Okta, Inc. ($OKTA)

Thanks, everyone, for joining us. With me today is Eric Kelleher. He's the CEO and President of Okta. So Eric, thank you again for joining us.

Great to be here.

I will maybe kick it off there's so much going on and there's a lot of questions I was putting together, but I really want to start from your perspective, there's so much has changed with AI. So if you think about the identity market year-to-date, what surprised you the most in terms of like the changes that you're seeing.

I think for year-to-date, so if you go back 6 months ago, prior to the beginning of this year, our conversations with customers were fairly broad -- and Okta has grown over the years from providing basic access management to also providing workforce and customer identity and then also adding in identity governance, privileged access management and new security products like identity threat protection, identity security posture management and developer products with the -- also platform. . Our conversations prior had really been about the breadth of all of those use cases for customers. And we would talk about the identity security fabric. And that fabric was the combination of all those capabilities to help customers solve identity and secure identity for all their use cases across all their products and other technologies. What surprised us this year, to your question, is while all of that is still true and we continue to innovate and develop for the identity security fabric and all those use cases, the only thing customers want to talk about today is how to protect themselves from the AI problem they know they have within their companies. Their employees are using AI. They're turning on agents internally. 91% of them report that they know they have agents that are live in production. 22-ish percent of them believe they have them secured properly. And so that's what people want to talk about. They have an exposure today and they need to contain it. And so I guess what surprises us about that is just how acute that transition has been, and it has really dominated our conversations. And as Todd mentioned on the call last week, it's really been a driver of pipeline for our new products. We brought for AI agents to market in November. And Okta for A agents came out last month, April 30. And so -- those products have really been built in consultation with customers talking to us over the past 6 months about how specifically they need us to help them secure AI.

So maybe we'll stick with the AI theme. And to your point, it was AI agents GA for a couple of weeks, taking the call last week that is the largest ages pipeline build for a net new product. So in terms of like how you go to market, like what are your reps doing to convert pipeline to actual dollars? And then we'll talk about the monetization. But just curious more so like what does that pipeline look like? And how are your sales .

Yes. It's interesting because the space is so new and it's evolving so rapidly, and the underlying technologies are evolving. So a few months back, Gemini was the predominant platform in conversation and anthropical leapfrog and now ChatGPT-5 Cybers out. And -- our customers are living in this world where they're not sure which technology is going to ultimately land. And so they need flexibility to be able to protect against all of them. So -- our conversations have largely been talking to them about the importance of neutrality and the importance of having a broad solution that can allow for the rapid evolution of these platform technologies. And 1 of the challenges we have we have addressed and continue to address with this particular pipeline is it's a newer motion. It's adjacent to conversations we've had about securing human identity and securing nonhuman identity and service accounts, but agents are are newer conversations. So for our sellers, they don't have depth -- years of experience positioning these products. There are new products that are just coming out. And so 1 of the things we've done to help them is we've stood up a Tiger team of resources from a sales standpoint, a product standpoint, and engineering standpoint with forward deployed engineers that are working actively with early customers, early enterprise customers to help us both design and build the products and convert pipeline. So that team has been really driving our strategy and our engagement and our specific feature prioritization over the past many months, which is why we're confident that the product that we brought to market is going to fit the need because it was really designed collaboratively with customers.

The -- I know it's a small sample set, but on the call, a lot of investors kind of focused on the commentary where I think you said if there's a I think it was more so on the call back, which was a customer comes in for an upsell, they buy IGA. They were to attach AI, that component would be larger than itself. Can you maybe just -- I know it's a small sample size, but just provide some color on what you meant like when you say larger in size, I assume absolute dollars, but now to the magnitude, and again, smaller sample set, but you take that to a lot .

Yes, what Brett talked about in the call is that it's about 40% larger, significant. And it is -- it really is testament to the value that customers place on the exposure that they're working to mitigate. And as every successful cyber attack earns a headline, customers get more and more aware of the fact that they have an exposure that they need to mitigate against. And so that helps them understand they need to control for all the use cases. The challenge that customers have today the enterprises have today is that they can control their employees. Employees are able to spin up an agent and give it access to their e-mail and access to their calendar and access to their Google Drive, which contains things that they don't even know they have access to. And so as that becomes more prominent and more acute and as there's more issues in the press, companies don't want to be tomorrow's headline. And so they know they need to invest to put themselves to increase their security posture and put them in a position where they can protect against having their own agents compromised or over permissioned or impersonated. And so that's really what we're seeing the level of investment we're seeing, that's where customers see the biggest exposure to.

And then if you think about offer and then Okta, where -- it sounds like Okta is having more success early on with AI just given the deployment because it's more internal versus all side. But maybe walk through those dynamics. And does that shift over time I don't -- I wouldn't agree with the premise of that question.

We're having good success with all 0 for a agents as well. We're you might not have heard us talk about it as much on the call because that is -- it's an upstream product. So to for AI agents is a tool for developers that helps them develop agents that can be secured. . And it allows, for example, for fine grand authorization where you validate -- when an agent takes action, you validate not only that it is who it says it is, but also that it has the specific permissions to do the thing it's trying to do. That capability is hugely important. Token vaulting, which is included there is hugely important. So we've had -- we continue to have good success with that product as well. From an enterprise security exposure standpoint, the more acute challenge our customers are facing right now is that they don't know 3 things. And we call answering these 3 things, the blueprint for the secure identic enterprise, where they don't know where their agents are -- they don't know what they can connect to and they don't know what they're authorized to do. And so Okta for agents is a platform that addresses those 3 areas, helping with the problem of discovery, how we find what agents are live in production for your company and then helping with the problem of connection and ensuring that you know what those agents are able to connect to and then help them with the problem of authorization and knowing what they're authorized to do.

I guess the question I have is -- and everyone ask within identities. If you think of privileged access that you have IGA, which you have further low the maturity curve and then think about your core it's more important or a bigger priority. -- for customers in terms of securing AI?

In terms of securing AI, so governance continues to be a very strong contributor for us, and we don't see that slowing down. in addition -- and governance plays a role in securing AI as well. And for example, 1 of the use cases there is just-in-time provisioning. So 1 of the challenges with agents is typically -- not typically, often agents are configured with standing privileges, meaning you allow an agent to connect to 1 of your systems and it just has access to that system, standing access to that. One of the areas we can address that with government provisioning and deprovisioning where you wake the agent up when it needs to do work and then you turn it off immediately afterwards. So you solve the problem of standing privileged to that. So governance is an inherent part of the solution for securing AI. But then the October agents platform helps you -- again, help out those fundamental questions. First of all, knowing where the agents are, so that you know which ones might have standing privileges and what to look at, knowing what they've been giving access to what they can connect to. So again, you can help you understand what that exposure is and then knowing the verbs, the actions that they're allowed to take. So is the analogy where it's -- everyone needs governance, but not everyone needs privileged access, meaning maybe to CSO, someone in the security team, I don't personally need access to turn on and off MFA.

So does that mean IGA is probably a bigger upsell than privileged access or do agents need both? .

There's aspects of privilege access that really help with AI and especially in particular, token vaulting, is a feature published access and also a feature for AI agents. And that token vaulting allows an agent to have credentials, which are managed in a secure location and have business rules that requires them to be rotated -- and if you go back 10 years ago, people would configure a service account and hard code a user ID and password and a config file and be done with it, and that would live that way for years. We've all learned as an industry that doesn't work. It's a huge exposure. So -- so the privileged access will become increasingly important for more of those use cases over time. As far as our business is concerned, it's a newer product for us yet. It's not yet a land product. We have now matured Okta identity governance to where it's now a land product for us. And in addition to being a frequent cross-sell for existing access management customers we're also going into displace entrenched legacy deployments from other vendors. So governance has matured to the point where where it is now a point of entry for us, which is great. Privileged Access is a couple of years behind governance and its maturity. It's certainly not -- we don't consider the land product for us today. But for the use cases that it supports around access management and governance, it's been very effective.

Could you share what percentage, I guess, of net new customers come through IG?

We don't break that out. We don't keep me honest with that to me -- so no, Sorry.

Now -- maybe talk about the demonetization. You mean this whole conference, everyone has been talking on the software side, how this plays out over time. How are you thinking about monetization of these agents, consumption, usage base? Maybe just walk us through what that trajectory looks like.

Todd talked about that on the call last week. The -- 1 of the challenges the industry has right now is the economics of AI and inference and compute are such that most vendors are bringing to market their offer with a consumption-based model. The challenge with the consumption model is customers don't know how to budget for it. And so leading with a consumption-based model slows down your sales cycle materially. And we have this issue internally as we're buying AI products and trying to figure out how do we budget for what our consumption is going to be. So what we're doing with these early customers is we are entering into 1-year terms, and we're only having a seat-based model for now. And we've talked to all these customers coming in that for the first year, we're going to have a seat-based model. Here's what it's going to look like and they can budget for that. And then we have the understanding that during the year, we're going to learn their actual consumption so that if we were to add in a consumption-based component at renewal, they would have data to ground their budget cycle on. And that the reason that we landed on that is just by working with these customers over the past 6 months is what would make it possible for you to get started. And that's been very effective so far.

Okay. Maybe jumping to, I think, some of the comments you made on go-to-market. You've made a lot of changes over the past, call it, 6 quarters. Most of last year was kind of ramping some of these net net reps. You called attritions basically at levels where you haven't seen it in a very long time. So you were senior level. A great level -- to be clear, that could go the other way. . Righ. So -- you've -- it sounds like you're kind of pulling back, I guess, a little bit on net new hires. Maybe just walk us through some of the commentary on the call in terms of why not silo.

Yes, it's interesting. We went through a couple of years of needed go-to-market evolution. And 1 of the questions we would get in these conversations is how much disruption is too much disruption? And are we -- how are we managing productivity and tenure and time and seat -- and we're looking back, we're very confident the move made with the right moves. So our waves is specialization, we're now in our fourth, which I'll get to, but our first few years back was to carve out a set of sellers to focus on public sector, federal and state and local and international. And that team has had a very good run focusing on those buyers in that budget cycle and those purchase cycles. We then, 2 years ago, rolled out a hunter farmer model in our Americas Commercial business. So we move reps who had been accountable for all new logo acquisition and upsell cross-sell and upsell. What we had found is that, that business was shifting so much in the cross-sell and upsell. And largely, as we brought more product to market, they had more to sell. -- that we weren't spending enough time feeding the top of the funnel by bringing in new logos. So we carved out a set of sellers and put them in a hunter motion where their focus was on new logo acquisition. And we've talked about -- it took about a year for that model to get to really get running at full speed. And that's what we had expected and how we've modeled and set guidance. And then this past year, 5 quarters ago, we rolled out a specialization model on our platform and our buyers. So whereas previously, every rep sold everything, we divided the sellers into 2 populations, 1 that sells to the corporate IT and security buyer, the CIO and the and they sell the Okta platform to that buyer and 1 that sells to developers, whether it's with software developers or B2B SaaS or even in-house developers and they sell the ACO platform. And that specialization we rolled out January of last year. And we talked about this throughout the year, it's proven to be very effective. So -- as a result of all of this, we have seen several quarters of improving rep tenure as attrition has gone down and 10 year has gone up and productivity has gone up as well. And overall, when you talk to the field organization, not just at our club event a couple of weeks back, you can feel the optimism and the people know that they can be successful and know that they can make money. And so when we look at capacity to get back to your question, we want to be careful to add capacity to drive up bookings, but not to add so much capacity that we disrupt territories that we disrupt books of business that we have to reallocate accounts and then people have to reset new relationships. We find that when people have accounts longer, they're going to sell more into those accounts. And we don't want to carve territories to the point where reps lose some momentum and feel like they're losing some opportunity to be successful. So we feel like we have that balance about right. We've added a bunch more capacity Q1. We're adding less incrementally during this quarter and next quarter. But we're going to measure based on our pipeline to make sure that we're not leaving deals on the table. So we'll adjust as we go.

Where is the -- you talked about public sector real quick, got a good strong Q1 -- what's driving that? It sounds like it was across the board, state, federal and local.

The -- overall, the team is well tenured and well productive and they've got the product mix that those buyers need. I think 1 of the things to keep in mind also is Q1 is typically a low quarter for public sector. The federal fiscal year starts in our Q3. So Q1 is typically -- we don't typically have a lot of renewals or a lot of net new purchases at the federal level, certainly. But state and local continue to be active, and we had some good federal deals in Q1 as well.

The 1 thing that also stood out to me was just the step up, I think, in partner-sourced bookings multimillion, I think, was up year-over-year. Along -- I guess, more so that was part of your shift and I want to talk about shifting more of your pro services to your partners. Obviously, that broadens distribution, makes it a lot more stickier. They're going to push your product. What was the move behind that? And part of it was part of the go-to-market changes. But if you think about just the partner-driven deals like what -- how does that look throughout the year.

Kind of it's 2 things. One, we've had for many quarters. We've had -- our most exciting growth has been at the top of the pyramid. And it's been upmarket. And we -- I think we did announce in this quarter at our customers over $100,000 in spend are now over 85% of our revenue. And our customers over $1 million represent over $1 billion in ARR. And so as we go upmarket, we're having bigger conversations. And as our product portfolio has expanded, -- we're having conversations that span multiple processes and multiple departments and multiple use cases. And as that complexity increases, our partners have an increasing need to be working with the systems integrator that can spend broader parts of their business. And so we have been working with the global systems integrators across all of this. And what we realized this year coming into a planning cycle is that we could take a deliberate step to refocus our professional services resources on enabling those partners to get greater scale. And so that's the adjustment that we made, where we basically shifted 1% of revenue capacity towards partner enablement and working with those GSIs and these accounts. And that's played out as we expected. It's both financially and from an engagement standpoint. We have active engagements at customers with all the major GSIs, -- and we're seeing that continue to improve. And ultimately, what will happen is as they build more experience, they build more bench capacity, they bring us more business because the GSIs sell what they have in their bench. And as they build Okta capacity on their bench, they're going to be bringing more opportunities to us.

When does that become more meaningful in the model? Is that Q3, just maybe when does it really start to step up?

Brett's comments last week was that we expect to start to see that contribute in Q2.

Okay. The -- maybe going back to AI. The question I had for you, and again, a question I've been asked all this whole week is incremental AI dollars being spent. Where is that coming from? Is it being pulled from endpoint network? Or is it like new buckets of spend that CISOs are getting?

It's really from it's both. So it's people rebalancing their portfolio, but there is a lot of net new investment coming in. And it's coming in because this is a board-level exposure. So as this market evolves and as threat actors get more published, boards understand that this is a real security concern. The problem of having agents having standing privileges to corporate assets. The problem people are understanding that they're behind in data governance and ensuring that they have proper -- they're finding their humans have access to data. They didn't know their humans had access to, but now agents are on they're 24/7, and they work at a much higher clock rate than their humans work. There's increasing exposure that needs to be managed and invested in. And so we're finding board-level engagement and investment in Agentic security specifically. So there's net new funding coming in. And then in addition to that, companies are looking internally at where they can harvest some savings to drive into this area as well. So we feel that it's very much a tailwind right now for the product, which is why we've innovated the way that we have with the offers we brought to market.

So it sounds like consolidation is definitely a trend. -- platformization. Obviously, 1 of your competitors brought it up. But with that, I mean, I guess, within AI, where is the most success? Is there a specific geography, a certain vertical that you're seeing more traction today than you probably wouldn't have expected 3 months ago?

On vertical or geo. These are pretty global needs. I think where we're having the fastest success in these conversations is where we're bringing in our specialists. And so the reason we've created a Tiger team is they do nothing but talk to customers about this specific problem all day every day. And so we direct that team to the -- what we perceive to be the largest opportunities. We have activity in all of our geos -- but I wouldn't say that there's a single vertical right now that is standing out beyond others.

If there's any has any questions, feel feel free to jump in. The -- maybe we'll talk about competition. I would love your input on what you're seeing or your expectations from Palo Alto with CyberArk. Obviously, it's more on the Privileged Access side. Palo making a big push into our identity. I think they came out with Alcon ID, which I think is more more of your core slightly, but very lower end products. Where is this market evolving to?

First of all, it's a huge pie. The market is enormous right now. Second, -- the 1 core differentiator that -- and we talked about last week as well, is our neutrality is huge. And customers don't know which vendor is going to own which parts of the stack, and they know that they need to be able to solve and secure identity across whatever technologies are coming in and out of the mix. So that continues to be a core differentiator for us. We also have, from a distribution standpoint today, over 20,000 enterprises that already trust us to secure identity. And so for them, expanding their secure identity to include any use case with agents is a very natural extended conversation of the work that we're already doing with them. We've done a really great job building those connections. And then from a partnership standpoint, I mentioned we have over 8,000 applications that are integrated with us. Those companies you mentioned, CrowdStrike and Palo Alto Networks, they're huge partners of ours. We have thousands of mutual customers. So those relationships, obviously, will evolve as they -- as we overlap more on the edges, but we see it on the edges right now, not in core head-to-head.

Okay. What do you think Palo Alto will have the most success with CyberArk?

I think CyberArk as strong is on privileged access, as you mentioned. They have many years of incumbent access there. Their solution is legacy on-prem. the work right now to bring it into a multi-tenant multi-tenant SaaS, and that's taken some effort for them. It will continue to evolve with Palo. So for firms that are Palo customers that have a privileged access specific need, they now then and now are going to be a natural offer for that. But again, that's not a land product for us today. So we wouldn't expect to be head-to-head in those transactions, where we're broader on access management and governance and security threat protection and posture management, we'll bring those offers in to help balance that.

For your early identity customers, to what extent do they have to consolidate identity stack across the 3 legs of the stool to make the best use of Agentic products.

The agenting products bundle all the capabilities that are needed from those stacks. So if a customer buys Okta for AI agents, they get the components of access, governance and privileged access that are required to secure the agent use case. That's how we would package it that way.

Because if I had sales force for governance and SebaroPAM, and I use you for access, could I deploy your agent solution or will need to migrate.

No, you could buy it full feature set. Yes. The Okta for AI agents to be clear. Also on the build side, but on the governance side, yes.

Can you maybe help us understand where does CrowdStrike compete? Is it more complementary? Is it more competitive? Where do they fit in the stack and then.

We don't see them head-to-head at all. The -- we talked about this on the earnings last week. The the 1 named competitor that we see materially is the SIEM has always been is Microsoft. And the primary play they have is bundle, right? So for customers that are that have a very specific need for the small subset of capabilities that Microsoft can provide in the E5 and E7 bundle and are under cost constraints. That's a legitimate competitor, and that's something that we grapple with. But our win rates against Microsoft haven't materially changed in many, many quarters. But that's something that is a constant conversation for us because if you're not using the breadth of our stack and you're only using a subset of access management and you have -- you're already a Microsoft bundle customer, we have some exposure there. So our account teams are very driven right now and driving our stickier products that get us in great things like workflows, things like governance and privileged access and our security products. And as we add all those in, we become even more of an established standard and people build on us.

Is there a step that you can maybe share like how much overlap how many Okta customers are using maybe parts of Entra, what Microsoft has?

We don't have published those stats, but we have many customers that are also Microsoft customers.

Is there a specific part of the Microsoft stack they're using that perhaps that you guys don't have? Like why would -- maybe walk us through like what -- if it's a Microsoft customer using Okta and Microsoft where is the overlap?

It can be a customer that's migrating over time and coming more -- it can be a customer that has a legacy Microsoft app that they're using that has been previously configured that they haven't yet brought over. Those use cases are very, very common. But usually, it is either that or it is people not really using the intra components, but they're using other components of the bundle.

Maybe talking about some of the cores, look, maybe I'll start with all. I mean that business has continued to do extremely well. checks in the channel on the developer side. It's just the library of applications. Where does that business trend? And what are the investments that you guys are making into Alero today that you think will kind of help sustain that growth?

The Astro continues to serve its core audience, which is developers. Those developers historically, we've talked about Astro a customer identity access management solution, which it is. And the reason that we've used to lump those in is almost anonymous once upon a time was because the primary use cases people were using a 04 was to build applications that serves customers in the B2C model. That's expanding now. Developers are building additional things. So we have a focus on B2B SaaS providers, people that are bringing applications to market. And so for those customers, you asked about innovation targets, for example, we are investing to earn FedRAMP's authorization for the AZ platform, which it does not have today. And the reason that we're doing that is because B2B SaaS vendors need to be able to sell their products to government. So we need all 0 to earn FedRAMP accreditation so that they can sell their products to government. That's an important investment that's well underway for us right now. And so in addition to B2C, B2B, we are now seeing agentic development as the net new category. And the reason that we've been investing in that product is specifically to serve as that use.

Okay. I guess is like how do you manage that investments across Okta and then Allco.

Today, our -- the engineering teams are distinct. They were both reporting to Rick Smith, our President of Technology. And Rick is and that's a recent change we announced a month ago. And Rick's Charter with pulling them together is to make sure that we don't duplicate efforts. So for example, we've spent many years earning FedRAMP authorization on the OTA platform side. It's Vedran authorized. We don't want to have to repeat that and start from scratch with the work that we're doing on users. So he's looking for programs like that where we can get some commonality and some some reuse of the skills that we've already developed in-house. And that will work in both directions for us.

The other impressive, I think, stand in the quarter, net retention ticked up for the first time, I think $106 to 107 sounds like perhaps maybe there is an inflection maybe later in the year. But walk us through how much of that is the core buying net new seats, AI? Is it IGA? Is it Privileged Access? Like what drove the trajectory to take it higher and then throughout the rest of the year?

Yes. So we've mentioned we expect it's going to travel around here, plus or minus a point for a while. It's all the things that you mentioned. The I would not attribute Privileged Access specifically to being a material contributor. But cross-sell and upsell across the entire new product portfolio had another really strong quarter for us. and that always helps. And we've -- as we've talked about in recent quarters, we've, at this point, flushed out the COVID cohort of exuberant purchasing in 0 interest rate days. That's helped us on the NRR side as well. So those headwinds have abated and the tailwinds of having new products that are available that are driving incremental value is really resonating as well.

Okay. You can see -- Are there additional. I guess why are you talking about maintaining the current level or through your go-to-market or you lap the large or customer loss in August -- so why wouldn't you expect improvement in the back half?

It changes every quarter based on the mix of new business and based on the mix of cross-sell and upsell. And so our expectation is for it to trend about where it is -- we're pleased to see it tick up from 106 to 107 and we're working to drive it further. But for now, we're calling it where it is.

If you think about the growth formula for the full year guide, how much of that is just maybe an uptick in net retention, better upsell and cross-sell, which you are seeing versus just net new logo growth? Like how should investors -- how should we think about the sustainability of, call it, high single-digit, low double-digit growth -- for the next 1 to 2, 3 years, obviously, you're not guiding. Walk us through like where is growth coming from.

Keep me out of trouble on this. What do we say on this one?

Yes. It's a confluence of all of the major is that we ingest -- so it's going to be -- what's going to come from improved market execution, new products, which includes some of the agent that that's a little bit further out, investing in the partner motion and shifting some of our revenue over to our partners. -- and success of large businesses. .

Yes. channels -- we haven't talked about channel much here, but the channel for us as well has been an area where we're pleased with the momentum that we have, and we're investing further in developing that. The GSI is 1 example of that.

What percentage of your bookings in today is partner-driven? And what does that look like 24 months from now? .

Yes. We said that over -- I want to hand numbers right here. Over 80% of our partner influenced. -- over 40% are partners sourced because that's when they actually source the deal and hand it over to us. We believe you can get those numbers up higher over time. That's what you see us making.

Okay. Now the partner influence versus partner source, like who's getting the pro service dollars? Is that the partner sourced deals.

For deals sourced with SI partners, they take prime on those deals, and we'll support them with expert services. For partner influence, it can go either way. It depends upon how the engagement is. Our preference is to prime as little as possible. We want to make sure we maximize the opportunity for our partners. But we -- there are some use cases that are either bleeding edge with new product capabilities or required deep work with our engineering team where it makes sense for us to be prime on those transactions. But we try to minimize that so that our partners have the opportunity to build their depth of relationship with customers.

Okay. Maybe 1 more financial question, I'll just end it with another maybe if you could help keep me honest. So if you think about the margin guide, op margin guide, free cash flow margin is now high teens, if not 20%. As you think about going through the year, perhaps if you then double down higher on go-to-market, new reps, the investments in R&D across all. Okta. What's the sustainability of margins and the trajectory there?

We managed to Rule of 40, and we we balance growth in free cash flow, and we work to keep them in balance and make sure we manage to rule of 40. Right now, as we've talked about, we have -- we believe we should be growing faster. And so we are indexing our investment to elevate our growth rate as fast as we can. And we feel confident we've demonstrated we can generate cash when we choose to generate cash. But what we want to do is, while we're focusing on efficiency, we want to make sure we don't starve investment opportunity that could accelerate our growth rates. So that's how we think about that balance right now.

Okay. Maybe the last question is, what's your pitch to investors today? Your stock has obviously outperformed since you guys reported. Execution has been a lot more flawless and more consistent over the past couple of quarters. So if you look at the stock today, your guidance kind of the tailwinds from AI, the monetization opportunity, I think most of us see within security because every agent needs identity authorization, et authenticate what every agent has. So maybe what's your pitch today? And then what does the story look like in a year?

It's a pretty good baseline right there. So securing identity has never been more important than it is right now. Human identity has -- we've known that's been important for a while. -- nonhuman identity and service accounts over the past 7, 8 years, people have realized they have huge exposures there. And now genetic identity and exposure has really poured gasoline on the fire. So from that overall perspective, there is an acute awareness of the importance of securing identity right now, which is going to help Okta. And I would say in addition to that, the challenge that new players are going to have in this space is the challenge of distribution and connectivity -- and our neutrality and our -- the depth of our network effect of our existing integrations makes us almost plug in play for any customer with all these use cases. And so we believe we're very well positioned to take this market. And we're working our asses off to make sure that we do that, excuse me. The -- as Todd mentioned on the call last week, we don't get paid for pipeline. We got a good -- our pipeline right now is really healthy. We've got to convert it to bookings. And so we're very focused on doing that.

Maybe just 1 last follow-up. You talked about the monetization. One question I get often is, I think SailPoint puts out their metric is for every 1 human employee, there's anywhere from 5 to 40, I guess, agents or chatbots, copilots attached. What are you seeing today?

It depends who you talk to. The IDC's projection is that we'll have -- collectively, we'll have 1 billion agents live in production in 3 years out. So it's hard for people to project. And as you just mentioned in the question, it's also important in how you define an agent, what's an agent versus a chatbot versus a piece of software tool that you use -- we -- I expect personally and I can say on behalf of Okta, we expect there's going to be a lot more agents than there are people. And all those agents bring with them security concerns and security exposure and they expose data governance concerns and data exposures. And so we believe all of that is going to bring agentic identity security to the point where it becomes much more of a prominent conversation even than human, which is already critical for folks. So 1 of the areas that we talk about frequently, that remains true is over 80% of successful cyberattacks, over 80% of successful cyber attacks start with compromised identity. And as the industry has learned that as the industry has become aware of that, they understand that they cannot have a secure company if they're not properly investing to securing identity. And we believe we have the right to earn that business to be that for them.

Quantum question. I know it's like a couple of years down the road. Is that a concern at all for customers today? If you think about Quantum breaking through any -- basically any kind of encryption.

From time to time, we'll -- I'll talk to CISOs about their thoughts on quantum, but what I can say is it's not a -- I'm not experiencing it as a top-of-mind concern today. It's on the radar, something they should probably be thinking about, but it's they have acute issues to manage today and that's something to think about in the future. Not there yet.

If there's any other questions, I think we could Eric -- all right. Thank you.

Thank you.