ServiceNow, Inc. (NOW) Earnings Call Transcript & Summary

Yes, we're seeing a very significant growth in what we call assist consumption, particularly as customers deploy more and more AI agents, which is what you would expect, because AI agents are delivering breakout value for those customers. And so I think the thing that we're managing is when a customer has to come back and say, "I need more assist, I need to buy more from you because I'm out of AI capacity," we should have a really good business case for them on what they're getting from those assists. So that's what we're seeing by and large. And when we find cases where a customer is saying, "Man, I'm burning through my assists faster than I like and I don't think those things are the highest-value tasks," we have to work with them and say, okay, let's make sure we're automating the things that are creating business value for you more than those things. And so sometimes it's about getting customers to pick the right places to deploy their AI, because they burn a bunch of assists on stuff they consider low value, that conversation about buying more consumption isn't a great one. But if they're doing it on things that are helping them reduce service desk agents or helping them improve their time to renegotiate a supplier contract or reduce their supplier risk, it's a much more valuable conversation for the customer and for us.

And so it's a bit of an education as we pivot from this traditional seat-based model to a hybrid model. And I want to go back to that in a second, but you said something interesting. In other words, implied in some of your earlier answer, you don't necessarily see ServiceNow moving to a full consumption model at some point in the future.

We'll stay really close to our customers on this. There are customers who have become used to models like that in hyperscalers. There's a model there. It's kind of commit to consume, right? You say, "I'm going to buy this amount from you," and then you can burn up to that amount. To the extent our customers say, "Hey, I want to start doing something like that with ServiceNow," I can imagine us being -- moving into that kind of a model. But it really needs to be something that we base on working closely with our customers. And this was the thing about seat-based pricing in the beginning, is it was a lot about customers don't want to just buy a bunch of [ assists ] at that time. Nobody -- it was very early on, this was 3 years ago, I think, and nobody knew very much about AI, generative AI at all at that point, not customers, not suppliers. And so there's a lot of uncertainty. As we're getting more and more certainty and seeing more and more of it play out, I think people are getting more comfortable by saying, "Hey, maybe I just want to buy a bunch of stuff." And with our platform, we can cover so many use cases. I mean I talked about from IT all the way to contract management running through customers and supply chains and everything that our customers have. So thinking about just saying, "Look, I don't want to have to contract with you on this product and that product and that product. I just want to buy your platform and go deploy it in all the places that make sense for me" is something that more and more of our big customers are coming back and saying is like, "Look, let's take the friction out of this. Let's just let me put it wherever I want. And when I use it, then you're going to create value and I'll pay you for that value."

So I guess maybe to put a bow on this, the big concern is kind of P times Q and fewer seats in the future is bad. When you're talking to these customers and they're seeing the ROI, are you seeing any friction on, "I love ServiceNow, but in the future, I feel like we're going to use less ServiceNow?" Or is it more of like, no, no, this is like it's changing the way we get work done. We see ACV growing on a customer basis...

It's absolutely the latter. Our customers are getting from the top of their organization, "We need AI. We need AI use cases. This should be" -- when I talk to anyone in a core business department, they're saying, my CEO is telling me to do my job as well or better with fewer people. How do I do that? And everyone tells me AI is the answer. So they're looking for those kinds of answers. And when they're able to use our AI to create some of those business value outcomes, it helps them be a hero in their organization. It helps them succeed in their job. And so that's something that is resonating really well. They're not saying I want to use less ServiceNow. They're seeing ServiceNow as the stuff we knew and this new AI capability, and they're turning to their peers and saying, "ServiceNow has a product in your department too." And by and large, a lot of the departmental work is very similar. Hey, how do we serve employees? How do we get our core work done? And how do we eliminate all that manual work that used to result from all the underlying fragmented technology landscape and all the manual processes that sit in between?

So it's a dovetail into the orchestration layer. I've long thought as organizations are building and deploying AI, as we are internally at RBC, just understanding what all these agents are doing is a critical issue, I think, for a lot of organizations. And you guys, we think you're in a really unique spot to be an orchestration agent across an organization, that agent of agents. Talk about how customers are thinking about that? Is that something that they think about, "I need somebody to manage all this?" And if not ServiceNow, it doesn't -- I don't know who else is in a good spot to do that.

Yes, there's a couple of dimensions to it. One is similar to what used to be the low code concern. It's like, hey, if I give citizen developers the ability to build apps and I have applications all over the place. So people were really worried about agents all over the place and they have no idea what those agents are doing. And so there's kind of a management capability to that and then a risk capability. And so we built something called the AI Control Tower. The AI Control Tower can monitor not just our agents, but third-party agents, and understand what they're doing, how often they're running, what kind of risky activities they might be executing, and start to look at the business value that they're creating by customers sort of assigning a form of value to the execution of that agent. And so I think we are in a really unique position to be able to provide that because of our place in the stack. We sit and complement, particularly in core business departments, we sit and complement a lot of the systems of record. And so they're going to have agents, and we believe it will be a heterogeneous agent world. And so you're going to want one place to really look and monitor how all those agents are running. And that's what our agent Control Tower is doing. It's one of our products that's getting the most interest from customers and growing in terms of interest faster than probably anything else.

So I can imagine as a knowledge worker, you're pulling information from multiple systems to get your job done. So is there an example that you have of somebody using your orchestration, your Control Tower to ingest data from one system, maybe create a PowerPoint in Microsoft? Are those examples happening? Or are we not to like the agent protocol perspective yet where that's...

Yes. So I think that I would say that's going to be at the agent level rather than the Control Tower level. So Control Tower will monitor those agents and what they're doing. But we have -- in our core products, we build out-of-the-box agents. So a lot of those out-of-the-box agents are doing things like resolution recommendations for agents. That's one of our most common agents that will run in IT. And that resolution agent will pull data from ServiceNow, but it will also pull data from other systems to make a recommendation to the human agent for how something should be resolved. And so that's an example back in that service desk layer, how I said we're making people more productive, that's a great example. Because now the human agent doesn't have to go start from ground zero and do the work. They can look at a resolution plan that was recommended for them. So that's one example of that. But we also see customers building custom agents with our Studio. And our Studio makes it incredibly easy for our customers to build an AI agent to do really anything. Because in our Agent Studio, all of the automations that exist in ServiceNow today are available as tools to a new agent. So if you have a flow that does approvals for time off requests or for expense reports, all of those are available for an agent to call on. All you have to do is kind of describe what the agent is supposed to do and then plug it into our orchestrator. And so then when a request comes in and says, "Hey, I want to request time off," the orchestrator says, do I have an agent that seems like the time off agent? And there's an agent that's been described as "I'm the agent that finds time off balance and schedules time off." So it was like, oh, yes, okay, do your thing. And it can just call in to our out-of-the-box integration to the HCM, Workday or SuccessFactors or whatever, and say, "Get me the PTO balance," sends it back through a conversational interface to that employee saying, "You have 8 days of PTO left. Do you want me to schedule you some?" "Yes, I do." Go back to the same agent and call a different spoke -- or a different -- we call integration spokes, that actually schedules the PO. So this kind of ability to -- the layers of our architecture and our AI platform for agentic AI that allow you to connect to all those systems is pretty unique in the market. And nobody has really built out from the [ action ] layer to the ability to create and modify our out-of-the-box, to the Control Tower layer and the ability to orchestrate that Control Tower, much less the ability from end user engagement all the way down through action.

Once Moveworks is completed and integrated, what are you most excited for? What are customers most excited from that they weren't previously getting from ServiceNow?

Yes, we have a lot of customers that are using both our products today. And we can see -- I've talked to some of those customers and they have Moveworks plugged into Teams or Slack or whoever they are, and their employees use it and they have conversational engagement through there. They do some search capabilities. And then they also come in to our employee portal, that the customer has deployed. We call it Employee Center. They'll come into our employee portal and be able to use ServiceNow's virtual agent powered by our AI. So a lot of customers are using both of those in 2 places. And they can call in to the back-end systems. Our virtual agent calls much -- has much deeper knowledge of ServiceNow and calls deeper into ServiceNow today. What I'm excited about is the ability to bring those 2 really powerful engagement capabilities together, let our thousands of customers who have our portal seamlessly plug in Moveworks the way they have into Teams and Slack and other interfaces, to provide a great, incredible conversational experience. I'm excited about the ability to make it possible for Moveworks to call deeper into ServiceNow and drive even more powerful automations from within the ServiceNow platform. And of course, there's customers that are using one, the other or neither. And I think the combination is just going to create an incredibly powerful way for us to accelerate sales of our AI technologies and for more customers to quickly get these conversational and search capabilities set up. And all of that will plug into the incredible power of that agentic AI architecture I just described, our orchestrator, our Studio, our Control Tower. So our front end and the chat and the search is just about to get better than it's ever been before, and we're going to couple that even more deeply with the power of that agentic platform.

That's great. As we -- you guys have delivered strong results this year, and it's obviously, it's like the new normal is uncertainty. The government is back open. But when you talk to customers -- a question that I always get is like, how do customers think about year-end budget? Do they -- is there a budget flush? Have we historically seen that? I can't imagine Gina is necessarily assuming anything abnormal there. But like from a customer lens, how do they think about kind of year-end spending?

Yes. I haven't -- to be honest, I'm not in, today, in a lot of conversations with customers where they're like, "Hey, I have a ton of money I just want -- I want to flush out." What I'm seeing a lot more is the conversations are all centered around AI. The buying conversation is, "I have money to spend on AI." And so I don't know whether that's a budget flush or it's just like, "Hey, we took all our money and we moved it out of this box and into AI." But spending money on AI, spending money on automation, spending money on driving costs down and driving efficiencies up, whether it's with AI or just sort of that deterministic workflow is a priority for customers. They're dealing with macro uncertainty. They're dealing -- and now nobody really knows what's going on in the background because we had a data blackout. But that's been a trend that's been growing, is macro uncertainty, tariffs, sort of erosion in the consumers that's kind of the low end of the market. And so customers are saying, "I need to cut costs." And so that's what I really see, is people saying, "I'm willing to spend money to save money," and coming in with a hypothesis that AI is the way to do that.

So effectively, what you're saying is you think there's more defined AI budgets today than there was even a year or 2 ago. And that certainly can play -- yes.

Absolutely. And look, a lot of people are saying, "Oh, when is the value of AI going to show up?" And I can tell you, we're seeing it with customers today and it is showing up. And so I believe that will be an incredible accelerator because no matter what part of a customer's organization they start using our AI in, the transferability of that automation into another department is so significant that I believe it will really help us accelerate AI adoption.

I love asking a question, like a moonshot question. And you, with your seat and your lens, I'm sure you have some interesting thoughts on where we go in the future. I guess like if we're sitting here 3 years, 5 years from now, like what is like a moonshot element that AI could bring to ServiceNow? I mean, what could it mean that is transformative that we don't even see today?

Yes. So I guess what I -- I think of this like, can you imagine using a product today that didn't have Internet connectivity?, Like every app on our phone, every app on our desktop, they all kind of depend on Internet connectivity. Even the things you'd think you can work on for a long time in isolation, like Word or PowerPoint or something like that, still you need to get to the Internet because that's where everything is stored and shared and that's where a lot of the suggestions that it will have come from. I think that's the way we need to think, is AI is going to be in the core of every single one of our products. Like you won't buy a product from ServiceNow that doesn't have AI in it. Today it's an add-on and it's -- we create more value, so we monetize it that way. As monetization models evolve, AI will be in everything we do. And so is that a moonshot? I don't know. It may seem kind of obvious today. But the amount of effort we're putting into making that happen is very significant because it impacts everything from feature design -- our designers can't design a non-AI and an AI version. They just have to design an AI version. So it impacts everything about how we build products, how we price products, to how we deploy products. And that's a really significant lift, but one that we are absolutely focused on, from our CEO to our COO and Chief Product Officer, to me and my peers and everyone who works for us, is we are figuring out how AI can be in the fabric of every single thing we do.

Yes. Well, and we need another half hour. We're barely scratching the surface here, Josh. So I -- unfortunately, we don't have any time for questions, but I know you're here and meeting with investors today. So from all of us at RBC, Josh, thank you, and Alex, thanks for coming.

Thank you. Appreciate being here.