Informatica Inc. (INFA) Earnings Call Transcript & Summary

Hey, everyone. I'm Pinjalim Bora, software analyst at JPMorgan covering SMID cap. Delighted to have here with me, CFO of Informatica, Mike McLaughlin. Mike, welcome to the conference.

Thanks very much. Thanks for having me. The turnout of engagement here is really best in class. It's a privilege.

Yes. Apparently, we are up 20%. So 20% grower. So...

At scale.

At scale. But maybe start off with a little bit of introduction about yourself and maybe talk a little bit about Informatica, for the people in the audience who might not know about the story.

Well, sure. Myself, I've been CFO for 1.5 years or so and had a long background in the banking side of things before coming over to be a CFO. Informatica is data management for the enterprise. So our infrastructure software allows the modern enterprise to discover what data they have, to connect their sources and targets of data together with data pipelines that include transformations and data quality and other enhancements so that you can actually get good use out of your data either for analytical or operational purposes. And it enables you to govern your data to make sure right folks have access to it and certain folks don't have access to it. You understand your data's lineage, you understand its providence. And then ultimately, through our MDM products, allow customers to generate a single source of truth for their customers, their suppliers and their products. Informatica invented the category of data management, I think it's fair to say in the mid-90s, and was a public company for many years through 2015. Company went private with Permira and CPPIB in 2015. And the goal of the go-private was not to slash costs and cut off unprofitable divisions, it was to transform the company from its pole position in data management at that time, which was an on-prem world with license, maintenance software called PowerCenter, which was the industry standard, and we still have thousands of customers running it today, but to spend $1 billion-plus in R&D to build a brand-new ground-up, cloud-native, multi-tenant data management platform that would solve the needs of the modern enterprise. It's really important to understand that what we have today is not repotted PowerCenter. It is brand-new stuff that was intentionally built to provide the best data management products across all of the categories from data ingestion, data quality, data governance, master data management. Ask Gartner Forrester or IDC, actually product category by product category within data management, we have the leading products. To provide those on one cloud native platform, single pane of glass, click to utilize the services, pay for those services using a single fungible consumption-based token called the Informatica Processing Unit. And to do that in a way that supports all the multiple vendors that an enterprise have, the multiple clouds that modern enterprise use and the hybrid environments that enterprises have today and will for years and years and years. So you put those 3 things together, the best products delivered on the industry's only true cloud-native platform, serving the multivendor, multicloud and hybrid needs in the enterprise is what makes Informatica truly unique today. And before you get to your next question, just by way of introduction, the final thing I'll make sure everybody understands is that because of the fact that -- because of our history that we were the leading and, in fact, inventing a company for on-prem data management, and we transformed ourselves through a lot of money and a lot of work into the leading cloud data management company, we're a transformation story and to our financials have an on-prem component. We've got about $1 billion of ARR from maintenance from on-prem perpetual licenses that we sold in the past, don't sell anymore. And we've got another $0.5 million of subscription revenue from on-prem or self-managed subscriptions that we also don't actively sell anymore. That's not a bad thing. It's just they don't grow. We've got thousands of happy customers that are candidates to migrate onto our cloud platform in the future, but we've got roughly $1 billion of ARR. That's going to be perpetually shrinking as we migrate those customers to our cloud platform. But we have $600-plus million of cloud ARR that's been growing at north of 35%, and we're guiding to grow 35% again this year. You sum all of that up. We're now at the phase in our transformation where the size of the growing piece, the 35% growing cloud business is big enough to counteract the slower decline of the legacy on-prem stuff. And our revenue growth rate and our top line ARR growth rate in total is beginning to reaccelerate this year, and we expect it to continue to accelerate into the high single digits and the double digits in the next 2 or 3 years. So I'll stop there by way of introduction.

That basically answers my second question. So I'll just skip through because that was the transformation question. You did a great job. But let's talk about macro then. It seems like we're still -- it's still coming up in conversations, macro still sluggish. What is your sense of the spending environment at this point from talking to people? What's the business confidence that customers are kind of showing at this point?

It feels pretty stable to us, Pinjalim. If anything, there's perhaps some shift of budget from what they thought they were going to do in '24 to what they're now being told by the industry they're supposed to do, which is start exploring generative AI. But it does kind of feel like it's a zero-sum game at this point, at least from our perspective. it's not getting any worse, it's not noticeably better. Sales cycles feel like they're about the same. It's a pretty positive environment overall, and it has been for the last 18 months, I would say.

Yes. So let's go on to the generative AI/AI aspect of things because I feel like Informatica is sitting in an interesting position to help people, help companies modernize their data estate, right? And I was looking through some of the transcripts. And Accenture, mid-last year said, 5% to 10% of companies are ready with their data infrastructure and AI capabilities to be able to take advantage of GenAI at scale, 5% to 10%. That seems very low, right, overall. And Informatica is kind of at the pole position to help the 95% of companies to modernize their data estate and be ready for the GenAI side. How would you characterize kind of the readiness of enterprise data at this point from talking to customers?

I couldn't give you an exact number, but that certainly feels consistent with what we're hearing from our customers. And it is an awakening for many of them to admit to themselves that, that is the roadblock or at least among the key roadblocks that they face today. We have numerous customers that are using our cloud platform, the IDMC, Intelligent Data Management Cloud, to do the data management piece, which is getting the data into the model with the right quality control, with the right governance, with the right lineage, with the right traceability into their AI models for use cases that they're doing testing and POCs for that they hope will be operational in 2025 and beyond. It's not a direct revenue generator quite yet, but it's happening and it will be. And how much will be in '24? I wouldn't want to predict, and our expectations are modest. But it is a real road block. People are getting the joke and they're realizing that we're -- I mean, not to be overly self-promotional but we're kind of the only game in town that can give you the best products you need on the only platform that can do it all from a single pane of glass, and do it in a way that allows you to access all of your different sources on all the clouds that they sit in and even your on-prem environment if it remains there.

With the inflection on the infrastructure side that we are seeing on AI, right, NVIDIA adding, whatever, $30 billion in last year alone, that has to permeate into software. It feels like data probably is the first layer where people start making -- putting that into kind of the structure they want, wouldn't you say? I mean, does it feel like that? I understand the revenue is not there yet, but the customer conversations are happening?

They're definitely happening, yes. It does feel like that, but any -- the hype cycle, right? There's the trough of disillusionment that will come at some point, and I don't want to get over our skis in terms of putting specific expectations out there, but directionally very consistent with what we hear.

So you talked about -- I think Ash was talking about growth drivers this year. You talked about ongoing data-driven digital transformation, cloud migration from PowerCenter and then GenAI. It almost feels like GenAI is feeding into the #1 and #2. Is that true? Is GenAI actually allowing or pushing your customers and PowerCenters to start modernizing even faster? And now you obviously have PowerCenter Cloud Edition as a tool to help you do that. But does that feel like GenAI is helping you do the first two?

Yes. So let me -- maybe to answer that, I'll step back for a minute to make sure everybody's got a base level understanding of how our business financial model is working. So if you think about the growth part of our business, which is our $600-plus million of cloud ARR out of our $1.6 billion that we've guided to grow 65%, 35% this year, we disclose on a last 12 months basis, we started doing this in the third quarter of last year. How much is from -- how much of that growth? How much of our NARR is from net new workloads and/or new customer? So new workloads with the existing customers or new-new customers versus migrations. So how much is coming from this $1 billion base of on-prem coming over the transom? And how much is -- are we winning net new in the marketplace? And right now, that sits at 78 -- 72-28, 78% net new, so 78% of our last 12 months cloud NARR is from net new customers, net new workloads and the other 28% is migration from our existing on-prem base, whether it's maintenance or self-managed subscription. And we think over time that, that's going to be probably a 2/3, 1/3; 2/3 net new workloads and customers and 1/3 migration. By the way, we also disclose what the uplift is or the flip multiple that we have enjoyed from moving a customer from the on-prem maintenance or subscription to cloud. And it's been up to now, I guess, we closed the last 12 months basis. It's been 2 to 1. So if we have $1 of maintenance and we're able to migrate the customer to the IDMC, the cloud platform, it turns into $2 of cloud ARR. So those are some of the basic drivers and facts and stats that will help you think about how to model the business. So when we think about the growth of that, total ARR of the cloud business growing at 5%, so as Pinjalim mentioned, we have 3 primary growth drivers that we point to. One is migrations. That's the 28% of the NARR that I talked about. And then the other 72% or 70% over time is from new workloads and new customers. So for that, digital transformation is the primary driver, and it has been for years, and it probably will be for years. So there's just tons of it to do out there as companies modernize and they find new things to do with cloud infrastructure and cloud tools as it gets more powerful. But we also highlight as a separate tailwind GenAI. But you're right. It absolutely feeds into the growth of new workloads. And it is driving to some extent, the migration from on-prem to the cloud. And you hear that from some of the hyperscalers recently as they talk about what appears to be a GenAI-driven acceleration of migration. So those 3 things all add up to what we think is a pretty healthy picture to keep that pace of growth going consistent with our 3-year expectations.

Yes. On the GenAI conversation -- I'll move away from GenAI soon. But on the GenAI conversations, what are you hearing specifically, which layer of Informatica is being used, right? Is it the core data integration is the MDM side? Is it mainly in the data quality or data governance coming into the picture? Where are the conversations mainly? Is there a concentration in one of these buckets?

Yes. And at the risk of doing it again, let me step back a little before I answer the question, so people also understand the context. So AI in the context of Informatica has 2 components. One is AI from Informatica, which is tooling that we are announcing -- actually, we went GA last week, and we have a lot of demos and hoopla about it at Informatica World this week in Las Vegas. If anyone wants to catch a flight, I think there's still seats available. That is our own generative AI tool that sits in front of our IDMC platform that allows you to use natural language. It takes advantage of the metadata system of record that we have for an enterprise's entire data estate to make you -- to enable you to use the IDMC faster, more efficiently, more cheaply, turning weeks of works into hours, if not minutes, to fine data, pipeline it quality control it, govern it, turn it into usable stuff. So that's AI from Informatica. And it's CLAIRE GPT and it's the tools that we embed in our products. Then there's Informatica for AI. That's the governance that you're talking about and the data quality and the data movement that you need in order to operationalize your data workload. That doesn't per se, take CLAIRE GPT. That takes all of our tools like data integration, like quality and governance. So finally, getting to answer your question, it really is all the above. So on the core data integration and ingestion, we're announcing a number of features at Informatica World that not only serve the nonstructured data needs for GenAI, whether augmentation or training, but also serve the unstructured data in a way that we've not before with text files and video files and audio files and that sort of thing. So just getting the data into the right place with our tools is a key thing that our customers are using. And by the way, we also can integrate the model. So one of the things we'll talk about is how we can actually be a catalog of models. And so when you -- here's all the data and here's how we want you to link it in and drop down menu, here is the model that I want to use it with. So it's that fundamental. And of course, the data governance issues and the data privacy issues are paramount as well. And that's another aspect of the challenge that companies are coming to, particularly recently as a real critical path and our customers are turning to us for that.

So good that you explained. So Informatica for AI and AI that Informatica is offering, CLAIRE GPT. But what has been the feedback on it? It's not GA yet, the CLAIRE GPT and what do you call it, the copilot. It's not GA yet, but the early feedback from customers, are you seeing that drive consumption up from those early customers?

So it was GA on May 8. One demerit for not knowing that, Pinjalim. But we're really rolling it out this week in Informatica World in earnest. We had over 200 private preview customers. So we got a lot of feedback, and we got a lot of real-time usage from them and that all informed the final state of the product. And the feedback is really great. And a number of them will be on stage over the next few days to talk about their specific experience. So what's it going to drive? So CLAIRE GPT, we are going to charge for it. We charge for it like any other service. The way our platform works is that you log on, single pane of glass, you've got 34 services, of which CLAIRE GPT is the 34th. And each one of those, you can consume with your IPUs, your Informatica Processing Unit. And it takes a certain number of IPUs to certain number -- to move a certain amount of data, a certain number of IPUs to do a certain number of transformations, catalog entries, et cetera. It takes a certain number of IPUs to do a CLAIRE GPT query. Now it's small, and we have priced it so that it covers our costs in a way that isn't going to be -- isn't going to hurt our gross margin. But we're not trying to lean on it for financial growth, CLAIRE GPT itself. Rather, we expect it to drive the ease of use, the power of the platform. So we win more deals when we get in front of prospective customers, and we show them what they can do with our platform that they can't do with others because they don't have the CLAIRE GPT tool. And they're going to do more with it. So they're going to build more pipelines. They're going to establish more use cases they're going to catalog more stuff. And so it really is that usage of the IDMC that we expect it to drive.

Yes. Let's talk about the migration. So $500 million or so of PowerCenter -- legacy PowerCenter migration or maintenance ARR that you have. You introduced this thing called PowerCenter Cloud Edition 9 months ago, about?

It was in August of last year.

Yes. And it seems like it has clicked and you're seeing more of the cloud -- the migration deals being driven by PowerCenter Cloud Edition in the last 2 quarters. I don't know if it's a fair question for a CFO, but what is the difference? Why has it kind of picked up steam? Is it much more automated, much more easy to use? Help us understand why that maintenance cadence has picked up because of that?

Our investors since going public have -- they always ask why are migrations from maintenance to cloud not happening faster because it's -- we talk about how great the platform is and all the stuff that it can do, which it can, and that's all true. So why aren't people just beating down your doors to do it? The reason is it's very complicated. With a PowerCenter estate that has been running for 10, 15, 20 years, you can have hundreds of pipelines with thousands of transformations and custom code that's been written to manage this particular source or target that isn't supported by its vendor anymore. And you used to have to before PowerCenter Cloud Edition. You used to have to drop a systems integrator in there, one that's been trained by us, using our tools, which are pretty darn good. But still, you have to unscramble the omelet and you've got to rescramble it for the cloud, and then you've got to test it and user acceptance it. And then over the weekend, you flip the old stuff off and you flip the new on. And that's just a big consequential undertaking for anybody, and you only do that when you're darn good and ready and you have the budget for it. PowerCenter Cloud Edition flips that on its head in that what we do is we take -- we figured out how to integrate the IDMC with your existing PowerCenter estate. In layman's term, which is what I am, it sits on top of PowerCenter and it reaches down in and it controls, manages and governs and does everything that is going on today. But you do it through IDMC, and it gives you all the power. So it's not just a user interface. It gives you all the features, all the power of and all the pricing and packaging of the IDMC, and you can move your individual pipelines or your individual transformations when you're ready. And the appeal to customers is that I don't have to move all my stuff right away. It's my intention to move them. People don't do this if they don't intend to move because it's not worth that. And I get all the power of the additional stuff that I get with IDMC that I don't get with PowerCenter today. And they have the ability to expand it to new use cases because it's all the same pricing model, et cetera, et cetera, et cetera. So what used to take 2 years can now be done in less than 6 months. And it's as you say, we have -- volume of those migrations is doubling year-over-year every quarter for the last 2 quarters and we expect that to continue.

You expect that to continue. So as that $500 million kind of shrinks over time and you get that extra bump on the flip multiple to cloud ARR, is there a place in that maintenance era that will never shrink below that? I mean there are probably some regulated industries that will never go to cloud, right? How to think about that floor in maintenance ARR over time? I'm not talking about this year, but steady state?

Yes. I wish we -- it's the right question, and we ask it of ourselves, and we haven't figured out how to answer it yet because we don't know. But you're right, it will never go to 0. And at some point, you can end of life PowerCenter and really force the issue, but we're nowhere near doing that. I don't think we're going to -- don't tell our customers we're end-of-lifing PowerCenter. We're not. But if you step back and you look at the medium-term plan through 2026 guidance that we offered at our Investor Day in December, you can, with respect to the migration piece and the pace that we committed to delivering between now and the end of 2026, we're still going to have 2/3 of our on-prem base running and paying us at the end of 2026. So $1 billion is going to go down to something less than 700 -- between $600 million and $700 million -- between $650 million and $700 million. So there's still going to be a lot left, and we still don't think we're getting anywhere close to that kind of asymptotic hard deck of the people who don't want to move.

Yes, understood. Now moving on to the other 500 million, which is your self-managed customers. We have not seen those migrations still now. It seems like in the last couple of quarters, sequentially, that has declined a little bit. We are seeing some of those migrations happen. But trying to understand what is driving that motion, because that is still -- those customers are using your new platform, but self-managed way, not cloud-managed way or Informatica managed. What is driving that motion? And the flip multiple there is a little bit different, would it be not?

Yes. It is -- we don't -- frankly, we don't have enough data points to have a really statistically significant number on that, to be honest, difference in flip multiple self-managed versus -- it should be lower, if you think about just the fundamentals of maintenance pricing versus subscription pricing. So directionally, it will be lower, but it's going to be positive. I mean we're not going to give it away unless it's where you're replacing shelfware or something. Those -- it's motivated largely by the same things as the PowerCenter migrations, except for the fact that just the makeup of the self-managed subscription base is a little different than the maintenance base. The maintenance base is sort of 70% PowerCenter and then you've got some MDM on-prem maintenance and some governance and catalog. In subscription, you've got more on-prem governance, our Axon product. You got more on-prem MDM versus PowerCenters. So stay tuned, watch this space, watch the Informatica World transcript. And there, we are offering new tools real time this week that are going to make the MDM migration easier, the governance migration easier. So we expect those to add fuel to the fire.

Understood. So going into the ARR equation. So you have 3 different buckets. You have cloud ARR, self-manage ARR, maintenance ARR. The cloud is obviously driven by -- partially driven by some of these migrations, but largely 70-some-percent is driven by new. Help us understand the sustainability of that new part, right? How should we think of that? Your -- I think in the Analyst Day, you were talking about the TAM growing somewhere around mid-20s, high 20s kind of a growth rate. Is that how to think about of the new portion ex migration?

Yes. Yes. While it may not be setting a high enough bar for ourselves, we're committing to growing at least as fast as the TAM with that -- with respect to the net new portion. So you can see how we build up the TAM and you're welcome to do your own TAM calculation. But the way we do it with IDC data shows that the data management total addressable market through 2027 growing in the mid-20s. And so we think with the products that we have and the platform that we have and the Switzerland of data position that we have, that we should be able to grow at that TAM for a good long time even at scale. And then the remainder of the growth rate that we have committed to over that medium term is for migration. And again, that's likely going to be 2/3 net new and 1/3 migration to get there.

So theoretically speaking, if the -- theoretically, if the cloud ARR growth from new stays constant and you're getting equal amount of kind of new -- net new logos, new workloads and you have migration pick up, shouldn't cloud ARR actually accelerate for you going forward at some point?

Hypothetically, under that set of circumstances yes, mathematically. But look, we are getting to be an at-scale cloud company, right? And I forget our numbers in 2022 in terms of cloud growth, but it was close to 40%. In 2023, it was 37%, and we're guiding to 35% this year. And if you look at our growth for -- that we guided to for the medium term, that is through 2026, we said 31% to 32% CAGR. So yes, it's -- we're being prudent about the natural growth at scale that every software company faces. But yes, if migrations continue to go through the roof and we continue to punch above our weight in terms of TAM, which is certainly possible, your math is right.

Yes. Let me wait and see if anybody has any questions from the audience. No? There's one there.

Could you help us understand those who are not very familiar with the story, how does your company differ versus Databricks, MongoDBs of the world? And when we hear different companies talk about it, they all are saying the same thing. We have a one-stop shop. We're going to put GenAI on top of it. We can do all these pipelines and extraction. They make it sound so easy, but it can't be that easy. So what are some of the pain points that you face even now as you are such a mature company?

Yes, you're right. It can be very hard to parse that and separate similar sounding words. And let's start with Databricks or Snowflake or even Azure. They do have data management tools that are native to and proprietary to their ecosystem. And they have those tools, I would say, not to speak for them, but I think it's fairly obvious, to drive more data into their ecosystem to ultimately drive more storage revenue, compute revenue and the things that they monetize. Good for them. That will always be the case. But we don't frequently compete head-to-head with them because that's not our customer. MongoDB, they provide the database infrastructure per se for using data in a cloud environment. So they have some tools to get the data into their database and some rudimentary quality control and so forth. And that's great. No -- we don't object to that or lose too much sleep over that. The customer we compete for and where we win is the customer that, as I said before, needs the best data management products across all the categories of data management, data ingestion, data quality, data governance, data lineage, master data management without compromise. So you're not getting B+ products versus the shiny new venture capital-backed Silicon Valley object. Our products, ask Gartner, Forrester, IDC, are just as good or better than those category by category and who have value from a platform that integrates it all together under one consumption-based pricing model and is the Switzerland of data that can serve your multivendor needs, your multicloud needs and your hybrid needs. So Databricks doesn't do that. MongoDB doesn't do that. Snowflake doesn't do that. I can name 27 different companies that are in our space and talk about it the same way that don't do what we do. For those who have a simple need and all they need is ingestion, there's 3 or 4 different companies that they could use, and we don't spend a lot of time competing for that customer. For the customer who's willing to live inside the Databricks ecosystem and not spend too much attention about their multicloud data infrastructure, we don't spend too much time chasing that customer. So where we focus, there's tons of opportunity, and it really is once you get to that level, quite distinct from what they're doing.

Anyone else?

[indiscernible].

So there's no technical -- the question was, are we just limited to large enterprises? Or are we -- can we scale down to the SME, SMB layer? There's no technical reason why we can't. But the enterprise is our sweet spot, and it's where we focus. We have essentially 100% direct sales effort augmented by partners. One of the statistics we talk about to emphasize the importance of our partnership network is that north of 30% of our bookings in any given quarter are sourced by partners, mostly GSIs and regional systems integrators. But the hyperscalers like Snowflake and Azure and Databricks are also sourcing 1/3 of our business in total any quarter. So super important. But it's all direct. I mean we -- you can actually go on our website and you can buy some IPUs and you can start doing data ingestion and -- if your heart desires, but it's enterprise.

Anyone else?

Do you see Microsoft as competition with their AI copilot Azure [indiscernible]?

So no, and here's why. And thanks for asking that question. So the question was, do we see Microsoft as a competitor with their AI copilot and their generative capabilities that sits on top of their Azure Fabric and otherwise? So CLAIRE GPT and CLAIRE copilot, which are our GenAI tools in the product, so GenAI from Informatica, are really good and powerful. But frankly, there's nothing particularly proprietary or moaty. It's proprietary because we developed it and we had to integrate it, but it uses the open models at the end of the day, private instances of them so that our customers' data isn't getting into the wrong hands. But it would be replicable by Microsoft or anybody else, except for the following fact. A key asset that we have because of where we sit, because we are the platform, data management provider for our customers is what we call the metadata system of record for our customers. So our platform knows where all your data is, what each data source has inside of it who created it, who should have access to it, what the quality issues are, where it's going, what its lineage is. And that's what CLAIRE GPT utilizes to deliver the value. It's not the natural language interface so much. While that's important, and it help democratize it, it is that it knows how to talk to our metadata system of record that we have for each and every customer in order to generate the productivity enhancements that I described. And Microsoft doesn't have that, unless you're a Microsoft shop, and you got all your stuff on Microsoft.

Let's quickly touch on 2 topics: your debt and your plan for that; and also M&A.

Okay. And I can probably do that in 45 seconds. So I feel good about our debt level. We're a cash flow generative software company with BB ratings, which I love being there. I was a banker for a lot of years until I thought a lot about tech capital structures, and I think we're in a good spot. We have more cash than we need. So our net debt levels are lower than they could be. But for now, we like the flexibility that it's giving us, but we're not going to keep $1 billion of cash on our balance sheet forever. But we're also not going to go out and spend like drunken sailors on acquisitions, because we have the stuff that we need to execute to our plan. We don't have big holes in our portfolio. We don't have a place where we don't have market share or customer access. We might do tuck-ins like we did last year with Privitar to enhance a capability here or there. But we don't need to consolidate and we don't need to expand our offering in order to achieve our goals.

All right...

Thanks, everybody.

We're out of time. Thank you so much.