Good afternoon, or I should say it's almost good evening, with the last session of the day. Thanks for bearing with us. Tyler Radke, Co-Head of US Software here at Citi, and we have the CEO of Teradata, Steve McMillan, joining us, President and CEO. Steve, thanks for making the time. I know we gotta get you to the airport, you know, right towards the end of this session, so would love to just dive in. Give us just your. I think it's been now about four years since you joined Teradata. Give us sort of your vision for the company. What are some of the big focus areas that you've done over the last four years, and what do the next four years look like at Teradata?
Thanks, Tyler. It's great to be here. It's always a pleasure. Yeah, four years at Teradata. Sometimes it feels like forty, and sometimes it feels like four months, so it's been a great journey. You know, over that four years, completely transformed the entire organization, you know, from our sales and go-to-market motions, from the overall strategy of the company, to developing customer success, completely pivoting our research and development towards a cloud focus away from on-prem, looking at how we build and leverage all of the great things that Teradata has built over its forty-year history. You know, you could say we invented enterprise data warehousing, but I'd like to think that we're working on what does the next set of data platforms look like inside our customers. And it's great to see that traction.
So if I just reflect on those four years and what's happened, we built a $500 million cloud business from pretty much zero when I joined. We returned the company to free cash flow positive, very healthy free cash flow generation. We're seeing continuously increasing operating margins, which is great to see. We're living up to that promise that we were gonna be a profitable growth company, and we're continuously making sure we're taking the right actions to make sure our EPS continues to increase on a year-on-year basis. So at a massive transformation, you know, we set out some really core principles when I joined the company, and that is we're gonna be a technology company at our core. We developed, at the time, a very large services organization. Over time, we'd we've been decreasing that.
We're gonna be a software company. You know, we had a great heritage from an on-prem appliance hardware, but at our core, really, if you look at where all of our secret sauce is, it's in the software that we've developed to run these enormous data warehouses for some of the biggest companies in the world. And the best software companies are platform companies. We made the decision that we'd be very focused in terms of what we're gonna do as a company. We'd focus our research and development, our investment envelope, into areas where we've got differentiation, and then we built it in two pillars. We'd be cloud first and partner first. So we'd leverage a partner ecosystem that would enable us to start a new logo motion.
We'd leverage a partner ecosystem, which would enable us to, you know, win the hearts and minds of some of the biggest organizations in the world. It's been a great journey over the last four years. It's a pleasure for me to interact with some of the largest companies in the world and deliver their data needs and continuously changing environments. So thinking about it from either cloud environments or hybrid environments, it's been great over the last four years. I think if we look to the next four years, four years seems really difficult to predict in IT these days. You know, you look back to November 2022, when the whole ChatGPT thing started and the whole focus around GenAI, and I'm sure we'll talk a little bit about that as that's coming up.
But from a Teradata perspective, you know, the foundation that we've built in terms of having an open and connected data platform that can work at a tremendous scale. We're gonna see that continue in terms of building out that hybrid, multi-cloud data platform that can serve all of the data needs of organizations around the world, especially in this new AI world that we're living in.
Yeah. Yeah, great. And I guess as we think about the AI opportunity for Teradata, how would you sort of articulate where Teradata fits in? Obviously, you have an industry-leading data warehouse. How does a data warehouse sort of tie into the world of LLMs and GenAI?
Yeah, it's really, we're seeing some really super interesting use cases being deployed inside our customers. You know, when we think about GenAI, there's some keywords that come out. When I speak to customers, it's trusted. Can we trust the output? And to trust the output of these GenAI solutions, you need trusted data, and we are in often cases the custodians of the organization's most trusted data inside that enterprise data warehouse. We see the word ethical. You know, how can we make sure that we eliminate bias from these models as they're deployed? And the ModelOps capabilities that we've got inside our platform help our customers eliminate that bias every day. And then sustainable, that's super interesting, the level of power, and what this power translates to.
It translates into cost of running these models and the cost of operating these models, so it's been really interesting just on that final point. We're actually running a proof of concept with two really two of the largest banks in the US, where we're actually running a language model in the Teradata, on the Teradata hardware platform, leveraging our massively parallel architecture on a CPU basis, and what we're seeing is we're running a Hugging Face language model inside Teradata, and the performance that we're getting on the CPU-based system is actually better than the performance of a benchmark against a GPU-based solution, so we see a great opportunity to not just leverage GenAI in providing data in a trusted environment, but also thinking about how can organizations run different-sized language models in the most effective and efficient way.
I think it's a really great example of the continuous innovation that we're delivering from a Teradata perspective.
Got it. Now, it sounds like pretty interesting use cases. I guess as we think about the impact that all this Gen AI excitement has had on budgets, we've heard from some companies that it's sort of exacerbated this macro uncertainty in the sense that obviously companies are very hyper-focused on Gen AI now, but budgets are still tight, and so maybe other parts of the IT wallet are getting constrained. Can you just elaborate on what you're seeing with your customers? And obviously, you know, I guess the context, your growth rates have slowed quite a bit over the last year, presumably due to some of these budget constraints.
Yeah. So I think we definitely see, from an overall perspective, Gen AI is a tailwind for Teradata. But I tend to agree with Gartner, although I did have a fairly active discussion with one of my customers down in Australia, that Gen AI is kind of entering into that trough of disillusionment, right?
Mm-hmm.
A lot of POCs have been done. POCs have tended to prove them to be expensive. They might not have got the business return or business value than there was initially anticipated. But I do think there are some leaders in various industries and various segments who have really embraced not just Gen AI, but AI and advanced analytics generally to really generate some significant business value and business outcomes. And, you know, when I think back many, many years ago, back in two thousand and ten, Teradata made an acquisition of Aster Technologies, which was full of essentially an analytics company, analytics capabilities. And when we re-engineered our Teradata Vantage solution, we also took the opportunity to build all of these super advanced AI, ML, and advanced analytic capabilities right into our database.
And by building those functions right into our database, we can essentially run advanced analytic workloads at tremendous scale. Because I think Gen AI is suffering the same challenges that most data science projects and programs have. You know, the statistics would say, you know, over 65% of data science projects, and I'll include Gen AI in that, fail. And the reason that they fail isn't because somebody doesn't have a great idea, that they can't do the feature engineering. It's not that they can't develop a model and train that model. It's the fact that they can't effectively move that model into production.
And so what we are seeing with our Teradata technology set and having that benefit of being able to operate at tremendous scale for the world's biggest organizations, we're able to help promote these models and capabilities into production, better than anybody else that we've come across.
Okay. Got it. So turning to the most recent quarter, and I guess Q1 as well, the. You know, this most recent quarter, you did take down targets-
Yep
... targets for the full year. Q1, you know, you saw some of the on-prem erosions. I guess now that we're, you know, past the first half of the year that was challenging. Do you? I guess, first of all, could you sort of frame what you saw for the audience here, if folks weren't as familiar? And then secondly, what sort of gives you the confidence that that was the last cut and you can kind of get back to this ARR growth heading into next year?
Yeah, I think as we looked at what happened in Q1, we did know. We've got a great handle on our customer base and what's going on inside our customer base. We've got a fantastic customer success motion in terms of understanding where our customers are moving to, what they're doing, what they're considering, where their strategic plans are, what competition is active inside our accounts, and dealing with the largest enterprises in the world, we know that there's always a competitive environment inside our customer base.
But what we experienced in Q1 was essentially so a couple of our largest customers who had made a commitment to move off of our technology before we had the roadmap and the cloud commitment and the cloud strategy and capability that we have today so over four years ago five years ago they made the decision to migrate to alternative technologies. And so they invested in a very significant SI program over the course of the last five years to execute that migration. And what we saw was that impacted our Q1 results but also I think what we started to see was just as you pointed out we started to see an impact in terms of what we considered the macro environment so we saw decision cycles starting to increase inside our customers.
We saw some opportunities that we had expected to close in both Q1 and Q2 start to slip out into later quarters during the year. And so we thought it was prudent to learn from what had happened in the first half of the year and look at our entire pipeline of opportunity that we have to execute against. You know, we're a typical enterprise software company. We're very back-end loaded in terms of execution and executing in a Q4.
But we decided to be prudent in terms of looking at that pipeline, especially some of the big deals that we have in our pipeline, and factor that pipeline against some of the most conservative historical conversion rates that we had, but also go through each of the major opportunities one by one to look at the likelihood of it closing. And as we did that, we decided it would be prudent to alter our guidance for the year. And we altered our guidance for the year from a total ARR perspective, because we saw some decisions in terms of some of the expected on-prem expansions that we were expecting to get start to slip out, as well as seeing deal elongation that meant that some opportunities were slipping into twenty twenty-five. And so we've reset the guidance.
I'll use the term we were prudent with the guidance that we've put out there. We are confident in terms of being able to deliver against that, but because we're a recurring revenue business, we see that impacting into twenty twenty-five, with reduced revenues for twenty twenty-five. So we actually had to put together a whole set of guidance from a twenty twenty-five perspective, which shows our recurring ARR returning to growth in twenty twenty-five. But it did mean that, for us to maintain what we thought was an acceptable level of EPS for twenty twenty-five, we had to take some cost actions. So we took some cost actions inside the business.
So we've taken a whole set of actions across the organization to ensure that we've got realistic guidance, guidance that we are confident that we can deliver against, both for the second half of 2024 and into 2025.
Yeah. So on those cost actions that you took, I think it came out to close to 10%, total workforce reduction, which was one of the largest you've done in years. I know in your tenure, you've been very focused on taking costs out of the model. What were sort of the areas that were impacted by that 10% impact, and how do you sort of think about the additional margin opportunity post this RIF?
Yeah. So, we're very prudent in terms of making sure that we took the right actions, especially not upsetting customer relationships that we have on the ground, and making sure that we maintain those relationships. We took the opportunity in our sales organization to restructure under a new Chief Revenue Officer, restructure our sales organization so that we became more focused. So, for example, we now have a global financial services sector, so moving away from the structure of having, you know, an Americas, an EMEA, an APJ team. So we're able to take out layers of management, which made that organization more effective, got the leadership closer to our customers, we believe. So very prudent in terms of the actions we were taking, especially from that perspective.
Very prudent in terms of looking at the investments we were making from a product engineering perspective and prioritizing that against the innovation that we need to deliver that's gonna drive our growth as we move into the future. And we also obviously took actions against our G&A spend, so that you know, I think there's continuously opportunities to improve efficiency and effectiveness, especially in our G&A spend, and so that those are the types of actions that we took. We were very thoughtful about it, very prudent, and it will give us the opportunity as well to reinvest some of those dollars back into the business to look at things like how do we run language models?
How do we have an open approach to bringing the right language models into our environment? We believe that, you know, small language models are gonna be super important, not just the big ChatGPT large language model solutions, but also how do we invest and make sure that we can deliver, you know, perform an open table format on Native Object Store and opening up the Teradata Query Engine to be able to access all of that data that might be in Native Object Store or object stores either on-prem.
I was talking to one of the banks in Canada, and they basically said, "Look, we have 10 times more data in object stores than we do inside our enterprise data warehouse." And so I look at that as an opportunity for Teradata to expand our TAM, to be able to get Teradata Query Engines accessing that data and combining it with the pristine enterprise data warehouse data, so that organizations can do things like better customer segmentation.
Mm-hmm.
They can look at the profitability of an individual customer, and that we can do that and deliver those solutions on a massive scale is something that I think is gonna be super important as we move into the future.
Okay. Okay, yeah, and I guess it's a good segue on to the Iceberg and sort of open table format conversation that's been very topical in the space, you know, Snowflake, Databricks, and you guys talking about that. I guess some investors sort of have a concern out there, like, yes, it is theoretically a good opportunity. You've done a lot of things to open up your query engine that can kind of tap into these data sources. But just as you think about the risk of the loss of storage or maybe that, you know, decreases the stickiness if all of a sudden it's sort of in this open table format, how do you sort of see the puts and takes on that?
... Yeah, I think, at enterprises, we focus on G 10,000. That's where our core differentiation is in terms of, operating at enterprise scale. And we know that customer segment is always gonna have the requirement for a highly performant solution for their enterprise data warehouse needs and their data platform needs. So if you like that gold tier of high performance, where Teradata really plays well. So, based on our massively parallel compute architecture, which we're now using to run those language models, based on our workload management, based on our query optimization, we know that the biggest organizations in the world will have a requirement to execute some of the most complex workloads and some of the most complex queries at tremendous scale, with a requirement for tremendous performance.
In order to do that, we believe that the Teradata Vantage platform, running in the cloud on elastic block storage, is exactly the kind of solution for those types of workloads. We think that the battle of the query engine as we move forward will essentially enable Teradata to operate at tremendous scale and effectiveness and efficiency for some of the world's most complex workloads. It'll be very difficult, as we've seen, from lots of organizations, for our customers to actually leverage open table format and Native Object Store at that level of performance and execution.
Mm-hmm.
Over time, that will get better, but over time, the capabilities that we have inside the Teradata engine will also get better. And you see workloads like GenAI workloads have the appetite for tremendous volumes of data tremendously quickly. And so I think there's always gonna be a fantastic place for a solution like Teradata that can deliver that high-performance capability, but also can start to query into open table format and Native Object Store.
Okay, great. Maybe that's a great, you know, jump-off point to talk about the competitive environment.
Mm-hmm.
So we often get asked, sort of, how does Teradata fit in with the world? Obviously, you have Snowflake and Databricks, pretty high growth players in sort of this emerging data platform player. Do you see yourselves sort of directly competing with those? And when you bring up some of these, you know, on-prem erosion events, is that typically who vendors are spending more with? How do you sort of see yourself coexisting and competing with those vendors?
Yeah, I think we segment the competitive environment into three. The kind of the traditional competitors, like an IBM or an Oracle. Then we've got clearly the CSPs, who have a set of native services, you know, Azure with Synapse, AWS with Redshift, Google with BigQuery, and then the cloud-native players, like Snowflake, like Databricks. We have clear differentiation against all three of those segments from a competition perspective. Clearly, our ability to effectively and efficiently run some of these workloads gives us some great advantage over both CSP solutions and the cloud-native solutions from Snowflake and Databricks, where a scale-out architecture is not always the optimal architecture to run super complex workloads, which have a high concurrency requirement.
And so our core technology enables us to deliver the most efficient, the most effective, price-optimized solution for these workloads in both the cloud and from an on-prem perspective. In terms of our work with the CSPs and differentiating from the CSPs, you know, Microsoft is a great partner for us. We had a recent event and we're featured as part of their new fabric, you know, the Microsoft vision for data into the future, one of five ISVs natively integrated into the Microsoft Fabric announcement. And that enabled us to position our new serverless query engine, we call it AI Unlimited, inside the Microsoft environment. It's the only query engine apart from Synapse that's available just now and natively integrated into the data fabric.
I think the reason that Microsoft did that is they see the tremendous power and capability of the Teradata platform and our ability to execute these massive workloads and massive requirements.
Yeah.
So, and then from competing with the likes of IBM, I think there's nothing better that we like to do than take out Netezza from a customer environment and shift that workload into the Teradata ecosystem, either modernizing it completely and putting it in the cloud or encapsulating it from an on-prem perspective.
Okay. Great, great. One of the areas from a go-to-market perspective that's been a focus the last couple of years is your new logo,
Mm-hmm
... growth engine, and I think you've talked about some you know improvements in terms of the volume of new logos, but it's still relatively small. Can you speak to sort of what you're seeing on the new logo front and any sizable deals that you're noticing in the second half?
Yeah, I think what we've actually seen is a bit of an uptake in terms of sort of the value of some of the new logo deals. We're also seeing, interestingly, an uptake in on-prem new logos, and larger on-prem new logos. And I think that's a desire for companies, especially in markets like the Middle East-
Mm-hmm
... where cloud penetration isn't as high, or in terms of regulated environments, where you know, an on-prem solution can actually play pretty well. And then in terms of winning new logos from a cloud perspective, they tend to be smaller. But they tend to be use case-based. One of the great things that Teradata brings to the table is the fact that over the last forty years, we've developed some fantastic industry data models, and we've built some great use cases on top of those industry data models. So banking is an example. Fraud and regulatory-
Mm.
-compliance, um-
Right
... is a good example, so going to a smaller regional bank and offering a Teradata fraud and regulatory compliance solution is something that kind of resonates. We take all of the great lessons that we've learned from customers like Citi over the years.
Mm.
Encapsulate it in an offering that we can take to, you know, smaller organizations across the world to, you know, leverage all of the learnings that we've had. Yeah.
Got it. Okay. On the partnership front, you talked about the relationship with the big hyperscalers and-
Yeah
... and all that. I know about a year ago, you hosted a pretty broad partner event. There were ISVs there, as well as system integrators, even some OEM partners. I think Dell is a big one. Some of them, I remember last year, talked about some pretty significant pipeline numbers that they're generating. How would you just sort of characterize the pipeline that you see from those partners, and how fast are they investing or growing their practices?
Yeah, I think, one of the reasons that we've deliberately reduced our consulting and services business inside Teradata is to take out the notion of competition with those partner ecosystems. Clearly, those organizations do a lot of business from an SI perspective, and if a company has a $700 million SI capability, they're gonna see you as a competitor. So as we focus more on our technology and, and that being the primary value creator from a company perspective, it's enabled us to essentially create headspace, room for our partners to execute.
Mm.
And so we see. I'll use Accenture as an example. Accenture have made significant investments in terms of the size of their Teradata practice over time, working successfully with the likes of Accenture and Kyndryl in terms of new projects and new project execution. One of the things that's really exciting is seeing, and I think Rich Petley, our new CRO, brought this focus when he came in, is regional systems integrators.
Mm-hmm.
The smaller regional partners taking those use cases to customers that they know, and really driving business from that perspective, and I think that's how Rich ended up getting more meaningfully successful results in the areas of the business that he was involved in before we promoted him to this new position as Chief Revenue Officer.
Got it. Okay, makes sense. As you think about the rest of this year, there's a lot of exciting things going on in the world, you know, elections, the federal fiscal year-end coming up in September. How are you feeling just about the pipeline? You always have sort of a pretty heavy second half, particularly Q4-
Q4
... weighted pipeline. How are things shaping up, and, you know, how does sort of the election uncertainty play into any of that?
Yeah, I think when we mentioned it on the earnings call, you know, as we looked at the end of Q2 and evaluated the pipe and assessed the pipeline, we saw more eight-figure deals in the pipeline. We saw three times the AR, total ARR value from eight-figure deals compared to the end of second quarter last year. And I think what that is built on is the fact that more and more of our largest customers have got to the point where they see the success that we're having from a cloud perspective.
They see the success that we're having in terms of migrating some of the biggest companies in the world into and onto our cloud platform, without some of the challenges that others have experienced with some of the cloud-native capabilities or the born in the cloud solutions.
Mm.
And so I think that has given them the confidence to actually commit to Teradata in the cloud, and to register interest. I think us recently coming out with VantageCloud Lake, our new architecture on Google, has opened up all of our customers, our enterprise customers that have committed to the Google platform. And so I think as we look at the opportunity, we're very encouraged by that level of opportunity. But as I said, as we reset the guidance, we were very prudent in terms of looking through that pipeline, being very prudent in terms of what deals we expected to close, ensuring that, you know, we did that from two directions: one, from an overall pipeline and conversion rate perspective, but also deal by deal.
What deals really have the compelling reasons to act within the certain timelines and within this year particularly? And so I think we're quite bullish on it, but we have factored in that deal elongation-
Mm
... some deals slipping out into twenty twenty-five, and we're confident in the guidance that we put out there.
Okay, great. Well, I know you gotta get to your flight soon, but I'll just sort of leave it with you. If there's any closing remarks you want to make or anything that you wanted to hit on that we didn't address.
Yeah, I think, the data analytics, the AI marketplace is a fantastic opportunity space to execute in. The transition and transformation that we've executed from a company perspective, you know, going from zero to, you know, over $500 million of cloud ARR in four years, and continuously delivering innovation whilst having a real commitment to return of value to shareholders. Yet, as our free cash flows develop, we've committed more, at least 75% of our free cash flow, in terms of, return to our shareholders, and we see that continuing into the future.
Great. Steve, thanks for joining us. Thanks, everyone, for coming to the last session of the day, and we'll see you tomorrow.
Thanks. Thanks, Tyler.
Thank you.
Thank you very much.