Informatica Inc. (INFA) Earnings Call Transcript & Summary

Okay. Thank you, everyone, for joining. I'm Pinjalim Bora, a SMID cap analyst, JPMorgan. Delighted to have here with me, CFO, Informatica, Mike McLaughlin. Thank you for coming.

Thanks for having us.

All right. Let's -- why don't we start with just a brief introduction about maybe yourself. And a few words on Informatica, where it is for some people who doesn't know about Informatica?

Yes, sure. Informatica is more interesting than me, so I'll focus on that, how about that. Informatica does data management. We've been around for 30-plus years in a lot of respects, invented the category of data management back in the days of on-prem data warehouses when that was a new technology. Historically, we were synonymous with ETL, Extract Transform Load, which was how you did data warehousing and data management in the on-prem world. The company went private in 2015. And the thesis of that go private was not to cut cost or to separate the company into separate divisions. It was to transform it to the leading cloud data management company that would address from a single platform built from scratch with a cutting-edge microservices-based multi-tenant architecture the needs of the modern cloud data landscape. Went private in 2015, spent $1 billion-plus on R&D to accomplish that vision and became public again in late 2021 and end of sailed all of our on-prem stuff fully at the beginning of 2023. So sitting here today, we have -- almost $850 million of cloud ARR on the IDMC, the Informatica, the intelligent data management cloud, which is our true platform for data management that does on one pane of glass, one login, everything from data integration, app integration, data quality, data catalog, data marketplace, master data management, all available on a single consumption-based pricing model. Everything else is end of sale as being where customers are ready to do it, migrated from on-prem to our cloud at a healthy uplift multiple. And some of that is a trading when the workload has gone away and so forth. The future of the company is the cloud portion, which is now 49.8% of our ARR, so almost 50%. The cloud portion is going to continue to grow. We've guided it to 25% year-over-year growth this year versus 34.5% last year. At that pace, it will be 58% of our total ARR by the end of the year and continue to be a bigger and bigger portion as the on-prem stuff gets smaller and smaller, which leads us to what we think is a very attractive financial profile in the years ahead.

That's great. So Informatica has gone through a lot of transformation. So let's -- as I think about it, right, you have this cloud business, then you have the self-managed business and then you have the maintenance business. Cloud is you said something around 65% -- or 50% of ARR. But it continues to grow partially by migrations, but a large portion of the growth is actually coming from new workloads, new customers, I think you said 65% of the net new...

The majority of the cloud growth is new customers and new workloads, not migrations or modernizations of our on-prem base. We disclose this every quarter on a last 12 months basis. And in the last 12 months, 67% of our -- I'm getting my IR head is keeping me honest. It's essentially 2/3, 1/3. That 2/3 of our cloud growth is from new customers and new workloads. Winning in the marketplace, head-to-head against every other alternative is out there, winning modern data workloads in the cloud straight up. The other 1/3 of it is coming from existing Informatica customers who have been using us for data management on-prem that are now moving those data management tasks to our cloud.

Talk about the kind of use cases that you're seeing on that new side, right, not the migration part, on the new side because I think -- if I look at the business, I think, in your deck, you had data integration, app integration is like 50% versus MDM and...

Data catalog.

Data catalog and governance 50%. When you're seeing that incremental growth in cloud from new customers and new workloads, where is that coming from? Is that DI side? Does that governance or MDM side?

It's pretty well balanced, Pinjalim, in terms of those three categories. The data governance, which is the smallest of the three of those categories, as we say, we don't give exact numbers, but it's roughly half data and app integration and half master data management and data governance. The data governance piece is the smallest of the three, and it's growing the fastest. But it's not growing 4x the rate that data integration is. So they're all growing at a very nice rate, plus or minus what our overall growth rate is. What we tend to see happening is that the land deal is a straight-up cloud data integration workload that I'm standing up a cloud data warehouse, and I need data management. And then the expand deal is, let's put some governance on that. I want to have 2 data warehouses or I want to have a data lake and a data warehouse, and those are different providers. One is coming from Microsoft. One is coming from Databricks. And Informatica is the only tool at enterprise scale who can manage them both, catalog them both, govern them both, move pipelines from one to the other seamlessly. And then Master Data Management can come before or after that initial data land, but is growing equally well. Master Data Management is a single golden record source of truth about a customer, call that Customer 360 product. So everything you need to know about a product in terms of where it's sourced and what its components are billing material suppliers, that is more like an application than the infrastructure software, but it drags and is tightly, tightly integrated with all the other data integration that you need to do for that, also growing in a very healthy rate.

So you are landing with MDM?

Absolutely. Yes, absolutely, for sure. And those are customers who never used our MDM on-prem.

Yes. Let's talk about the motivation for somebody to kind of do data integration for the first time, right? Because one of the biggest motivation right now is AI. And the theory for some bulls on Informatica would be that data -- enterprise data is kind of in a mess and Informatica is in almost every Fortune 100 company or Fortune 500 company. You should see a lot of kind of workload from that base as they try to figure out the enterprise data landscape and prepare for AI agent and all that stuff, right? Is that coming into the conversation? Are you hearing that from customers at this point?

We're definitely hearing it. It is the prerequisite to having accurate and governed enterprise-class AI to have the data management underneath it sorted. It's not a one-to-one yet where customers are coming to us and say, we have this fully formed GenAI work case -- workload or use case that we know is going to work, and we know it's going to generate this amount of data management need and we need to buy IPUs that's our consumption unit to do it. It's -- we are standing up this proof of concept on this use case. And we're going to use IDMC to run the proof of concept. We have 175 customer orgs that are running GenAI-based workloads right now. Most of them -- almost all of them are still in the development or proof-of-concept phase, not at scale. That's the way the conversations work. And even in the conversations where customers are not explicitly saying AI, that's in the back of the mind of most everybody that we know we have to get our data to the cloud. We have to have it properly integrated and managed before we can use AI and it's an encouragement to keep the pressure on to move forward in the cloud.

Yes. I want to bring the discussion a little bit current. In the last quarter, not this quarter, Q1, but Q4, you had a little bit of hiccup around cloud, retention, right? And the retention numbers dropped, I think, 200 basis points?

It's about that. We're not being accounting precise on that, but it was, call it, 2 percentage points versus what we had expected.

And now fast forward in Q1, you're basically saying, okay, sequentially, it has improved. So maybe talk about what happened, what did you do? And the changes it seems like are taking effect is there a way to kind of understand the magnitude, right? Are we back to the level there was before Q4? Or do we still have kind of room to go?

So the miss -- it was a miss in terms of our gross retention, our renewal rate in the cloud was about 2 percentage points versus what we thought it would be. So it wasn't massive, but it was enough to get our attention and be noticeable in numbers. What happened was not that there was a bug in our software or that people woke up and realized that what we do is no longer essential or that what we do is no longer the best in class out there. It was execution around a handful of deals, many of them sizable where we didn't identify the renewal risk early enough. We didn't involve field sales in collaboration with the New World team, early enough to fix the situation or find ways to utilize the IPUs for new workloads. It was kind of classic stuff that we could do better on, and that's what we've been focused on over the last 3.5 months, and it's working. Q1, as we said, is -- was better sequentially. It's on track with what we thought it was going to be when we guided for the year. But it's the smallest quarter for new bookings. And therefore, it's because most deals renew on an annualized basis. Most people sign a 2-year or 3-year or 4-year, not a 27-month deal. So it's also our smallest renewal quarter. So we're not declaring victory yet. But it's playing out as we expected and hoped and still feel good, therefore, about the year, meeting our expectations.

Okay. So the cloud guidance, when you gave the cloud guidance, I think at that time in Q4, you had said that it's based on a lower level of cloud retention. So now when you're saying it's improved sequentially, you're still on track. It doesn't mean the initial guide, which the assumption that was based on, it's not exceeding that level at this point you're kind of?

I wouldn't say exactly that, and I don't want to get too again, accounting precise on renewal rates because particularly, Q4 is -- Q1 -- sorry, Q1 is our smallest quarter. So because the renewal cohort is small, saying that we're above or below that could just be noise. Sequential improvement at or better than what we had forecast for the year is the message.

Got it. Obviously, when there is some kind of a hiccup for companies, a lot of conspiracy theories kind of come a float. Maybe talk about the competitive landscape, right? Was there any impact on cloud, new cloud specifically, on new cloud ARR. Was there any impact from competition? What are the -- what are you monitoring that makes you -- gives you confidence that there's not no kind of competitive -- competitive-driven outcome?

Not that we can see from the -- either the aggregate data we can pull from our various CRM platforms about why we win and why we lose, neither from anecdotal conversations that we hear from customer interactions. There are competitors out there and we live in a competitive world, but no change from the past. And just as a reminder, we compete against kind of three categories of folks, one of which is the point providers that basically started developing their cloud data management products more or less the same time we did, sort of in the 2015, 2016, 2017 period. And they now have fully scaled but narrow products to do the same things we do. Like there are qualified vendors out there who do just cloud data ingestion. There are qualified vendor who do just the transformation piece. There are qualified MDM vendors, all of whom are private in all these categories -- catalog vendors. But nobody does it across a platform. Nobody does it in the integrated way that we do. And by the way, if you look at Gartner and the other third-party analyst categories, even point by point, our products are ranked higher. So we've got products that are generally better than theirs and we put it on a platform, but we compete against them. They generally would compete on price and they're usually cheaper than us. And that's why we lose to those folks is when there's a simple use case and you only need catalog using the cheaper catalog is -- may make sense for you. But that's not really where enterprise is. The second category are, I don't -- they don't like it this way, but the legacy providers. So there are still big 3-letter software companies out there that have data management products that are widely used and because their incumbency will continue to be widely used for a good long time. But also, if you look at the third-party vendor rankings, third-party analyst rankings of those customers or those companies, those vendors, that their products are meaningfully behind ours in terms of cloud, data management capabilities. And the third category in this, I think is what most people in this room worry about is the hyperscalers and the cloud data warehouse providers themselves. So Microsoft has Azure data factory, which has data management tools, and they're okay. Azure -- AWS has Glue, which are cloud data management tools. Databricks has some tools. Snowflake has some tools. And we do compete against them. But their tools sit within their walled garden. Their tools work great with their stuff. And a varying level of degrees, their tools may be easier or harder to use in terms of -- we have a fully sort of drag-and-drop low-code/no-code development environment, for example, their tools have limited connectivity outside of their environment. They are not a Switzerland of data, like we are. But if you have a simple use case, and you already have a big Microsoft commit to stand it up and use ADF, that could work fine. And we -- there are some deals where that happens. The fact is, though, we don't see those deals and when we do, we try not to waste time competing for because that's not where we add value. Where we add value is in a multi-cloud, multi-target complex enterprise global environment. We're having a platform of all of these best-of-breed capabilities knitted together that works across all your clouds and all your on-prem environments, that's where we win. And there hasn't been -- to get back to your original question of this overly long answer. We haven't seen the intensity of that competition change.

The handful of deals that you talked about or customers -- handful of customers that you talked about and that led to kind of the cloud retention problems. Within those conversations, let's say somebody was using or thinking of using 30% of your IPUs for MDM and I think you said at that time, executive sponsors saying changes and that didn't happen. So that 30% maybe didn't materialize, right? So they reduced the IPU total commit by 30%, let's say. But in those conversations, was there any competitive angle at all or now that you're talking to those customers at all?

I never want to say never, but that was not the mode of the conversation. It was -- the use case did not develop as the customer thought it was going to 3 years ago or the sponsor changed halfway into the process. These are long -- generally frequently long consequential implementations that take lots of -- they're cross-functional, they're cross department, in many cases, M&A. There have been some big splits in corporate America in the last 12 months. And one of them, two of the three divisions that split kept our product and one just thought that they could consolidate it on to something else that division ends up being back at the table right now and is likely to be an IDMC win later in the year because they've realized that actually, our stuff was pretty good.

Yes. Interesting. Okay. Moving on to -- from cloud 50% of the business, you have self-managed, which is 25% of the business and then 25% maintenance. On the self-managed portion of the ARR, that is the decline is kind of starting to be higher going into Q1, right? So maybe talk about those dynamics, specifically, are sales reps being proactively compensated anyway to move those to cloud, is that migration incentive -- a new incentive structure, is there a migration part of it for this year? Is that part of that conversation? And are you seeing any change on kind of the natural churn for that business?

Yes. So there's a bunch of different parts of that question. I'll start by pointing out that the migration or the modernization portion of our business, that's our on-prem customers moving at an uplift to our cloud. That's accelerating. And that part of our business feels stronger than ever. But it's not accelerating because we've changed compensation or given our sales reps free vacation, if they sell a modernization deal. And it's not because we're lowering price on the IPUs or the MDM records for modernization. It's genuinely because the need to modernize and the capabilities of our product once you modernize are compelling. It's not accelerating because of anything we're doing to accelerate it other than just singing the praises of our product. The compensation is a little different between a modernization deal and a net new deal for our sales force. And the nature of that difference is that on a modernization deal, we only give you quota credit and commission on the net new piece of it. So if it's $100 a year maintenance customer that we're moving to the cloud for $160, you only get paid on the $60. You get paid at a higher rate on the $60 than if it were $160 net new, but you're only paid on the $60 so that you've got a lot of incentive to grow the pie versus just milk the base that doesn't generate a net. But it doesn't, on balance, create like an outsized incentive to chase monetization deals versus chase net new.

And that's for all modernizations, including...

All modernizations, including MDM and maintenance. So there's kind of an X-Y axis or 4 box for modernization. You've maintenance that modernizes. You have self-managed, which modernizes. And then you have PowerCenter, which is both maintenance and self-managed master data management, which is both categories as well. But that net new concept is true across all 4 boxes.

And what about the natural churn in that business?

So natural churn is higher in self-managed, for sure. It has been -- we've been clear about that for as long as I've been here anyway. The primary reason for that is that the self-managed, the term license bit is less seasoned. We started selling -- the self-managed term license was a bit of a bridge between our legacy on-prem license maintenance -- perpetual license maintenance business, which when we went private, that was all we did. We didn't even do subscription then. We started the subscription shortly after going private before the cloud product was fully ready. And so those started being sold in 2016 or maybe 2017. So they're just not as seasoned. And there's nothing that predicts renewal better then how long it's been running. If it's running for 20 years. It's like the 20 -- renewing in 21st year is highly likely. That's been running for 4 years, that renewal is much less likely. That's the primary reason in self-managed, so there's more situations where the product wasn't fully implemented or where the use case has proved to not be essential or they consolidate the use case with something else. Those that are going away in self-managed. And with maintenance is not predominantly because they're going to a competitor. That's typically less than 20% of our non-modernization churn is we can see that they decided to use something else but keep the workload. It's a vast majority of it is the workload goes away or is consolidated or some M&A event or something like that.

Yes. Let's go to PowerCenter migration, which is your last kind of quarter of the business, right, maintenance. That maintenance line has been -- I was looking back in the model, has been very stable for 5 to 6 years. I mean, it's going down like 4%, 5%. Now it's kind of picking up pace, right? But it's been fairly stable. And we understand, okay, there's a modernization wave that's kind of driving that decline to cloud. But there is some other mathematical stuff going on as well, which you call the maintenance roll-off, right? Maybe -- maybe talk about that a little bit. And specifically, when does that pressure go away for you, right? When does that pressure release so that we can see more of a natural downtick?

Yes. So just to try and explain this as simply as I can because it gets complicated pretty fast. We have in both categories, maintenance and self-managed. We have natural churn, which is customers that are turning off the software for any of the reasons I described. The use case went away or in some cases, they go to competitors. And then there's modernization churn. Those are customers who are turning off their on-prem but moving to the cloud. You have to think about those very differently because the drivers are very different. The economics are very different to us. Natural churn and the maintenance business has been quite consistent at kind of 4% to 6% a year for a long time. It's consistent with what we're forecasting. It's very seasoned. The average customer has been using PowerCenter maintenance for almost 15 years, average customer. In the self-managed business, that natural churn is kind of twice that, so call it, low double-digit percentage because it's less season for all the reasons I talked about. All of the decline in those two above that is modernization. So modernization is accelerating. So the amount of modernization churn is going to be higher, and that's in our forecast. And it's particularly higher in the maintenance in the short term in terms of the percent of the total decline. And one of the reasons for that is because the way the accounting for the modernization works is that, that modernized on-prem revenue or ARR, doesn't roll off of our revenue stream or AR stream until the modernization is complete. And so if you sign up for a modernization deal, you start paying us for the cloud, and we give you credits for your on-prem, so you're not double paying, but that ARR revenue according to accountants has to stay on our income statement until the modernization is done. So prior to a product we call PowerCenter Cloud Edition that we introduced in the third quarter of 2023, that amortization took 2 years. And so that on-prem revenues stayed on our revenue and ARR basis for that full 2 years of the modernization period. So anything that we signed up to modernize in 2023 is rolling off this year. But then in the third quarter of 2023 didn't start to scale until the fourth quarter, we launched PowerCenter Cloud Edition, which allows you to modernize in 6 months. And if you want, I can talk about how we achieve that magic. But accounting-wise, it means that the modernization is done in 6 months and the roll-off happens to 6 months. So in 2025, in terms of the shrink, the modernization shrink of the on-prem maintenance and self-managed, we kind of have a double whammy, that the 2-year modernization deals that we signed in 2023 are rolling off in 2025 and the 6-month modernization deals that we signed in the second half of '24 are rolling off. By the time we get to the first quarter of 2026, this will be done with all those 2-year deals and so that double whammy will be gone, but we'll still have that 6-month dynamic. So this is one of the reasons in addition to some things like renewal term length and so forth that this is a noisy year in terms of our results for better or worse.

So Q1 2026 is kind of the first period when we should see that alleviate a bit?

Yes. That's right. It will be better. It will be close to normal in Q4 because we -- in Q4 of '23 which is 2 years prior to Q4 of '25, we still offered the 2-year version, but most of those sales were PowerCenter Cloud Edition, which was the 6 months. So it was a non-zero, but not very large number in that comparable period.

Yes. I want to ask you one question on which I'm getting more and more is this notion of uplift multiple being a little bit lower than before. The uplift multiple when you move somebody from PowerCenter to cloud used to be 1.8, 1.9, even 2, right?

Even 2x, yes.

Now you're talking about 1.5, 1.7. So there's a little bit of confusion on why that is happening. And if that's something to do with pricing or competitive, I think there's some conspiracy theories around that. So maybe talk about that a little bit.

Yes, thanks for asking that question. It is an important point. So the process for determining what we're willing to sell the cloud software to an on-prem customer for is not -- you're paying us $100 for maintenance today. We need a 2x multiple. So how about paying us $200, that's not how it works. It starts with what are you using PowerCenter for today, let’s get in there with our technical team and our estimating resources and how many IPUs, which is our cloud token or pricing unit for the cloud. How many IPUs do you need to replicate that workload in the cloud. Once we determine that, you're in this price band. This is the list price. And now let's talk about the discount that we're willing to give you because you're our favorite customer ever. And the uplift multiple is an output. We do not charge a lower price per IPU for modernization deals than we do for anything else. Our list prices are the same, our realized discounts are the same. So it's -- we're not lowering price to incentivize people. We don't give the amount of credits that we give to modernization deals have been stable over time. We're not increasing the amount of credits we give. So we're not lowering price. We're not increasing that of credits. So why is the uplift multiple going down? It's because we are now allowing our salesforce to go after those on-prem customers and on-prem workloads that will naturally need fewer IPUs to replicate it in the cloud than the deals we incented them to target in the early innings of our modernization journey. To give you an example, a customer that's had PowerCenter on maintenance for a long time. One customer may be using the exact PowerCenter was sold on a number of cores. They may be using their cores really, really efficiently. And so -- and they're paying $100. And there's another customer who's using the exact -- is paying for the exact number of cores, the exact same price, but they're using them inefficiently. That guy who's using it efficiently is going to take more IPUs and therefore, is going to generate a higher uplift multiple to migrate that person. That inefficient user paying us the exact same amount, has exact same cores entitlement is going to take less IPUs to move that, and there's -- we can't do anything about that. We don't want to oversell the IPUs. So we're letting -- because we now know after having experience with the modernization, how positive those deals are, they drag expansion sales at the time, in addition to the IPUs, they need to replicate the workload, they buy more IPUs than they need because they want to add data quality. They want to add data catalog. They want to add data marketplace to it. The utilization rate of those IPUs in term is very good and the renewal rate is better than average. So we become comfortable that go after those low uplift multiple deals, those low IPU deals because the lifetime value is so good. And I'll tell you that the uplift multiple varies a lot. I have to approve any deal that has an uplift multiple below a certain threshold. And those deals that come through, we're basically selling the IPUs at list, which in enterprise software is very rare for anybody because they need so few IPUs that -- that's -- we can't charge them anymore. But then deals come across my desk that are 5x uplift that I don't have to approve those, obviously, but they see them. So it's a really wide standard deviation in that 1.5 to 1.7 is an average, but it's not because we're lowering prices, not because we're lowering -- we're changing incentives. We're just going after a different set of the base.

Okay. Let me see if there are questions in the audience? There's one here.

My name is [ Katherine ] from Durable. I had a really big quick picture question because I'm super new to the story. But if you -- modernization has accelerated. Part of it is people want to take advantage of generative AI and AI generally. So you're modernizing your data state. I guess, like on the other side of this, let's say, 5 years from now, where -- like what is Informatica's place in this world? Like I feel like people are going to stick more workloads and possibly data in Databricks, in Azure. And then it's like there's less stuff to move around? Or is that not the case? So I'm just trying to figure that out. I would love your view given your background.

Yes. Thanks for the question. It is a common line of thought out there for sure. But here's where it breaks down. You use the word, And. People are going to put their data in Snowflake and they're going to put it in Azure, they're going to put it in BigQuery. That's where we come in. Unless you are putting your workload entirely in one data warehouse, you need a data management capability that will allow you to manage your data across the multitude of places where it sits. So if one of those vendors, you mentioned, ends up with 75% market share of the world's data, that's not good for us. But in fact, every third-party survey that you can find because we've looked for all of them, tell us that enterprise data producers, big companies, the Fortune 2000 that is our core market, they expect their usage of data -- cloud data targets, data warehouses, data lake, data lake houses. And by the way, the sources of data, which continue to multiply and get more complex, they expect that to get more complex, more multicloud, more multi-warehouse, not less. That's where we win because we're an independent third-party vendor with the best products sitting on the only platform that can do all your data management tasks from a single pane in glass, unifying everything globally, tracking the metadata catalog and controlling governing, et cetera. So those -- the thesis that the emergence of Snowflake, Databricks, BigQuery would be true if you feel that unlike any other technology trends and software at least that have happened to date that, that leads to this, a historic consolidation of all your data in one place, you have to believe that for that to be the negative for Informatica. If you believe that, that proliferation of endpoint technologies will lead to people using them and picking and choosing which one they use for what workload, that's very much a positive for us.

All right. I think we are out of time. Thank you so much, Mike.

Thanks, Pinjalim. Thanks, everybody.