Teradata Corporation (TDC) Earnings Call Transcript & Summary

Okay. Cool. Why don't we get started, guys. Welcome to day 2 afternoon of day 2 of the Flagship TMT Conference. My name is Erik Woodring. I lead Morgan Stanley's hardware coverage here. I am delighted to be joined by John Ederer, CFO of Teradata; Sumeet Arora, Teradata's Chief Product Officer. Both of you guys, thank you and welcome to the conference.

Thanks for having us.

Before we begin, let me point everyone to the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. So with that out of the way, this is the first time that I've been able to host both of you at the TMT conference both relatively new to the role that you're sitting in right now. I'd love just to start the conversation, just quick background, maybe 1 or 2 priorities for each of you as you think about the changes you want to kind of enact in your relative seats, and then we'll go from there.

Sure. Well, I'll kick things off, even though Sumeet predated me by about 30 days. But yes, I joined last May and very excited to be on board. I've been in the software industry for 20-plus years. And before that, I actually started in your chair, I was a research analyst for about a decade before making the jump over. And so spent a lot of time in enterprise software, most recently with a company called Model N that was publicly traded and then joined up with Teradata last spring. As I came onboard, we, as a company, had obviously gone through, I would call it a challenging 2024. And so a big priority for us in '25 was really just kind of getting the business back on track. And there were some things that we felt were really critical. One was getting total ARR back to positive territory. The second was continuing to improve on our operating margin. Third was really starting to demonstrate durable free cash flow generation. And I think fourth and perhaps maybe the most important was just demonstrating some consistency on a quarterly basis as we went through the year. So I think we were able to do all of that, and that was probably goal #1 coming out of the gates.

Yes. I'm Sumeet Arora, and I'm the Chief Product Officer at Teradata, joined roughly around the same time as John. I have 30-plus years in tech industry have spanned domains. I did Cisco for 20-plus years and built some of the largest routing systems that operate the Internet today. And then I have spent more than spent 5 years at a much, much smaller company, focused on Agentic analytics, natural language analytics way before ChatGPT became popular. And in the last 10 months, I've been at Teradata. My priority is, 1 or 2 priorities. I think first is to make sure we continue to earn the trust of our customers through our product. That's most foremost Second is really, really focusing on innovation, both foundationally as well as, in some ways, almost like a startup, operate like a start-up in terms of the velocity, in terms of discovering product market fit for some of the newer things that are happening. So it's kind of almost too faceted there, and those are my priorities.

Okay. Great. Now that's a great way to start the conversation. And Sumeet, not everybody gets exposure to you. And so I want to take advantage of this and start with you. And maybe, again, started at a high level, just the demand landscape as we look at Teradata Solutions, especially how they could be changing with generative AI risk around -- risk or opportunities around Agentic AI. What are the solutions that you see right now kind of specifically gaining traction with customers at Teradata?

Yes. I think the landscape I'll just say that the landscape above data platforms like Teradata is shifting. And from traditional applications to reporting, which are still there, but there is absolutely a motion towards Agentic AI, experimentation with that, actual stuff. And I can see that just based on the demand for MCP servers and certain types of queries that show up, right? And I can see that shifting right now. So that's kind of part 1 of the story. The part 2 of the story is the demand for context. Everybody is talking about that. But what is context? Context is the ability to give a world view to agents so that they can be more enterprise-grade explainable, high accuracy. And Teradata platforms like us have the knowledge of the data, the metadata, also the knowledge of the industries that we have served for more than 4 decades, and our ability to combine these to deliver the right context for the use case for the agent so that agents are useful, actually useful, deliver real ROI is the second piece that I'm seeing in my conversations. And the third piece is people want AI to be co-located with the gravity of data. There is no AI without knowledge, without enterprise knowledge, enterprise data. And we are known to be that trusted repository. And people want to locate AI where that data is instead of moving the data to where AI is. And that is an amazing opportunity. I see that in all my conversations. And I'll just say that underlying all of these 3 that I outlined is the demand for moving fast with innovation, but minimizing risk. Moving faster with innovation, but like having cost efficiency and cost economics, moving fast with innovation, but really having the governance and the trust. And I think Teradata is really well positioned to deliver on those.

And if I could just tack on to that quickly. If you look back at our business over the last 3 to 4, even 5 years, we were very focused on building the cloud side of our business, and I think we demonstrated a very strong ability to do that. But the emphasis has shifted and whereas even just a few years ago, it was still very much about migrations and getting to the cloud just for the sake of getting to the cloud now that emphasis has really shifted to some of the things that Sumeet was just describing from an AI standpoint, and there's an eagerness to get going on those investments and using the data where it sits today, whether that's in the cloud or on-premise.

Okay. No, good. And we'll get into that kind of Teradata 3.0, so to speak. Before we -- I want to touch on competitive landscape. Before we get there, I'd be remiss if I didn't kind of let you share your thoughts on AI disruption, right? Obviously, kind of a key topic here, a key topic in the market. Teradata should play an important role in enabling in enabling AI. There's broad risk or concern that new LLM tools can kind of disrupt existing software platform. So just maybe lay out how you think about the risk of AI disruption or the opportunity with AI disruption just as it relates to Teradata.

Absolutely. From a Teradata vantage point, if you look at the new enterprise tech stack that is emerging, there is essentially a knowledge fabric, which is the data, the processes, the documents, the important videos and maybe what's in the heads of people, that's the knowledge fabric. That's the true effect of an enterprise. Then there is a layer that essentially uses that to deliver governed context, trusted context. Then there is the agent flare, which is the layer that is kind of approaching or trying to replace the SaaS layer, right? And then there is the interface with the stack itself, which is going outcome centric. So if you look at that stack and then you look at Teradata, we are absolutely well positioned to play the role of the knowledge fabric and being part of that, both with our structured data capabilities and our absolutely first-class vector capabilities because we can understand unstructured really well. So we have a big part of the knowledge fabric. And I think as long as I and my team and Teradata, we make sure that this is usable by agents, which is the layer above, trusted context, which we are absolutely invested in at Teradata delivering that to agents. And people can build agents in Teradata now. They can deploy agents close to literally in the Teradata environment. They can govern them or if needed, they can build agents elsewhere, but they can come to us through the MCP servers. So we offer both because we want that Agentic layer to have the flexibility to run on-prem, like John was saying, run in the cloud, run in the cloud and safely access on-prem data. So we are really enabling the Agentic layer as well, both in terms of supporting other parties, but also first-party agents inside of Teradata. So we see us playing a huge role in at least 3 layers of the stack all the way delivering on trust and governance as the primary property.

Okay. And let's touch on kind of the competitive landscape in Teradata's differentiation. 2024 was a year of higher-than-average erosion. You've been able to bring that under control. John, you talked about kind of executing in 2025. I know your focus is on hybrid workloads also continue to move to the cloud. So I'm just kind of setting the stage for -- if we think about these large-scale workloads where Teradata excels in price and performance, is that the impenetrable competitive moat that you believe Teradata has or across kind of product pricing performance, et cetera? Where is that impenetrable moat?

Maybe I'll take a shot at it. I think the first piece is our ability to give choice to our customers. They have the choice to leverage -- leverage the same technology stack on-prem or in the cloud, which is increasingly important -- increasingly important because the deployment type dictates economics and security. The second piece is the choice we provide in terms of storage options. And the reason that's important is Teradata can do both block store and object store well. The advantage of the block store tightly coupled architecture is low latency and high concurrency, which is very amenable to operational workloads increasingly being delivered through agents. The same engine is built for the always-on economy. Remember, what my customers tell me is Teradata is the best when the workload is always on or when the workloads are mixed because we have this amazing workload management technique. Guess what? The cloud story was built with human usage patterns in mind. We go to bed at night, we take time off at weekends -- things shrink in the cloud, and that delivers economics. But now we are entering an agenda -- a time when agents don't sleep, they're rigorous. They're 25x more rigorous in terms of the amount of data they access. So in that era that always on economy, we are the price performance option. So that's kind of one choice we offer. We also offer the elastic compute option, which is amenable for things that come and go, things that are experimental. So they get a choice of 2 types of compute, again, price performance, types of storage, types of deployment environments. So we -- I call it as the best engine, the best context. Remember, we spoke about context earlier in my conversation, our understanding of industries like we understand financial services. We understand telcos. We understand airlines. We can bring that knowledge and map the crazy taxonomy of data that exists underlying to give a world view to agents that agents can be much more enterprise grade. So best context, best engine. We've built -- I think we have already spoken about this, so I can say it, which is we're building this AI studio capability, which is right next to our data, our knowledge and that allows people to really bring their AI outcomes right there next to the data without data having to move. And that's a beautiful option. So these are some of our differentiations that actually resonate with my customers.

John, you were just going to say something.

No. Okay.

Okay. Great. And Sumeet, maybe just add on to that. When we talk about the innovation coming out of Teradata, these agenetic architectures and protocols, that are very rapidly emerging. What's the criticality of the role they play as we think about Teradata innovation over not just like 2026, but as you're looking out and you're thinking about how Teradata needs to remain the most relevant for their customers or the trusted choice. Talk to us about that a little bit.

Yes, that's a great question. Look, one of the things that innovators forget is the value of distribution. Right? It's important to build great technology. But if it's not easily consumable and it's not distributed well, then you don't get -- derive the value and you're not able to invest in it. So if I look at the innovation we are unpacking this year, whether it's a complete transformation of our hardware, complete like a huge investment in AI, a huge enablement of Agentic workloads. It is also what we are trying to do is trying to make it easy to consume, easy to use and really focusing on some building blocks that allow me to play the market to continuously discover product market fit because I think as companies in this era of AI, we have to almost behave like start-ups every day like days because things just change so rapidly, Erik, like Anthropic drop something and like whole world is like reacting to it, right? So in this era, you want to have the building blocks in place so you can play the market based on what the emergent needs are. And I think I would characterize my -- our set of innovations this year as also the building block for that. And I think you're going to see us do a lot of partnerships in big ways this year as well, which is another way to drive distribution and leverage.

Okay. Awesome. Going back to the kind of Teradata 1.0, 2.0, 3.0. I've been privy to see all evolutions of that. So IntelliFlex Box is on-prem, Two, we're going cloud first. Now it's kind of the hybrid approach. This could be for you, John or Sumeet, but just talk to us about why this is the approach that gives you kind of that long-lasting competitive differentiation. What sets you up for the most long-term success?

Do you want to talk about it from a product standpoint?

Absolutely. So look, one of the things I get , which is the most pleasure for -- a part of my job is the opportunity to meet customers and prospects around the globe. And there is increasing value attached to the ability to deliver innovation wherever it's needed. It could be an on-prem appliance. It could be a private data center. It could be a sovereign cloud, a local cloud. It could be the public cloud and the Teradata tech stack ships in all those locations. Not only does it shift to all those locations, we ship the same stack, the same set of innovations. The same AI that you can do in public cloud, you can do it in all those areas, on-prem, appliance, wherever it is, you can do it anywhere with Teradata. And it's not like one lags the other by years. No. Within a few weeks, everything is available everywhere. So that is the velocity at which we are moving. We're making sure that the best is available in all these locations. Customers love that. And not only do we ship these and deploy these in different locations, they can talk to each other. So you have real -- you can mitigate risk, you can put the right economics, you can deliver based on privacy and security requirements. And I think we're really, really well positioned. And that is the conversation that happens around the world. So I feel very, very strongly about that.

And maybe, John, this can be for you as a jump-off point from that is, so what does that mean when we think about the underlying drivers of growth when it comes to new logos versus limiting erosion versus migrations versus expansions? When we talk to this more sustainable ARR growth, where is that coming from if we kind of dissect it across those different avenues?

Yes. I mean you kind of touched on every element there. But I would say what's different in terms of the go-forward view is probably the relative weighting of those things. And so I alluded to this a few minutes ago, if you were to go back a year or 2, there was a heavy emphasis on migration activity, right? And that was used as one, a point to get customers to the cloud, but two, a catalyst to engage with the customer again and upsell and cross-sell and trying to expand the relationship overall. What we're seeing now is really a move towards the adoption of AI, and that is a primary driver for us. And so that drives the increased workloads and that's what fuels a lot of the expansion activity for Teradata. And so I'd say, going forward, what you would see from us is a much more traditional land-and-expand type of a model, where yes, we will continue to focus on new logos. We'll also look to expand existing customer relationships.

Okay. Perfect. Awesome. Now maybe Sumeet, last question before I turn it over to John and kind of turn this into numbers is just going back to kind of the product set. We've talked about MCP server. You've talked about the kind of NVIDIA partnership that is relatively new. What are you most excited about from a new introduction standpoint, what does this mean for the pace of innovation that's going to come out of Teradata over the next 1 to 3 years? You talked about, I think, it was 150 AI and Agentic engagements in 4Q. Bring that all together for like where is Sumeet getting excited about where I need to push that product to, so to speak.

Yes. I think, look, this is a year where we are bringing to life through execution, raw execution, our vision of an autonomous AI and knowledge platform. We are moving the addressable data for Teradata from structured data to also making sure that we are able to handle all sorts of unstructured information inside the enterprise, whether it's videos, images, documents, whatever it is that you need for AI. We are evolving to the trusted knowledge platform. So that's kind of part 1. The second piece is we're delivering AI outcomes. Everything starts with ROI with measurable outcomes, work backwards from there. We're delivering AI outcomes right where the knowledge gravity is where data gravity is and in different environments like we discussed. And to support that evolution to the autonomous enterprise, we're delivering our own stack to be fully Agentic and autonomous. So that's kind of the vision and strategy. What am I most excited about? I think this is a big year for us. I think if I look out the next few months up until summer and so on. We have a full transformation of our on-prem platform underway. We'll transform on-prem to an autonomous AI and knowledge platform. We'll enable people's AI factories with knowledge and help them deliver outcomes, right? So that's a big deal. In terms of cloud, it's going to be an exciting year for us where we combine the low latency, high concurrency, always on system with a femoral elastic compute that delivers both the ability to innovate fast but also deliver price performance for stuff that needs to be always on and proper and structured. I'm very excited about that prospect, both in the cloud and on-prem. So that's the story. And I'm super excited about our AI studio launch, which is going to bring together -- I know we didn't dwell on it, but I'll take 30 seconds and finish this off, which is the AI studio brings the ability to do entire agent life cycle, build, deploy, govern agents with the MCP server allows the ability to do all sorts of analytics with 150-plus in database analytics functions that we have, allows the ability to run applications that are oriented at specific use cases and outcomes. And last but not least, allows for scalable Python compute to leverage both normal CPU compute as well as accelerated GPU computes in both on-prem and cloud environments. My customers are loving this vision and strategy and the execution, and I hope to bring that to life this year.

Cool, I'll maybe transition more to John now and find out how we're turning that into numbers, so to speak, on the financial statements. So John, as someone that's covered for Teradata for a decade. It's exciting to see stabilization in top line, a modest acceleration in the business, 2% to 4% ARR growth, flat to 2% recurring revenue growth. We've kind of talked through how you expect to get there, so to speak, with a lot -- with expansions kind of primarily the driver. Talk to us about how you see kind of upside and downside risks to that. That could be macro related, that could be company and product related, just up and downside risk.

Yes. I mean, I think that well, look, prior to this week, I would say more focused inwardly on our own execution. Obviously, there's a lot going on in the world today, and that's a very large wildcard for all of us. But I think really, when we look at how we're operating, how we operated through '25 and then how we set ourselves up for '26, it really is about execution for us. And so we did get back to positive territory on total ARR growth last year. That was an important milestone, especially relative to '24 to get us back on track and to start to stabilize the business. We did launch some products last year. We've got a lot more slated for this year. And we did start to do some things from a go-to-market standpoint to hit the ground running in '26. We talked about on our earnings call, the 150 proof of concepts that we did on the AI side. So we have our forward-deployed engineers out there working very closely with customers, developing use cases, creating these POCs. We've added a services team, our AI services team to come in behind that and take those proof of concepts and spin them up into actual projects and get customers launched and going on this initiative. So those are some of the ways that we're starting to drive -- combine the product side with the go-to-market effort to really start that expansion activity in '26. And so I think if I look at sort of the upside and the downside risk, I do think a lot of it comes down to our own execution in '26.

Yes, yes. Okay. Perfect. And we talked about expansion, again, key driver of growth. If I were to maybe just push on that a little bit, cloud net expansion that's a rate that has been falling. I realize that cloud isn't necessarily that maybe the KPI that it was during Teradata 2.0. This is the hybrid story. But to be fair, cloud is still kind of the primary driver of growth, so to speak. When does that rate kind of stabilize? Or just walk us through where you expect to see these expansions happen.

Yes. No, that rate has been shifting as we've moved through the migration period. And especially if you look at the total ARR growth rate, that was obviously impacted for a few years by migration activity. And now we're starting to come down on the other side of the bell curve of that activity. And so I would say we're close to seeing that normalize. And so what I would expect going forward is a little bit more convergence between that net expansion rate and the total ARR cloud growth rate.

Okay. Okay. Helpful. And then I know you don't -- I know you don't guide beyond 2026. I don't expect you to give that. But it would be helpful for us to better understand if you see 2026 as kind of an inflection year that should enable a more consistent return to growth or even how you guys are thinking about the revenue growth trajectory from here? Because presumably, the fact that ARR is growing this year would set you up for more growth in 2027. But like where are you on this journey from stabilization to potential inflection?

Yes. I mean at the risk of oversimplifying it, we're looking at this journey as a bit of a crawl, walk, run scenario where 2025 was the crawl that was kind of getting us back to a stable business, getting us back to the positive side of growth from an ARR standpoint. '26, we're certainly looking to build on that. And so even just from a total ARR standpoint, we're at 1% constant currency growth last year, 2% to 4% guidance this year. So making some incremental improvement and continuing to build the foundation. The other thing what we're doing in '26, and we talked about this on our call is that we're really prioritizing our investments. We're still looking to optimize parts of the business, so we can drop incremental benefit to the bottom line but we're also reinvesting some of that. And in particular, we're investing on the product side. And so a lot of the things that Sumeet described and we're working on now and we expect to launch over the course of this year. will become drivers for '27 and beyond.

Okay. and then sorry.

Yes, yes. And so -- we're certainly looking to start to drive that glide path. I'll stop short of giving you guidance for '27, but obviously, we're looking to continuously improve.

Fair. Okay. And then I want to touch on -- I'll ask you kind of 2 questions about margins, gross and op. Just on the growth side, cloud gross margin is still modestly dilutive to the platform, I believe. How do we think about, again, these expansions and kind of the hybrid approach. When does that stabilize gross margins? And then second to that, these initiatives on the cost side, which is reinvesting in innovation but also finding areas of efficiencies and rationalization. What does that mean for the trajectory of OpEx because then we can all kind of back into how we think about op margins can kind of go from here?

Yes. So kind of -- we'll kind of walk down the P&L there. So first, from a gross margin standpoint, and I think -- you kind of mentioned cloud gross margins, but I think it's really the total recurring gross margin, which includes both our on-premise subscriptions and our cloud business. So today, the on-premise subscriptions have a higher gross margin. than the cloud side and cloud has been growing faster. So that presents a bit of a headwind to the gross margin now. We have been incrementally improving cloud gross margin each year. We don't disclose that externally, but we've been making progress. We're very focused on that side of it and particularly for '26 and beyond I feel like we really need to start making some stair-step improvements there to ultimately drive future operating margin improvement. If you look back over the last several years, we've made great improvement in operating margin over 500 basis points over the last 3 years. But I would say that would be -- that was largely on the backs of streamlining operating lines, right? And we still think there's a path to do that even with reinvesting in R&D in '26. And so our guidance would suggest another point of improvement on the operating margin in '26. But I think to really sustain that and maybe even take that stair steps higher, we really need to focus on the gross margin side of the equation.

Okay. Okay. Perfect. As we want to kind of think about -- let's say it this way execution is key this year, remains key. As we want to hold you guys accountable to say you're executing the way that you intend to execute. In addition to what I imagine is ARR growth, that's kind of one of the most important KPIs. Is it earnings growth? Is it operating margins? Is it free cash flow growth? What are the other KPIs we all should be thinking about that's like, hey, we're going to hold John and the team accountable. These are the metrics we should be tracking?

Yes. No, I think that's fair. I think that, yes. Certainly, from a top line perspective, the total ARR number is the more forward-looking number, right? And so -- if you're thinking about what's going to drive growth in '27, you should be focused on the total ARR growth in '26. Certainly, I think it's valid and important to continue to see how we're doing on the operating margin. Are we continuing to drop incremental benefit to the bottom line. When we talk about profitable growth, that's exactly what we mean. We're looking for ways to invest that will drive future growth, but we also want to do so in a responsible way that we're dropping that incremental profit down. And then from there, I would probably focus on free cash flow. I think that earnings per share is also a good gauge, although there's some different dynamics at play there. I think if you look at the free cash flow number, I view that as a critically important number. And even again, if I go back to last year, we had given guidance that was roughly $260 million to $280 million in terms of free cash flow for the year. And we had done $277 million in the year prior. And as I came onboard, I was very focused on the amount that we had done last year. And was very focused on making sure we got to that level or higher in '25 and I was pleased that we were able to come out a little bit better than that. So I view that as a critical foundational element for us from a stock standpoint, frankly, and from a valuation standpoint. I feel like that gives us a good floor for investors.

And just on free cash flow, kind of the underlying drivers there? Obviously, we hear about profitability or the expectation for profitability improvements, that's clearly one. You've had some restructuring costs in the model the last couple of years. Are those rolling off? Do we expect more of them? Like what are the building blocks to get that -- this year, I think it's 10% free cash flow growth roughly is the target. But what are the building blocks that get to that more kind of sustainable free cash flow growth as you're kind of outlining it.

Yes. It's -- I mean, certainly, the face of the P&L is a big part of it, right? And so continuing to drive that operating profit improvement and looking at even below that, looking at the EBITDA performance that's your front-end driver to it continuing to be focused on the working capital side. And so doing a good job on managing collections and payables and those elements being diligent with our capital spend, our CapEx. We've generally done a good job at that. Now there may be some investments we make this year particularly around the hardware side of the business. But in general, I think we've done a nice job of managing that. And so we'll focus on all of the elements that ultimately drive them.

Recently, you guys kind of settled with SAP, you got a $480 million gross payment that nets out to, I think it's $355 million to $362 million of kind of net cash. What do you do with that?

That's a good question. Fair question. And the honest answer is still to be determined. And I think that we'll look at our entire capital structure. I think it's strong today. We've got a good amount of cash on the balance sheet. We're generating healthy free cash flow. So we're in very good shape there. We do have a term loan around about $450 million that comes due in June of '27. And so we'll take that into consideration. We're already active in the market from a buyback perspective. We announced a refresh of that program, another $500 million at the end of 2025. And so we're actively in the market and have committed another 50% of free cash flow going towards stock buybacks. And so we'll take a look at all of those elements, and we'll talk it through with the board in a few months and just leave it at that.

And maybe as we think about that, those are maybe -- as you've outlined, it sounds like those may be the most likely opportunity.

Yes, I think like we're going to look at everything holistically, right? And so you've got debt, you've got -- what are you doing in the marketplace from a share buyback program. Maybe there's reasons to preserve some cash if we want to look at some potential M&A activity. And so I think all things would be on the table.

Okay. With a minute left, I just wondering if you can give you guys the dance for kind of last comment here. The message that you want to leave with everyone maybe what people might underappreciate or undervalue about the story, each of you want to matter, but...

Well, I would just say that Teradata is changing, while keeping the best of what we have done over the decades, we're changing. We are really, really enjoying this market in terms of innovation and really operating like a start-up in many ways. So that's what I would conclude my parts with. But you have few seconds.

Yes. I would just tack on to that and say that it's a business that's undergone a lot of change over the last several years. And certainly, the transition from on-premise to the cloud was a big driver of the business for several years. And when you're going through those types of transitions, it can get a little muddy. The story is sometimes harder to understand. So I guess I would encourage people to take a look and do the deeper dive because I do think a lot of things are starting to tilt our way. I think customers are starting to tilt our way and how people are deploying Agentic solutions, again, is starting to tilt back in favor of Teradata.

Cool. That's a perfect place to end. Thank you, guys.

Thank you, Eddie.