Informatica Inc. (INFA) Earnings Call Transcript & Summary

Okay. I think we're live. Welcome back, everybody. Brad Zelnick, Deutsche Bank's software team. Delighted to be here for this session at our 2024 tech conference in sunny Dana Point, hosting Informatica, Chief Financial Officer, Mike McLaughlin. Mike, thank you so much for joining us.

Thanks a lot for having us. It's a great location and a great conference.

Yes. Great to be here. Format of today's presentation is a fireside chat. I've got a number of questions. I don't know if we're going to get through all these, Mike. But we've got a lot of fun things to talk about. Exciting time for Informatica. And maybe with that, I'll dive right in. Just maybe to kick things off, you reported Q2 earnings about a month ago now. What's the key message that you hope investors took away from the event?

Look, it was another good quarter of consistent execution for Informatica, which is our goal. Nothing really stood out over anything else. The cloud growth, which is what we exclusively sell today, was 37% year-over-year versus our guidance of a little north of 35%. The decline of the parts of our business that we don't sell anymore, the end-of-sale maintenance, the end-of-sale on-prem term licenses, declined at a rate consistent with our expectations and with our guidance. Our operating expense management continues to be rigorous, delivering operating margins close to 29%, all of which gave us the confidence to raise our Cloud ARR guide for the rest of the year to 35.5% growth versus the 35% that it was going into quarter end, and to raise our operating income and free cash flow guidance to deliver north of 33% margins by year-end. Our growth in our overall ARR, the combination of the growth of our cloud and the decline of maintenance and self-managed, was 6.6% for the quarter, and it's on track to be consistent with our guidance of 7.3% for the year. And as -- we'll probably get more into the moving parts as we go forward in this conversation, but that ARR growth is going to be higher in '25 because the growing cloud bit is going to be almost 50% by the end of this year and be well north of 50% into the first and second quarter of next year. So that growing bit is going to be bigger. The declining bit is going to be smaller. And so total growth is going to be higher in '25 than it will be in '24, and it's going to be higher in '26 than it will be in '25. That's why it's such an exciting time. It's not that one cylinder is firing hotter than the others. It's all going according to plan and is giving us increasing confidence that our financial trajectory is going to be of accelerating growth, making money while we do it.

Excellent to kick things off. It is an exciting time for Informatica. And I thought rather than just show up and ask you the very fluffy questions, as the Chief Financial Officer, we could start through the lens of the numbers and then use that as a basis to kind of get deeper into the story and the why behind those numbers. And it was good to see both subscription ARR and Cloud ARR guidance for the full year raised, which I think best captures underlying performance. However, full year revenue guidance was reduced modestly due to a couple of factors that I was hoping you can just help unpack. The first was lower contract duration on self-managed renewals, which doesn't impact ARR but does impact in-period revenue recognition, which I think most people here understand. My question is why self-managed duration is coming down faster than you previously expected? And how much of this is customer-driven versus any behavior that you might be incentivizing in? And should we view this as a leading indicator of customers looking to move to cloud, maybe if not today, at least within the next year or 2?

So I'll answer that by backing up the half step just to make sure everybody is on the same page. Since going public, we have disclosed and guided to ARR, we also guided to total revenue. But ARR is the more meaningful way to track our business. We're a transition company that has a long tail of legacy on-prem maintenance and self-managed, which has all the ASC 606 noise. And as an enterprise software company, we have a portion of our business, that's professional services, which is nonrecurring and low margin and again, isn't a driver of value for those 2 reasons. So the disconnect between ARR and revenue is unfortunately driven by those 2 factors primarily. And it leads to noise, and it's important to understand why that disconnect occurs from time to time. In this particular quarter, we brought down our full year total revenue guidance for 2 reasons. One is a decline in professional services revenue. And I'll start there as opposed to the direct part of your question, which is the self-managed duration piece. We provide implementation and high-value consulting services to our customers, most enterprise software companies do because we want to ensure that customers get to value quickly and that the product is installed correctly. Over time, our products are so widely used, we have such scale, and they are so popular, if you will, with the GSIs because they're part of major transformations in most cases, with companies that our partners train their people up to do that instead of us, and that's awesome. This year, we expect 12,000 or more individual certifications by GSIs on Informatica's products. So Informatica-certified implementation people at GSIs around the world that are doing that professional services business, which means we need to do less of it. So it's customer pull saying, we want Deloitte to do it instead of Informatica because it's cheaper, and they're doing other stuff for us. So that's why professional services is going down, and it's going down faster than we thought it was going to decline. We knew it was going to decline this year, and we guided to that at the beginning of the year, but it's going down faster than we expected for the reasons I described for good reasons. So that's that bit. The second bit is ASC 606 upfront revenue recognition, which we still have in primarily the on-prem term license part of our business, which we call self-managed subscription. This is just one and the same. So every time those things renew, we have to accelerate a significant portion of the total contract value of those engagement, those contracts just because of ASC 606. There's a portion that's ratable, but most of it is accelerated, and we have to recognize it on the day we signed those renewals, and it creates this lumpiness that drives all of us CFOs with this problem crazy. So -- and the duration of those renewals matters because it's a TCV concept of how much you have to accelerate. So we have found that the duration of our average renewal in that on-prem term license base has been declining, not like by 50%, but enough to be noticeable like a number of months on the average versus a multiyear average. So why is that? And that was the other half of why we brought revenue down because it doesn't affect ARR, doesn't affect cash flow. It just means that the renewal -- the next renewal comes up sooner. And it is primarily driven by customers setting themselves up to move that workload to the cloud at some point. They're not ready yet, but the writing is on the wall that they've got to move at some point. And so just for prudence and flexibility that many of them are saying, I'm going to do a 2 year instead of a 3-year. I'm going to do 18 months instead of 2 years, so that I've got the flexibility to move sooner. And we don't view that as a bad thing...

Seems like a good thing.

Because we've got the catcher's net for them when they're ready to move. It's our products that solve those needs the best, and we already have the relationship, and they're happy customers. So it is unfortunately an accounting impact, but in terms of the medium term of the business, I don't see it as a negative.

I appreciate you taking the time to explain it because maybe somewhat counterintuitive to the casual observer. It sounds like it's a good thing. Or hopefully, we'll find out in the years to come, but...

Yes, we think it is.

As we see the move to cloud, maybe in addition to the duration dynamic and the impact to revenue, which I don't really think has been a source of pushback. We've also seen subscription gross renewal rates decline in the past several quarters, driven again by the self-managed piece. I think it's important to acknowledge you're no longer actively selling these offerings, but self-managed is still about 40% of subscription ARR and a big part of the longer-term cloud migration opportunity. Can you just help us understand the dynamics behind self-managed gross renewals? Is there anything you can do to help reduce the churn from these customers? And when should we expect overall subscription gross renewal rates to perhaps recover?

So again, I'm going to answer a different question first, and then I'm going to get to your question. Sorry, I have a habit of doing that.

No, it's okay, as long as we get there.

Subscription revenue and subscription ARR, actually, I think, is confusing. It's a legacy from the IPO, where we were still in the midst of a transition from a mixed model, selling both on-prem and cloud, and we're transitioning that on-prem from perpetual license maintenance to subscription. So it was meaningful to show how much of our revenue and ARR at that time was subscription-based versus the old perpetual license maintenance base. We are end of sale all of our on-prem products. We did that at the beginning of 2023. And so -- and we haven't sold a perpetual license of any materiality for years. So that subscription versus nonsubscription differentiation isn't the right way to look at our business anymore, and we're going to discontinue that. Exactly when, we'll put in our footnotes so you will have it. So you want it as long as you want and you can do it because all of it is adding cloud subscription on-prem. It's not a higher math. The way to think about our business is cloud subscription will be 48% of our business at the end of the year, growing at 35% according to our guidance. Maintenance on perpetual licenses sold in the past, about $400 million, declining at a certain rate that historically has been about 5% natural churn and then additional couple of percent on top of that, that moves from maintenance to cloud and self-managed subscription, which is on-prem. And that's declining at a faster rate because it is less seasoned. This is where I get to the point of answering your question. So if you're going to group things now and going forward, don't think it as about subscription and nonsubscription, think about as cloud and everything else. Cloud and on-prem, okay? So the right question is not subscription renewal rate because that's a mixture of cloud and self-managed. It's about the renewal rate and the gross retention rate of the on-prem stuff, which is maintenance and self-management. Now the dynamics in those 2 are a little bit different. Maintenance is very, very seasoned. Some of those -- our average customer tenure is almost 15 years. Some of us have been with 20-plus; very, very sticky. So that declines at a very predictable rate, mid-single digits. And then on top of that is the migration of some of those customers to our cloud. Self-managed is less seasoned. We've been selling it for less longer -- for a less amount of time. Average ratio of those or tenure of those customers is 5 or 6 years. And so the churn rate -- the natural churn rate is just higher. There's -- in some cases, there's under-deployment. In some cases, the use case didn't work out. There's some; shelfware here and there. So that is just higher. And the increase in the churn rates that you can see that we disclosed is merely a fact of that self-managed is going to where we fully expected it to be in light of the fact that we held up a big neon sign at beginning of 2023 saying, this is end of sale, it's not the future. And so the initial churn rate of that in the first few quarters after we end-of-sale'd it was not where we knew it was going to go to because not a lot of renewals has happened. So it's not going to come back. That renewal rate in the self-managed piece is right about where we think it will be. It's going to look higher next year because despite the fact that we're not selling it actively, sometimes it's still bought because there's customers that still need on-prem because they're in a geography that has data residency requirements or they're a federal government customer or whatever. So there's still some new going into the top of that on-prem subscription bucket, but that amount new is going down and so that shows up in the decline rate of that ARR from year to year. So don't expect it to go back up, but do expect it to be consistent with what we forecast and consistent with what we've guided to, both in terms of the next 2 quarters and in terms of the medium-term guidance that we've given for '26.

Got it. Thank you for taking us through that. I think it's helpful. It's helpful to have it transcripted and memorialized, so we can all go back and we can remind ourselves everything that you just said. Maybe just looking forward, net new ARR guidance that you have for the second half doesn't really seem overly demanding relative to historical trends or the first half for that matter. What can you tell us about the pipeline heading into year-end? As is typically the case, I know new business is seasonally weighted towards Q4. But anything else that we should be keeping in mind as we think about the back half?

Business has felt very consistent for us as the year has gone on. And it felt consistent with how it felt at the same time in the prior year in 2023. Now I just started here in 2023. So that's my only personal data point. But the sales cycles don't appear to be getting shorter or longer, the amount of scrutiny and the level of sign-off required doesn't seem to be shorter or longer. The pipeline conversion rates are very consistent with what we've seen historically through '23. So it's that issue as you go. And as we were sitting here at this date last year, the amount that we still had to go to make our yearly guidance was about the same as what we have to go this year on a proportional basis, maybe a little less this year, and we're a little bit ahead of the linearity that we had last year, but I wouldn't call that meaningful. So again, it gives us the confidence that our guidance is still good guidance for the full year.

Fair enough. Mike, maybe one of the biggest questions we get from investors is around the confidence in your medium-term target. So beyond the year, you've got a target to hit $2 billion and be growing double digits by 2026, which based on current guidance would require acceleration of net new ARR growth over the next couple of years. You did a good job laying out the 3 major drivers of that at Investor Day last December. But I wanted to dig in a little bit to each of those. So starting with digital transformation, which is one of the 3. As companies adopt hybrid and multi-cloud, Informatica has positioned itself really well as the Switzerland of data with IDMC as part of this. You've got unique partnerships with some of the leading data platform names that we would know in modern generation, Azure, Salesforce, Snowflake, Databricks and in many cases being among the first natively offered data management apps on their platforms. How important are these relationships to your goals? And are they real long-term differentiators or perhaps just more a reflection of greater opportunity in the industry?

I think it's more of the latter. We are differentiated inasmuch as we are the largest data management provider out there. And that scale gives us the ability to spend more money than our competitors can who have point products and much smaller scale, on all the connectors and API updates and things that you need to do to be both natively and second-party integrated with all the platforms that are out there. It's not just the top 5 now, it's OCI, Oracle. I think we're going to go with them. We're their native default option, and that's a nice growing part of our business. So we're able to continue to be the Switzerland of data handling the multivendor, multicloud and hybrid on-prem needs of our customers because of our scale, because of our leadership, because of the fact we've been in the business for 30 years, and that differentiates us from our competition. But it doesn't mean that we're dependent upon that relationship or we're dependent upon any one of those. It's not like this partnership with x hyperscaler is driving our growth and we're leaning on it. It's a balanced growth across the ecosystem providers. We provided some data about that in our Investor Day to convince people that actually we are growing at or faster in Snowflake workloads, Databricks workloads, OCI workloads than they're growing their core business. And we depend just as equally on the partners, the GSIs and the people who help us go to market. So it's balanced across that in terms of what is contributing to our growth.

Helpful color. And just another dimension to it. Cloud subscription customer account has been rising around, I think, 10% the past few years. And the last time you updated us was in December of 2023, that new customers accounted for more than 1/3 of trailing 12 months new cloud ARR. Is there anything that you can share about trends year-to-date in 2024 in terms of quantity and quality of new customers and the contribution that they're adding to new ARR?

So if you look at our customer count, as disclosed in our periodic filings, it might make you scratch your head because it doesn't look like our total customer count is going up very much. And sometimes it's flat, 5,000, and something, I don't even memorize the number. What's going on is because we have these 3 parts of our business, the cloud, which is all we sell, that's growing and generating new customers and then the maintenance and self-managed subscription bit, which we don't sell anymore, and therefore we lose some of those customers, the net doesn't look like an exciting number. And what's happening under the covers is that we're gaining new cloud customers, net new logos and those are at an ASP of X. And those that we lose are small ASP maintenance customers, by and large. So it makes the customer count look wonky. If you look just at the cloud, that 10%-ish number, I think it was 11% when we disclosed it in December, is still about the right number versus our total growth of 35% in the cloud. And that's a function of the fact that we're already -- we already have great market share with the Fortune 2000, which is the primary part of our market, 65% market share or something like that in terms of logo count, although that's across our business, not just cloud. And the net retention rate is so good once they get on the cloud, that once we land them, they expand themselves at a 125% rate on an end-user basis. Those new customers are important, but they're a balanced part of our growth, that growing with existing customers and their needs is just as important to us as landing new logos.

Got it, Mike. And that's maybe to my next question. One of what I think is a really encouraging sign around your cloud-only IPU consumption model is Cloud NRR, which has been resilient at a multiyear high of, I think, 119% for the past few quarters at an entity level and actually still rose at an overall parent level. Can you unpack the drivers of cloud NRR? How much is net new use cases or starting to use additional modules versus increasing usage of existing use cases versus maybe benefit from migrations to the cloud?

Yes. So I won't be able to give you specific numbers, not that we don't know it, but more detail than it would be prudent for me to put out in the public...

Too much information isn't always good.

But it is -- and it's probably not a terribly satisfying answer, but it's all of the above, that we shared actual data at our Investor Day in December, that a new customer to the IDMC, Intelligent Data Management Cloud, after they were with us for 6 months is using on average 3 services on the IDMC. We've got a platform, single pane of glass, IPU consumption based. And you can just use services. You don't have the new provision, you don't need a new token. You can just start using it, which is one of the reasons why the net retention rate is so good, 3 services on average after 6 months. After 18 months, they're using double that, 6 services. So they are finding new ways to use the platform and new ways to consume IPs. So that's part of it. Part of it is the fact that many, if not most of these new cloud deployments are the first step or one of the first steps of a broader transition to a cloud-based architecture for their data and analytics and operational backbone. And so the first cloud land is we're moving this workload or this portion of workloads. And once it works, now it's time for the next workload to move, and we're there because our platform solves everybody's needs because it's so broad. And the third one is not just expansion of a number of services, but we found that in term, because of the telemetry we have, we can see what IPU usage, how it's going on a minute-by-minute basis for all of our customers. We have algorithms and inside salespeople that use those algorithms to determine who every day is using a lot of IPUs and who deserves a call from an inside sales rep to say, Hey, it looks like you're getting great value out of the IDMC. Tell us about the use case? How's it working? Sounds like you could use some ideas? And all you have to do is tell me and it's done. You don't need a new contract or -- and the price is already set. And that's been a really exciting part of our business over the last 6 quarters, and it's growing very, very well.

Cool. It sounds low friction.

Very low friction, very low cost and high customer sat.

Great. Mike, the next driver I wanted to talk about is migrations, where you've seen great progress with PowerCenter Cloud Edition. But once customers are migrated, what's the hook to get them over on to fully managed IDMC? How difficult of a process is that? And what's the advantage for Informatica?

So just to make sure everybody is on the same page, PowerCenter Cloud Edition is a new way to migrate yourself if you are a PowerCenter customer, which is the name of our legacy on-prem data management, primarily ETL that Informatica used to invent the category of data management back in the '90s. That's all -- that's what the bulk of the maintenance is based on PowerCenter. Moving that to the IDMC, which does all the same stuff and more, requires unscrambling the omelet that you have now with PowerCenter that's working great, has been working for 20 years, and re-scrambling it on to the cloud. And it takes some time, and there's some technical risk with it. And it's just a fact of life that you could have hundreds and hundreds of pipelines with thousands of transformations that you've built and maintained over the years. And so you got to get it right. PowerCenter Cloud Edition is something we introduced in August of last year, that instead of requiring a full lift-and-shift migration from your on-prem PowerCenter state to the IDMC with rescrambling everything, doing the user acceptance testing, picking a weekend where we're going to shut everything down and turn the new stuff on, it's taking IDMC and putting on top of your existing PowerCenter infrastructure as a control plane. And so we built all the connectors and all the functionality, so that IDMC reaches down into your existing on-prem environment and runs it through the IDMC with all the services, all the drag-and-drop capabilities that you have from the IDMC. But what it's talking to on day 1 is your existing PowerCenter. You don't have to change anything in the underlying infrastructure. So that makes it very quick and very low risk, and enables the customer to move their pipelines, their on-prem pipelines to the cloud and workloads when they're ready. And so it's kind of the best of all worlds for them. And to the point of your question is, okay, you've got PowerCenter Cloud Edition IDMC sitting on top, so their -- they're a cloud customer for us. It's fully ratable revenue. We get an uplift in terms of what they were paying because it's cloud and SaaS and so it's more valuable. But they still have this PowerCenter running under the covers. When do they move that? And will they -- some get stuck. We're finding -- nobody does PowerCenter Cloud Edition because they might move their PowerCenter. They do it because they intend to. So we're not seeing people just leave it. But we're seeing them take their -- the pace at which they move the underlying stuff does vary. But from our perspective, as both the vendor from a technical standpoint and from a financial standpoint, it doesn't matter because they're paying for us for the IDMC. They're getting value out of it. It's ratable cloud. It is ratable cloud revenue. It just happens to be that the stuff that it's talking to is on-prem stuff that they're running on their environment, just like if the IDMC was talking to non-PowerCenter stuff that's on-prem that we also manage in a hybrid environment. So it doesn't create like a long R&D technical tail cost for us. It doesn't create a revenue rec issues. The customer has the IDMC, so they can expand themselves to other use cases. So it is not a risk, and we're happy to let people move at their own [ path ].

Fair enough. I think last quarter, you talked about optimism for even better migrations into 2025. Why is that?

Well, PowerCenter Cloud Edition is a big part of it because it reduces from what was a 18-month to 2-year cycle to do that lift and shift to 3 to 6 months. And it does it for the reasons I described, with a lot less technical risk and a lot more flexibility. So that's a big part of it. 80-plus percent of our migration deals that we're signing today are PowerCenter Cloud Edition as opposed to the old lift and shift model and it will converge to 100% pretty soon. So that's a big reason for it. And the second reason is that one of the primary compelling events for a customer to move to the cloud -- to the IDMC, our cloud is because they're moving the rest of their data and analytics estate to the cloud. Like I said, if it ain't broke, a lot of people don't want to fix it. And that's the case for PowerCenter in a lot of cases because it runs like a top, it's been running for the last 20 years. And unless I have a compelling event, I'm going to just renew it for another year and think about it next year. And so that compelling event increasingly is to get ready for AI, to use modern tools, to use modern lake house architectures, I need to move everything. And so that's what catalyzes it, and the number of customers that are hitting that milestone is increasing at the same time that we're making it easier to move with PowerCenter Cloud Edition.

So with all of that momentum and goodness that customers get in making that move. But I get your point of if it's not broke, there's the -- inertia is a powerful force. But with all that, you still only have about 6% of the non-cloud base that has migrated to date. Why not get more aggressive incentivizing migrations to capture the associated 2:1 ARR uplift that you get and the better expansion dynamics once they're there? And maybe to give customers access to CLAIRE because that to me, I assume, is the ultimate goal, not only the uplift, but the expansion, the retention and value realization that the customer ends up having, is it possible to maybe do something along the lines of like, let's say, what an SAP has done in helping to use maybe a little bit more stick along with the carrot. How do you guys think about that?

It's possible, for sure, and it's something we think about regularly, how balancing the carrots and sticks to encourage migration. But we think about it, it's not exactly this, but it's all about the net present value of that customer. And being too aggressive with particularly the sticks can reduce the net present value of that customer big time. It's got to feel good in that period where they reluctantly move because you end of life the product or you did some confiscatory price increase on them and they'll move because they have to, but they're pissed. And they're not going to net expand with you and they're going to look for other vendors. And the long-term value of that customer goes down at the expense of the short term. And we don't want to do that. Likewise with carrots, right? We could lower price, so that it's even cheaper, and we take less of a 2:1. And we might do that. That's less risky in some sense than the stick approach in the long term, but it also sets them up at an initial price point that isn't as attractive if we did not offer that care. But there's ways to fix that. So we do think about it. But for now, we are very happy with the mix of migration-related growth, which is about 1/4 of our net growth per quarter in the cloud, and 3/4 of it is winning in the marketplace net new customers and workloads. That's a good balance for us. The migration is growing a little faster than the net new stuff. So that may grow to 30% or 1/3 over time. But that's plenty and strikes for now, we think, the right balance between customer satisfaction and net present value to the customer.

That makes sense, and it's logical. With respect to time, I want to hit on the third and final driver, and I hope you appreciate my framework of kind of stepping through and going through the different drivers. And to me, it's the most exciting, maybe we shouldn't have saved for the last, but it's AI. CLAIRE GPT has been out for a few months now. Can you just update us on any usage trends that you're seeing, early feedback, maybe even the response from prospective customers trying the free promotion that's out there?

So again, let me frame the question in the way that I think make sure that everybody stays with us.

I like how you do that.

GenAI for Informatica has 2 components, which are pretty different, both in terms of what they are and their impact on our long-term financial success. There is GenAI from Informatica. That's CLAIRE GPT, that is our in-product GenAI natural language interface that allows the users of the IDMC platform to sit down in front of a natural language, query it for data discovery, data exploration and operationalizing that data through data pipelines, data quality rules, data governance. It's in our product. It makes our usual products better, faster and smarter. It helps us win that business versus competitors who don't have it. The financial contribution to that is going to be small. Probably, you won't notice it because the goal is not to make money off CLAIRE GPT. The goal is to win more business because it's there and to enhance the productivity people use it so that they can find more ways to use the platform that we have. The second part and the more impactful part of GenAI for us is Informatica for GenAI, that's using the IDMC platform to provide the data management you need to have your data ready for GenAI. Our tagline for our GenAI strategy is, everyone is ready for GenAI except your data. You hear that same tagline in one way or another from the third-party analysts out there, you're hearing it from customers, and it really is true. Gartner had a webinar 2 weeks ago where they talked about the data aspects of GenAI. I would encourage you go to it. They said, look, to be -- to have your data ready for productive and fruitful use of GenAI, you need to do 3 things. You need to invest in metadata management, you need to invest in data observability, and you need to invest in data governance. We couldn't have written that slide better. That's what we do. And that doesn't require CLAIRE GPT in our product. CLAIRE GPT makes you do those things more efficiently and be able to do it with natural language as opposed to drag and drop, which is what you would do otherwise. But the core functions that Informatica, IDMC provides for data management today and for the last several years are what you need to get your data ready for AI and to operationalize it. And we have customers today, we talked about them on the Q2 earnings call, who are running POCs doing exactly that, and it's not yet a meaningful contributor to revenue because those POCs aren't in production yet. And -- you tell me when they go into production, I'll tell you how much the revenue contribution is going to be. But we're -- we feel like we're in a very attractive place in the space, in the scene, if you will. And it's not because of CLAIRE GPT, per se. It's more about Informatica for GenAI and what our core capabilities can do to get you ready.

Maybe just with respect for the few seconds here we've got left, maybe we can get a quick one in here. Since the start of '23, you announced 2 separate cost actions that have helped drive nearly 10 points or so of operating and free cash flow, margin expansion. As we start lapping the impact of these in the next few quarters, how should we think about the levers and pace of expansion ahead to bridge back towards your medium-term 2026 targets?

Yes. It's a good question to end on, and I can be more concise than I was in some of the other answers on it.

Stay tuned.

2023 -- yes, we'll see. 2023 was a year of step function improvement in operating margin because we end-of-sale'd the on-prem products within the year. So we were able to take out in 2 separate actions a lot of the duplication that existed in the cost structure in go-to-market and R&D because we were developing to and we're selling both on-prem and cloud. So 2 big step function actions that will land us at north of 33% operating margins by the end of the year, hundreds of basis points improvement. From here on out, we're not planning anything like that. It's going to be more linear enterprise software operating leverage economies of scale in sales and marketing and R&D. So going forward, margins are going to improve, and you can see that in our medium term guidance. But it's going to be more linear. It's going to be scores of basis points per year as opposed to hundreds of basis points or percentage points per year.

Cool. That makes sense. With that, we're out of time. Mike, thank you so much. It's always great to see you, even better here in Dana Point at the DB Tech Conference.

Thanks a lot, Brad.