burn down. They've been trying to do -- large any want to trying to do something similar to what clouds do. Where you commit model, you can burn down different of your AR usage. But over time, I think there will be a lot more commoditization going on there as well. They will be tiering. I would expect in some of the use cases you want to use an older or cheaper or even nondifferentiated model. In some very complicated use cases, you might use the most latest one, but over time, it just becomes like chips. I mean you don't really care what chip you are using in your application. I think it will be the same thing in this infrastructure layer. It's going to be cost-driven less differentiation. And how many times you want to keep on testing them. See what happens with every version you have to do prompt engineering again, you have to redo your work. And that's the problem I think a lot of customers are realizing when they say you want to build something completely brand new on a large language model. because every time the new version comes out, you're to retest. It's not backward compatible. It's not feature forward-look feature is not future-proofed. So a lot of those issues we have to take away and that's what we're trying to do with our software stack is to remove that barrier of usage without customers having to deal with it. We deal with it for them. We have started out enough then we give you the value and outcome or a solution on top of it. So I expect a lot of these things to change over time. And many of these model companies will start having to figure out how to solve some of those problems.

Yes. Let's look at a few years. You guys laid out the $30 billion to $32 billion 2030 subscription revenue target. A lot of pieces in that, but what people noticed was it's like high teens 20%-plus CAGR between now and then, AI, 30% of ACV at that same time frame. This last quarter, you guys kind of -- if you think about the CRPO kind of growth curve, it's kind of in that high teens area as well. So it kind of implies that you guys are going to be able to sustain that level of growth, maybe even inch up a little bit between now and then. Help us kind of frame some of those components that are then you guys have talked about it and what gives you guys confidence? I mean, Bill, obviously, sounded very confident about it, but help us maybe kind of get closer to where he is maybe?

Yes. I think at Financial Analyst Day, we wanted to lay out our plan for -- until 2030. And I think there are base cases, and I think there's a lot of upside cases, right? So I think Bill is definitely -- and we all see there's a huge amount of headroom available for us to grow. What we already have, and we want to make sure that there's a clarity in terms of where we will definitely be there. The $30 billion to $32 billion is our base case, and we'll continue to focus on accelerating more of that going forward. But the reason we feel very confident, one, I think our innovation cycle, the product we're building is resonating with our customers. We are still the proven and most adopted enterprise kind of workflow as well as the whole idea of bringing AI workflow data and security into one platform. I don't see anybody does that today end-to-end, right? So we give you that confidence level for our enterprise customers. And also unlocking a lot of new things for us. I think if you look at the growth engines we have, security and risk growing at a very significant rate. As we said, we kind of crossed $1 billion, and we'll talk about where it's going, but it is 1 of the fast-growing areas and then the road map has been accelerated. Second, if you look at data platform, data and analytics, the workflow data fabric. We talked about Raptor DP, which is our postgress TAP database, $100 million ACV in less than a year, right? So it's like a very fast acceleration path. And this is just untapped market. It's all available to us. Then you add the work for Data Fabric, the connectivity with the zero-copy adapter, the data analytics and the data platform we're building out. That's one of the fastest-growing areas, and we talked about a potentially very fast $1 billion business for us, right? Similarly, if you look at what CRM is, customer service, 1 of the fastest business to $1 billion plus. We're on target for 2 I think what we shared at the financial analyst days. So that is getting -- continue to accelerate. We've done a lot of modern work with AI, voice and capabilities of our SFM. CPQ acquisition of Logic is really driving a lot more conversations with the customers, customers like NVIDIA and others who are using that for CPQ today. So we have a large growth engines which are all substantially invested in innovative, differentiated, while we have a very good base platform, right? It's completely AI driven, but also it's very deterministic in terms of it can get the work done. Guaranteeing an outcome. And then our -- the domain expertise in IT, HR, CRM security allows us to really go into a lot of those conversations and give us the confidence that we can really accelerate growth in each of these areas. And all these technologies are pulling in more and more of our products, whether you look at it with ITOM, ITAM, ITSM, what we're doing with portfolio management, what we have capabilities in our enterprise architecture, I mean it's just a source to pay. So we have a very substantial capability, but it's on this 1 platform. So you don't have to keep on replicating every time you want to add new features. That's why we're able to innovate faster, which is very hard for many other vendors out there where they have to keep on rebuilding a lot of things because they're not built on the same platform. So that's why we feel very bullish about the headroom we have in front of us, and the pipeline and the customer traction is proving it out so far. And I think our base case is laid out, which we feel very confident and I think Bill mentioned that fab that he has a lot more expectation and I think we all know we will be there, so.

Bill seems like he has very high expectations. Very quickly, just on CRM, you hit on it a little bit. I think the quote you guys had was going to blow through $2 billion. When you're seeing success with that product, what does that look like? Is -- are you displacing other CRM systems out there? Or are you kind of coming in, in areas where there's a little bit of greenfield and growing from there, look like? Because a lot of enterprises have a pretty robust CRM solutions?

I think CRM is a broad market. So just to be very clear, we are -- the space which we have been seeing a lot of the investment we've been doing and seeing growth is customer service. which is very logical for ServiceNow, right? Because if you're doing case management, we understand incidents, we understand issues and how we resolve it. And that has driven the early growth of CRM business for us. associated with that field service management in terms of how you operate all the people who want to help you. Then you add layer on top of that all these complex orchestration work, like CPQ is a very complex orchestration. It's not just asking for information. You're going to do something, which is, again, very much action-oriented like ServiceNow has been always, Same thing with some of the things we're doing around distribution like order management. So the portfolio we built is in areas where we know we have expertise, differentiation and the credibility and the capability building the platform. And that's why we feel very confident blowing $2 billion, as you heard at financial it's a pretty large number already in a very short amount of time. And we are displacing people in some cases. I think in customer service, there are many incumbents. But I think a lot of what's happening in customer service, given that most of customer service products out there, customer has been very fragmented and very old architecture. So people are rethinking voice, for example. They're rethinking how omnichannel intake happens. People are not just calling people on the phone for support. They're doing it in multiple mediums. So for us to build that in a platform because AI platform is what is supporting our CSM product is very quite quick. And today, we have customers who now are using this multichannel way of interacting with our systems and replacing what they had or sometimes consolidating because the opportunity is to really modernize, make it more efficient and get into this technology stack, which allows them to get into many more areas for customers to do much more efficiently. So that's where we're displacing or interoperating in many cases. And it's just a modern stack and proven capability, allowing a lot of people who are running CSM, we're also running the IT systems. So they're used to our platform for them now extending that into CSM, become much more natural because you're not redeploying a completely new stack. It's very easy for them to now add CSM from us or CPQ and other things like that, given it's built on the same platform.

Yes. Very insightful. -- thank you for coming here and giving us your wisdom. I think we'll all be watching service at very closely.

Thank you all for your interest and support as well. Thank you.