PagerDuty, Inc. ($PD)

Hi, everybody. My name is Koji Ikeda. I am one of the software analysts here at Bank of America on the research side. I am thrilled to have Jennifer Tejada, Executive Chair.

Yes.

It's the right title now, John Duo.

DiLullo.

DiLullo, who is the new CEO of PagerDuty. Thanks so much for doing this. Super appreciate it. So there is a CEO succession plan going on here. .

I guess first question, maybe to Jen, why did you feel now is the right time to make this succession? And then, John, I'm going to ask you a couple of questions.

Yes. Thank you for the question. Well, now is the right time because of really two things. One, we felt that we've stabilized the retention -- some of the retention challenges that we've seen in the business. And we're starting to see growth levers accelerate. So whether you look at 5 consecutive quarters of more than 600 new logos, starting to see some of the green shoots that we're seeing through our pricing transition going from a seat-based pricing model to a platform and usage-based pricing model, things in the business were starting to really point in a positive direction. And that gave the Board and I comfort provided we could find a great leader that we felt would be the right person to lead the company through its next chapter of growth. And John is that leader and have become even more convicted about that in being able to work with them over the last several weeks. He's somebody that comes to us with a deep expertise and academic background in engineering that has worked in infrastructure and cyber software for many, many years. He's the multi-time CEO, multi-time Section 16 officer, has worked in private and public companies, high growth, also scale and just had a really unique mix of skills and experiences that we felt would really benefit the company and customers and shareholders. So those two things coming together, along with the fact that we've been running kind of a thoughtful succession planning process for some time made it actually a pretty easy decision. He's also a nice guy. Like we've enjoyed -- I mean we're enjoying working together. So that's been great.

Awesome. Thanks for that. And John, what did you see? What sparked your interest? Tell us a little bit about your journey that led you to the stage today.

Well, I will tell you that I've been a customer before and I've enjoyed using the product. I've also been a partner. And you could have pushed me over with a feather when I got the call. I don't believe that this was really happening. Jen is an iconic figure in the valley here, and as they've done some amazing things with the company. But not only did I know the company, I know the space that it's in. And I'm super excited about everything that's happening in software right now in software development and all the new code that's being deployed, all the all the new capabilities. And along with that volume of code and volume of new capabilities comes a lot of surface area and a lot of opportunity for risk in the operation of the infrastructure. And I got a feeling that the asset was kind of underappreciated. I think that's the vibe that I had because I see PagerDuty as a company that's going to help enterprises really liberate all of the capabilities that they're going to be able to enjoy with AI and it didn't feel like that was necessarily being communicated in the price of the -- in the way investors were valuing the company. So I couldn't have been happier. Since I've joined, I've gotten to know the people, I'm blown away by the capability of the people. I can't believe the level of connections we have within large enterprises and how customers value us, how customers are rooting for us. It's been the most exciting 3 weeks and 2 days in my entire career.

Good segue into my next question. I was going to start with how many days have been on the job.

There has been a holiday in there. So it's -- we'll say, 16 days.

16. Okay. 16 days. You've done a lot. You've done your due diligence. You've been here for 16 days. How would you categorize high level, some of your most near-term initiatives that you want to act on, these are the things I want to do. And then, how are you thinking about the long term?

Well, the short term is all about listening right now, and that's what I'm focused on is trying to understand exactly how our customers use the product, what are their care about? What are the elements of the business and product features that they care the most about. And so that is super important. But I'm also listening to the organization. I'm also listening to our partners. I'm really getting out and exploring everything. My #1 focus is working with Jen on the smooth transition. And so she has kept me incredibly busy. And I can't -- I mean, the first week, we did PagerDuty on tour, and we got out and met a lot of customers. The second week, she listed me off to our Champions Club where I got to meet all the top performers in the company at a recognition event. And then we had earnings last week. So she...

Perfect onboarding plan. This is 16 days.

16 days, yes. So but she manages to work into the weekends as well as you can imagine. But -- and then this week, we have our quarterly business reviews.

Next week, we have a Board meeting.

And next week, we have a Board meeting. So the schedule has really been quite ambitious, and Jen has been really thoughtful, and it's all written out if you want to a 30-page document. But it's -- we're -- I'm really happy with the progress that we're making, and I feel like I've never been involved in anything quite so good. I shared with somebody the other day the last -- this is my fourth CEO job. The first 3, somebody basically handed me an empty Manila folder and said, "Good luck." And I'm really -- I couldn't be more happy with the way that the team is handling this and working side by side with Jen is amazing.

I'm so appreciative that you made time out of your busy schedule. I think you have going on dig it out.

It wasn't an option. They carpooled up.

It's in the transition plan, Koji. .

So I'm most curious about, you mentioned during the early days that you spoke with a lot of customers. And so maybe tell us a little bit about what was most refreshing that you were hearing from the customers out there as you were speaking with them?

This is an emotional type of a thing I'll tell you, but the customers have an emotional investment, I think, in the company. I mean, they really want it to be successful. They really want to see us enjoy success and to continue to help them and to continue to thrive with them. They count on us on their worst days. Our customers rely on us. And that over years, you build up an incredible reservoir of trust and loyalty and that is one of the things that has really resonated with me, just how loyal our customers are. The other thing is we are operating at a very high level inside the company. So you amount of the number of CEOs that proactively reached out to Jen and subsequently reached out to me and said, yes, we're here for you. I said, "Well, that's interesting because I'm the vendor in some ways." But they really do want to see us succeed and that I don't know, Jen, I think that's probably something you experienced as well, and that's a great tailwind to come into a new role.

Yes. I'd say, too, in the last year, 18 months, dialogue shifted a little to customers talking about improving their DevOps and modernizing their digital environments to help us get ready for deploying AI at scale. A lot of our largest customers are still in experimentation mode, and as they start to really contemplate what it means to deploy AI at scale in production, in highly regulated environments, in high stakes customer environments, the operational stakes go up meaningfully and the financial stakes of that go up meaningfully as well. So they're looking for thought leadership. They're looking for someone who can scale and meet that moment. . And that's -- that, I think, has started to show up in the types of customers we're expanding with in the quarter. We talked about an automotive company, a manufacturer, a large, highly regulated financial institution, et cetera. And historically, these were not target verticals for the company. We were really focused on early adopters, tech e-commerce, travel and hospitality, transportation, et cetera. But we're seeing some of the largest of the Fortune 100, Fortune 200 come to us and say, "Okay, now we have to really expedite our modernization to gain the benefit that AI promises, but if we're going to do this at production scale, we have to be able to operate effectively." And so that's really interesting. And then at the same time, when you look at the new logo land numbers, a lot of that is coming from AI start-ups. And the difference between the class of AI startups that we're seeing versus the classes of software startups we saw over the past several years is they have a higher requirement for resiliency and reliability earlier in their growth curve, because if their products don't work, they lose trust and they lose business. It's very different than when you're shipping your MVP like the first viable offering to market. AI fails differently. It's harder to identify when it fails. But when it fails, the blast radius tends to be wider, it can have a much broader financial impact, much faster than traditional software failures.

You said something interesting there, Jen, where I've talked to a lot of customers over the years, and I concur. Customers love PagerDuty. You guys are one of the biggest brands in your category for sure. And you mentioned something right there about these companies scale quickly and failure is bad for the brand. Does that drive for your category to go directly to the most well-known brand right from the start because a lot of other technologies like you hear them, "Hey, maybe we'll use this tool, and then we'll graduate to this tool and this tool, and there's kind of a path, but PagerDuty...

Certainly, I would say that the more recognized brands have higher exposure of a public failure. So you've all been in an airport where an airline has a public failure and operation stop. And it used to be that a failure in the tech stack didn't impact flight operations. But now because of automation, everything from crew routing, to catering, to boarding, to the safety checklist before you take off, all of that is flowing through software. So something central in the infrastructure or central in that ecosystem fails, the business stops. And that's probably the biggest change as people have gone from being worried about protecting their brand's reputation to actually preserving revenue and mitigating risk associated with compliance misses and escalations. It's not just about saving money and lost time anymore. It's actually that many of the financial transactions that drive business outcomes rely heavily on a very complex web of technology where the dependencies are unclear. And so when there's a failure, it's not always obvious, what's driving that failure. And so observability is a part of that, but observability only gets you so far. And you've actually got to be able to quickly reduce the blast radius of that failure. You've got to be able to resolve the issue, and then prevent that issue from reoccurring. And so the big brands have become much more sensitive to this. The other thing we've seen happen over the last several years is these large, more traditional consumer brands or big banking brands, they're hiring more technical leadership. So the CIO used to come out of consulting or the big 4 or program management, et cetera. And increasingly, they're coming out of the cloud hyperscalers, they're deeply technical. They've built Prad, and they're reinventing the way these companies do business. So they're more likely to engage in a conversation around how do you operate your infrastructure, your digital environment, now your AI environment more effectively and more efficiently.

You guys just reported first quarter results last Thursday. Nice share reaction positively on Friday. We have Howard in the room, CFO Howard in the room instance. So I will ask you maybe highlight the key metrics, but Howard, you could always jump in just in case. So if you get -- make sure we get the numbers right. Key takes.

I mean, I think, the key metrics, we beat the top end of guidance, both top and bottom line, significant beat on operating margin at 25%, and raised our op margin guide. We also are seeing significant, I think, leverage in both free and operating cash flow, 37%, 36% there. We're very encouraged by the sequential improvement in gross retention. Retention has been an area we've been focusing on as a lot of our customers have been impacted by it's really the number of jobs and people that they're licensing and so on a traditional seat-based licensing model. That's been a headwind that we've been working through. And then I'd say that we've also seen a really positive reaction to the new Operations Cloud pricing model we put in place, and we shared that 10% of our total ARR is now coming from usage-based products. So that's the first sort of milestone we've kind of put out into the public markets. . The other thing that's been encouraging is where we're moving customers that are larger than $100,000 in spend from more of a traditional a la carte start with instant management and maybe add new products and services to more value-driven total Operations Cloud offering. We not only removed the friction of seat-based usage and seat-based license administration, where the CIO goes and passes a tin cup from one department to another to make sure that different departments can afford to be licensed, that friction goes away. So you see this opening of new use cases, but it also means that customers on the new operations, cloud packaging have access to all of our products and services. And that's driving a more rapid adoption of our usage-based products, including PagerDuty Advance, which is our AI offering, which includes our Agentic solutions, but also things like our event management solutions, AI, ops, et cetera. So the feedback there has been, I think, very positive as well. And I think the demand signals continue to be strong, 5 consecutive quarters of new logos above 600%, I think is enviable to any software SaaS company. And we're demonstrating that we can win in both native AI. So the fastest-growing AI providers are engaging with us and also the largest enterprises in the world, and those are going to be our focus areas. Still work to do, absolutely, but it feels like it's an inflection point, and then a good turning point for the company in terms of the strength of the foundation that we want to accelerate growth on.

One thing I think about, and I think you guys talk about it this way, too, is AI is an operational risk layer. I mean, you can do a lot of cool things with AI, but boy, can it cause a lot of problem?

It is the new operational risk layer.

What's the easy way to think about an example of how that plays out in the PagerDuty world? And how does PagerDuty help solve that?

Yes. And I think we're very early in that curve. AI becomes even more of a tailwind when you start to see it failing at production scale. And so right now, people are trying to manage AI in, I would say, well cordoned off experiments, particularly in large enterprise, right? But even the frontier models who work with us, they are becoming more and more concerned about the elasticity of the platform, like can we scale to meet their volumes? Can we scale to meet their needs. And we've been able to demonstrate through many cycles of growth, whether it was the rideshare platforms and multisided marketplaces that drove a lot of traffic on the platform or the semiconductor players, who are managing not just IT and DevOps but also supply chain operations on PagerDuty. As more and more scale and ingest and events flow across our platform, we're still scaling at 86% gross margins because we have an architectural advantage in terms of how we consume and manage compute, but also because the architectural strategy was designed for reliability and resilience at scale. And as people start to ship more and more AI into their production environments and into their more regulated environments, those operational failures become much more expensive from a top line revenue loss or revenue capture conversion perspective from a compliance risk mitigation standpoint, and then also from cost because when AI does fail, customers tend to retreat and try something that they feel like they can put their trust in. So we're seeing increasing demand on the platform in terms of usage, whether that's number of events, transiting the platform, number of automated workflows, et cetera. And frankly, the number of users is going up as well, but it's simply not the best, I think, economic metric associated with the value that our platform provides in a world that becomes much more automated. And part of our architectural advantage is our platform doesn't really care if it's dealing with a person or an agent, right? It just cares that it's an encrypted secure transaction.

So you mentioned scale. Maybe that is one of the answers to this question. And so you just won 600 new customers, too. So clearly, they see something in PagerDuty that's driving them to you. And it sounds like scale might be one of them, but I just want to make sure when you are winning the customers, what are the 1, 2 or 3 core things that they always tell you, we are picking you because of your ability to do this.

Yes. Part of it is ecosystem, the fact that out of the box, we connect through our APIs to over 700 of the most important solutions in the market. And just to give you an example, like we have a lot of customers that connect us into their ticketing systems, their help desks, their observability providers. And we're, in some cases, on generation 15 of an upgrade integration. It's not simply data flow. And so that means that you can bidirectionally propagate anything that happens in an incident workflow in both systems. So your system of record is capturing it. You've got an audit trail from a compliance perspective, but you can also respond and triage an incident in minutes and seconds as opposed to like waiting for something to come out of a ticketing queue. So that ecosystem has mattered to a lot of customers because it reduces the cost to deploy, it improves the time to value and reduces the total cost of ownership over time compared to systems you have to custom build a lot of these things. And then the second thing is they look at the future road map, and they say, great, your chat first, they're very excited about the Agentic SRE, because instead of a person being called an agent is the first responder to triage is this a new and novel incident? Is this a familiar incident? Can we solve this immediately with an automated workflow or run book? Or do we need to bring more people into the call? So they can see those tasks being completed. And because our SRE agent runs on our data model, which has more than a decade of context in it, it's going to outperform the other sort of generic SRE agents you see out in the marketplace that run on public LLMs in terms of speed, accuracy and fidelity. So they get really excited about that because we can show them that running in their own incident data. I think the last thing is it's future-proof. They look at the investments we're making in our MCP-driven AI ecosystem. So much in the same way we built a context moat through our API ecosystem. We've already got more than 30 partners through our MCPs. We've been out in market now for a year. And that means that like our customers who are building their own agents, they can engage with our agents, they can use our platform. And I love that like people are making news talking about headless environments, but we have customers that have been working with us in a headless fashion for years because they want to use a chat-first experience. And they want to use an agent-first experience, that's fine, too.

You mentioned data there, your data moat. On one side, I always think about companies that talk about decades of data as a differentiator out there. On the other hand, I think about the explosion of data that's going to come that could also power kind of this Agentic future. What's more important, the historical data, the future data? Or is it a combination of both? Like how do you guys think about your moat?

So like data for data's sake is not an advantage. it's do you have the context from multiple types of data and multimodal data that enables you to solve problems faster or enables your agents to solve problems faster? So there's genuine domain knowledge, and there are kind of 3 kinds of incidents that a customer is going to encounter, ones that are very familiar. You've seen them before. They repeat themselves all the time, by the way, more than half the incidents on our platform have occurred before. And that's not because people are stupid. It's because these failures tend to -- the root cause tends to not be identified and the failures continue to surface themselves in different ways. And then there's partially understood failures whereas you recognize some of the pattern, but there are still things that are happening that are new. And then there's totally new and novel failures. And I think the percentage of new and novel failures as a result of AI is exploding. It's going to proliferate. And so you need more context to triage those. You can't simply rely on pattern recognition, human memory and tribal knowledge. And that's where the context of understanding how people operate, how agents operate, where dependencies tend to fail together, why certain events storm to become major incidents and others don't, which workflows solve certain incidents, where fragility lives in an organization's infrastructure. Those are all things that our platform knows because every time an incident runs on that platform, it's learned. And that also leads to, as I mentioned earlier, speed in resolving an incident, and that really matters. If a major incident is costing you more than $1 million a minute, right, you want to resolve the number of minutes or $1 million an hour or whatever the case may be, and that could be in lost revenue, not cost. The faster you can compress the time that incident is having an impact on your business, the better it is. And then accuracy. If you, in trying to solve one incident cause another because of poor fidelity in terms of what your platform is telling your agent to do and what your agent is allowed to do, that's -- or what your people do, people make mistakes as well.

So you guys have Agentic SREs out there. I kind of view that journey as SREs as an assist, like, "Hey, we'll help you." And then like, hey, we'll help you. And then the goal is autonomous SREs, like where are we in that journey, not only from the technological capabilities to do that, which is probably beyond what customers are willing to do today.

Yes. I think the customer is the governor on that, not the technology. So deploying an autonomous SRE is not a technical difficult problem, and we can do it internally, et cetera, and some of our customers are asking for it, but the vast majority of customers want human-in-the-loop automation. They want to be able to manage the permissions around what an SRE agent can do. But what you see in the adoption kind of pathway of a customer is they'll start with a very kind of protective and conservative mindset around what they'll want to allow an SRE to do, and then they see it start to work, and they very quickly get used to it running on its own. So I don't -- I think that's a this year, next year problem that gets solved. I don't think it's a long-term problem.

There's also a lot of learning. The agent has the ability to learn. So it's reinforcement learning with human feedback. And so you get better and better. Every time you handle an incident. And so you also get a persistent memory of the incident that lives in the platform as well. And that has some real potential value down the road.

Is that persistent memory shared across your customers? Meaning will all customers benefit from all incidents happening on the platform?

They do benefit from the general model, but we don't share -- obviously don't share one customer's data with another customer.

Got it. Okay. So full thesis, trying to think about it from an AI tailwind perspective. What's the best way to think about how AI is a benefit, long-term tailwind for PagerDuty?

I think as you -- this idea of as AI moves to production scale, you're going to see the failure landscape explode much in the same way the cyber landscape is exploding. And the incidents that our customers have to manage are going to become much more complex, higher stake. We're already seeing. Like if we just look at the usage data on our platform, more incidents, more frequent incidents, more major incidents and larger blast radius associated with those incidents. And we think we're in very early days. Like we haven't really seen the tipping point towards really production scale outside of maybe the frontier models and some of the pure plays that we're working with in the AI native space. So I think that's one. The second is really monetization, right? Because the usage growth and demand signals are there. It's getting through this transition of being able to go from monetizing the number of users in the platform license to the platform to monetizing the usage itself. And that's where we're starting to see that. We can see that in the cohort of customers that have made the switch, right? We also still have a lot of early customers that are happy to grow on a seat-based licensing model. It's simple. It's easy to administer and deploy when you're small, and they're growing, but at a certain point in time and probably sooner in the life cycle now than we would have done it in the past, we want our customers to have access to the whole platform, be thinking about the platform itself as the operational control plane for their enterprise.

Yes. You guys disclosed 10% of ARR as part of the new pricing model now.

Coming from customers on usage-based pricing.

Yes. How do we think about the balance of that as that continues to become a greater portion of ARR versus potential seat compression and the growth curve that's all kind of underlying that?

Yes. I mean we're in a transition, but you can see we shared that gross retention improved sequentially and that we expect to continue to improve throughout the year. We believe that dollar-based net retention stabilized in the quarter and it's going to start improving through the course of the year. So I think we're coming out of the hard -- the most difficult part of that transition where we're working through the seat-based headwinds, but didn't have our usage-based pricing ready for prime time. And now as we're releasing that into more customer cohorts, we expect that to benefit growth.

I should know the answer to this question, but I can't remember. Is there a goal of percentage of ARR that...

Not one that we've shared publicly, but yes.

Okay.

Yes. I'm so filtering. I like that.

I tried.

I like that. That was really a nice way.

Because you haven't told us, that's why I remember. Growth levers. So how do we think about growth levers, maybe enterprise, PLG, new AI products? Is there something that you're leaning in more versus the other? Or how do we think about that?

I think the Operations Cloud, like giving our best and largest customers access to the whole platform instead of going through this sequential sort of a la carte menu that is very traditional in SaaS. It produces a lot of friction for adoption and growth. I think having a new CEO with fresh eyes and fresh energy is a benefit. I mean, honestly, like when you think about this transition, you have the benefit of continuity in terms of our relationships, all of our stakeholders, talent, et cetera, but I want to be out of the way so that John can make the meaningful changes that he wants to make. I feel like it was a great 10-year run. We 10x the company in 10 years, went from a single app to a multi-app platform that is the leader in digital operations and becoming the leader in AI operations. But I think it's still very early days for the company in terms of where the upside is. And one thing I've been consistent about since before our IPO is this is a long game. Like this business will have a legacy that lives well beyond me or Howard or any of the people here. It's because it is essential infrastructure for the largest companies in the world and the most innovative and important disruptors in the world. And so those 2 segments I think, remain important growth levers for us as well.

I'm going to sneak in one more. I know we're over time here, but you guys are GAAP profitable. It does sound like you're hitting an inflection here on growth and the new pricing model. How do we think about the balance between growth and profitability and really around, as demand improves, what are you -- is there some sort of trigger out there where you may be like, do we take -- do we slow profitability down and invest more for growth?

Yes. I think it will be a good question to ask John, next quarter. But what I would say is we have very consistently over many, many quarters, demonstrated operating leverage. And that's not through announcing ginormous job cuts, et cetera. It's been through structural programs. It is part of the culture and how we operate that we want to continue to optimize. And John was just telling our team this morning that you're either building something, you're selling something and we're optimizing something. And we've been building and optimizing, I think, really well. We're very well positioned to accelerate growth. And now it's about making sure that we can monetize effectively. But I don't think if you surveyed our entire Board or our leadership team, there would be any question that we see ourselves as a growth company, and we'll continue to grow responsibly.

And Koji, I would just add, before I embarked on this whole CEO thing, I was a 4x CRO, 3x in a public company. So I definitely have a growth agenda and I see a growth opportunity, but it's all been disciplined growth. So I do think that there -- I think it is a great opportunity to really navigate that where we can break and gas at the same time and in the right proportion to give the best result back to the investors.

And they're really -- like we're already seeing a significant uptick in the velocity of our feature deployment as a result of using AI in our own development team, right? So AI leverage inside the company, not just in the products super important to us, and we're finding lots of interesting ways to do more with the same, do more with less or do something differently and better.

We've talked a lot about the benefit of adding agentic agents to drive automation in the workflows, but also by leveraging this technology, we can interpret signal with a lot more clarity and be much more nuanced in determining when a customer is having a problem or proactive and preventative. And those are all new opportunities that await us down the road.

Makes sense. We're all out of time. Jennifer, John, thank you so much for doing this. Appreciate it, and we'll catch up soon.

Thanks for having me.

Thank you.

Thank you.