For joining us this afternoon. I'm Skee Venders, part of the software research team here at Citi. With us for our next session, we have the team from OpenText. We have Tom Jenkins. I think you're Chairman of the Board. I have that right?
Executive Chair and Chief Strategy Officer.
All right.
Used to be a Chair. Now I got all those other titles.
Yeah, now you get the extra stuff with it.
I actually used to be that about 10 years ago, and before that, I was the CEO.
Okay, glad to have you here and welcome back to that role. Maybe just to start, I think it has been a bit of a transition time for the company, for OpenText. Maybe we can talk a little bit about what's led to some of the leadership changes and now how you're thinking about that moving forward and what you're looking for in a new leadership team.
Yeah, so uniquely right at this moment, the company's actually in a search for a CEO and a CFO. That's why I came back full-time. How that happened was that at the end of our fiscal year, we had planned for a CEO change, but what caught us by surprise was the CFO change. That's because our former CFO, Chadwick Westlake, his old boss had a heart attack and died a week before the end of the fiscal year. He obviously went back to his old company to take over as CEO. That created the unique moment for us where both executives were gone at the same time. We have an interim CEO, longtime exec at OpenText, James McGourlay. I've been working with him for more than 30 years, so he's a steady hand. Cos Balota , he's a longtime VP of Accounting.
We've got two steady hands while we do the search. The board kicked off two searches. We'll hope to do that in the next few months.
Okay. As you're looking out in the marketplace for both those roles, what is it that you're looking for? What's kind of the right profile or the right personality type to lead OpenText forward in the future?
On the CFO side, we're looking for someone that can obviously handle that scale, handle the complexity of multiple units, that kind of thing. A steady hand there, but not anything exotic, if you will, in terms of M&A or capital allocation. Not any particular unique skills. On the CEO, however, I think you'll see the board, and by the way, I'm not on the search committee, so the search committee will do that. I think the board will lean towards the pendulum going the other way towards more solutions and sales. Our previous CEO, Mark Barrenechea, was like me an architectural engineer, but in between Mark and I was John Shackleton, who was a solutions engineer. If I had to bet, there'll be an emphasis to go back to solutions.
Okay. All right, that makes sense. The pendulum swings the other way.
Yeah, no, it makes sense.
Yeah. Maybe again from a high level, as you think about the opportunity that's in front of OpenText and you know what would you, I guess, pitch an investor on, what would you say is kind of the core opportunity and strategy for the business today moving forward?
Training agentic AI. That's the number one, number two, and number three priority. Part of the reasons why we made the move and the change was to go back to basics. I think investors will see us become a single concept and not a multi-business unit because it became too complex to understand. I think with the opportunity going on with training of agentic AI and the need for content and curated content, that's right down the middle of the lane for us. I think you'll see us pair off some of the business units that don't make sense to that vision, and you'll see us execute from that. It's a great market to be in. It doesn't take a lot of thinking to come to that conclusion.
Sure, maybe we can dig in there a little bit just on the future of the OpenText portfolio, because I think to your point, there's a lot of business lines, a lot of business units that are within the OpenText arena. I guess what do you consider as kind of the core business for OpenText moving forward? I guess, you know, what do you think of the areas where maybe it makes sense to rationalize and look for opportunities to divest?
Yeah. First of all, the issue of the business units within OpenText got exacerbated by buying all the HP software catalog when we bought Micro Focus. We were already creeping into multiple business units and that accelerated that. At the time, when all this is going on four or five years ago, agentic AI, ChatGPT, it wasn't a thing yet. The business was expanding into multiple places within enterprise software. Clearly now everything is about AI. It's about the productivity gains of agentic AI, etc. In retrospect, what the company was doing five, seven years ago made a lot of sense for that environment. Going forward, training bots for very specific tasks within enterprise, that's consuming the world, not just enterprise software, but the whole world. It was easy to come to that conclusion.
If you look at our business units, effectively when you're doing something like training bots, you're talking to the CIO at General Motors or Coca-Cola or whatever. At the same time, it's probably not a good idea to also be selling software to mom and dad at Best Buy, right? That's what some of the consumer products do. Remember, the Hewlett-Packard portfolio was the portfolio of a massively scaled organization that did everything from developer ops to consumer security and things like that. It's actually not that difficult to shape the portfolio to go back to basics because management has estimated that that's about 15%, 20% of the total revenue. It's not like a huge amount, but it's enough to distract management, distract investors, and to cause overall growth rates to go down.
By pairing that part, I think it gives the company an opportunity to reduce debt, gives the company an opportunity to focus on a single story. Besides, it's obvious to anybody, if you have the opportunity where you have one of the largest archives of content in the world and the world needs that content to train their chatbots, it's pretty obvious what you should do.
Right. That makes sense. As you think about this next stage for OpenText and the parent of the portfolio, how long do you expect that to take to play out? What do you do with those proceeds after the fact?
Yeah. With the announcement of all the changes, which is three years ago, we also announced that there would be a committee of the board dedicated to doing this because we didn't want management doing it. We wanted the board to do it because, you know, sometimes when you look at different business units, you fall in love with them a little bit and you think you can fix them all. You probably can, but that's not the point. The point is to be focused on one. The board took over that responsibility to just do the capital asset allocation. I think you'll see them move very quickly. I'm not on that committee either. I think you'll see them go at a pace of one a quarter. They shouldn't go faster than that. That would be unwise. I think that's what everyone can expect.
I think they'll probably do two or three business units over the next year.
Okay. As we think about that, is there an approximation for how many units will kind of be?
No, I don't know that they've identified that yet. I think, one, looking at our portfolio though, just as we've said, it's pretty obvious. There's probably about 15, it's probably somewhere between $750 million and $1 billion out of the $5 billion and change that we do today. I haven't seen the detail yet, but it'll probably be something like that. One of the things we are doing, though, so that investors can follow along with us, we're going to provide more detail on all the segments and their growth rates and what have you. Everybody can say, yeah, that makes sense, that doesn't fit with the other pieces. We're going to try and get that out in the coming weeks.
Okay. I'm sure that'd be very, very welcome by the investor community. That's helpful for sure. As you go through this process, maybe this then kind of dovetails into the other part of capital allocation. What do you do with the proceeds? What do you do with the kind of go-forward capital allocation strategy?
I think there's two parts to that. The one-time use of the capital, as we've done in the past when we did sell another business unit a year and a half ago, we'll pay down debt because we have two tranches of debt. One layer of our debt, our original debt, is long-term paper, 4% fixed. We're quite happy with that cost of capital. There's another layer, which is around $2 billion. We have about $6.5 billion in total debt. There's about $2.5 billion that we'd like to reduce from the proceeds of the sales because that's higher variable. If we do that, investors will be happy because they'll see cash conversion rates come back up to what they historically were. We got caught just like everyone else when the dot plot went 200 basis points higher than all of us were expecting.
I think you'll see that at the sort of asset trading level one time. On an ongoing basis, investors can expect us to keep our historical EBITDA percentages somewhere around 35%. I think they'll see us allocate that capital into three places like we have been doing. We'll keep our dividend and grow our dividend. We will keep buying back our stock. At the current course and speed, we're probably close to buying back 10% of our stock in the past year. We'll just keep doing it. There's a lot of cash flow, so we'll keep doing that. The third thing, investors will see us do tuck-in acquisitions, small. We won't do large ones. There isn't really any more compelling products or install base out there. You'll see us speed delivery, just as we were talking about in solutions. I think investors can expect that.
There's about $1 billion in free cash flow, roughly. I think we'll see a division between those three things. In other words, expect us to continue what we've been doing.
Sure. I guess does it maybe change the mix of how we should think about, you know, those buckets between dividends, buybacks, and the reinvestment back into the business? I guess primarily when I dig into the last part about investing back into the business and what that looks like moving forward, is that a continued area?
We've invested a lot back into the business because of AI. The production of the OpenText™ Aviator AI product line, to be able to see what happened, is that we've had content in enterprise for, you know, since the beginning, 40, 50 years ago. We didn't have the specific connectors to large language models because they didn't exist. We didn't know how they were going to be trained because that entire technology had not existed. We previously have had predictive analytics, decision support, and APIs, and what have you. We spent the better part of the last two years while we were integrating Micro Focus, also building out the entire product line to be able to interface. You don't interface with one LLM. You have to interface with dozens of them because in the enterprise, people are going to pick different large language models to build inside the firewall.
That took a lot of work. That work's done.
Okay.
That is like an entire refresh product cycle that we've just gone through. That's why the emphasis on selling now. Let's go sell what we built over the last two years.
Okay. Maybe I'll ask a little bit on, I guess, to that point. Now that you've made this investment, what does that mean for kind of go-forward R&D? On the other side, since it's now time to go sell, does that mean the mix of sales and marketing should tick up from here? Is it kind of reallocating buckets? Just how do you think about that?
Yeah, absolutely. It just makes sense. We're through it. In many ways, enterprise software companies go through these cycles all the time. The pendulum swings, and you'll see us spend more in sales and marketing and less in R&D. It's normal. That's a normal cycle.
Sure. Okay. You mentioned M&A before. I do want to ask around kind of, you know, what that will look like. Before going into that, just as we think about Micro Focus, maybe we do a little bit of a postmortem on that acquisition. What were kind of the learnings from it? What is there maybe that you can apply moving forward as you think about the OpenText strategy from here?
I'd say first off, would we do it again? Absolutely.
Okay.
If you were to go on Perplexity or ChatGPT or any LLM out there and say, who has the most data connectors to train agentic AI in the world? OpenText. You ask why. It's because they bought the HP software catalog and archive. That was a massive archive. There were only five players. There was Oracle, SAP, IBM, Microsoft, and Hewlett-Packard. The other four are not available.
Sure.
That was a huge jewel in the enterprise software universe. In retrospect, I would say that all of us, both on the board and the management team, wish that we had gone faster to concentrate on the core. Hindsight's 20/20. We should have worked on some of this stuff faster. From an execution point of view, the EBITDA, the overall company, is at 35%. It was executed well. We just needed to do it faster.
Sure. Okay. Maybe to then dovetail into the other side, just as you think about the go-forward M&A strategy, it seems like there's some tuck-ins that you would kind of think about. Just how, what kind of opportunities are you looking for? Maybe what are the areas where it would make sense to kind of lean into, you know, lean into that a little bit more versus, you know, what are areas maybe you'd be a little bit more kind of cautious about entering?
I think the days of us so-called buying enterprise content archives for $0.10 on the dollar are over.
Okay.
Everyone has figured out the massive value that they are. We just saw recently the sale of Informatica, which is a kissing cousin company to OpenText to Salesforce. That went at, I don't know, 5.5x, 6x revenue and 25x, 30x EBITDA. Our ability to pick up assets in the content and data space are quite limited now. Besides, we're at a cycle now where our emphasis is really picking up go-to-market solutions companies, resellers, especially in the era of nomenclature. When you're dealing with oil and gas or you're dealing with automotive or pharma, as you train the bots, it's actually the nomenclature, the knowledge of the supply chain, the knowledge of the word transmission in automotive has a very different context in pharma. You really need to have teams that can accelerate your ability to convert. I think you'll see that will be the emphasis going forward.
It's really about finding the context that helps you train those models.
Yeah, absolutely.
Okay. Are there certain, I guess, maybe like technology areas or market expertise or customer footprint that you would be kind of targeting for those tuck-ins?
Regulated industries.
Okay.
Absolutely. Because those are the industries that we have built our biggest enterprise content archives with and have the greatest industrial knowledge, greatest customer relationships where we can compete and win.
Yep.
That will be our focus.
Okay. That makes sense. Maybe wrapping a little bit of a bow on some of the M&A or portfolio side of it. I guess once you kind of go through this bigger transition over the next kind of couple of years, what does the new OpenText look like once you're kind of through the portfolio rationalization? How do you kind of think about, you know, is it just kind of a content-focused business or are there kind of other areas where maybe it makes sense to?
Content-focused business because, you know, everyone may be getting tired of talking about AI and agentic AI, but the reality is we're going to be spending the better part of the next decade rewiring the world. That rewiring will require training with content, curating the content. Remember, when we speak about content, a lot of times we as human beings, we think of the content in the public web, let's say TikTok, you know, and Meta and what have you. That's actually only 5% of the world's content. 95% of the world's content is behind the firewall. I wrote a book called Behind the Firewall to explain the fact that in the early days of the internet, Boeing, for example, had an intranet, an internal internet that was many times the size of the public internet.
This will take us the better part of a decade to convert all that content. The content is in three types. The content is human-generated content, machine-generated content, and content between organizations or business networks is what it's called at OpenText. Those three types of content are all critical if you're going to train a bot to remap a supply chain for General Motors. You need all three. You'll find that those three business units within OpenText with a cybersecurity wrapper will be the core of the business.
Okay, that makes sense. We're about halfway through this. If there's any questions in the audience, we want to make sure to get to those. Yep, we have a mic coming.
With the strategic changes, what changes in terms of your financial policy with regard to financial leverage might there be?
That's a great question. I'll repeat the question. Would there be any changes to financial policies, approach, et cetera? Actually, no. The current plan for the fiscal year was approved by the board. We had no issue with that. I think the margins, the growth rates, we hope to shrink to grow. We hope to take some of those business units that are growing slowly simply because they weren't perhaps getting as much attention as they should have gotten. In terms of margin, all these units are generally ± 5%, all about the same kind of margin. It's not going to have any shaping from that point of view, not that we can see. We did sell a mainframe unit several years ago that had like 85% margin. That had a big impact, but we don't see a big impact going forward. It should be about the same.
Was there a comment about leverage in there?
Oh, I'm sorry. On the balance sheet, oh yeah.
Is there anything magical of the three times leverage target, or might that evolve?
I think you'll see us come down below 3. Historically, for 30 years, we always kept our leverage sort of around 2, 2.5. I think that's what you'll see us come back to. I think leverage is obviously a very investor-friendly thing as long as it's in the right ratio. That's what you'll see us do.
Okay, perfect. Thanks for that question. I want to dig into the product side a little bit more here. Just on, you know, I think you talked about we're in a content age. Generative AI is going to be the next, is the decade for the next, or the story for the next decade. Maybe we can talk about OpenText™ Aviator AI. Just where is it at today? How are customers thinking about or adopting that solution set? What is kind of the future of the, you know, the generative AI strategy look like for OpenText?
I would say that this is an evolving area. I don't think anyone can say that this will be the dominant LLM. I think that there are many layers to your question. I think CIOs are debating where do they go from here because we've had a massive move from where we had the so-called IT white tower on-prem inside the firewall to the use of whether it was private cloud or with hyperscalers. I think with AI, it's a bit of a different game. They have to decide what parts of their business they're okay to keep data at a hyperscaler level and what parts of their business do they want to keep data in a private cloud or on-prem. CIOs I've talked to, once you start that discussion, then they start to say, okay, then what does that mean? Am I a hybrid company now?
Because they were all moving to the cloud and to the scalers. It's a real dilemma for them. I don't have a crystal ball to tell you which way they'll go. We, just by history, have both because we began in on-prem, so we can support on-prem. Over the past decade, we've moved to cloud services, private cloud, as well as to a SaaS model on hyperscalers. It's really up to each individual company to decide how they're going to do that. Now with Aviator, we built our product line so that it would satisfy all those three areas. Obviously, that's more complicated to build. It took us longer than, say, some of the pure play content players. It gives our customers maximum flexibility. The other layer, so you could call that a multi-cloud architectural approach.
Sure.
You also have a multimodal architecture approach in the sense that which models will exhibit themselves as being preferred in particular industries, et cetera. Part of the dilemma that we have is that most of the models everyone knows in the audience were trained effectively on public information. There's so much training you can get from Reddit and Wiki and what have you. As you go deep inside the automotive industry or you go deep inside the pharmaceutical industry, will there be models that emerge that are preferred by that industry? We don't know that yet. We're architected to be both multi-cloud and multimodal because we can't predict it. We'll leave our customers to decide because, you know, one day they may want Anthropic or maybe they want Llama.
Yeah.
We don't want to decide that. We've certainly made strategic partnerships with various LLMs just to give our customers a sort of a pre-designed choice. We have to stay in an open architecture. We've always been that. We have been the Switzerland, if you will, of content for more than 30 years. I think our clients will expect us to keep doing that. That's what made Aviator a two-year development process because it had to be multi-cloud and it had to be multimodal. That took a lot of thinking. The team did a great job on it.
Okay, that's great to hear. As you think about what the future of the portfolio looks like on the generative AI side, what does the roadmap look like? How do you think about the use cases that it makes sense for OpenText to address and own for customers?
Okay, that takes you to another level.
Sure.
Architecturally, the big debate is a 30-year-old uses an LLM as an operating system. A 50-year-old uses it as basically a search engine. We have different user cases, and the dilemma is we believe we have to satisfy both. You can't tell the 50-year-olds, you know, we're expecting you to have a conversational experience with your LLM app, et cetera. You can't tell the 30-year-olds, well, we're constraining you that you can't use an LLM as an operating system. There is a real diversity in the install base. There are cases, I don't want to call out specific customers, but there's cases where we have to satisfy both uses. Even though these applications could in fact be GUI-free, because they really are, we actually are still building GUIs to interface because that is what a large part of the population expects.
I think what we'll find is as we get better, let's call it agentic coaching, if you will, I think we'll find that we will obviously go to a GUI-less app, but we're not going to get there anytime soon. Human beings are human beings and they will take time.
Yep.
That's the complexity. When you talk about, as we go down into the classic enterprise applications and how those interfaces occur, that's going to be a complexity. The second complexity that we see, I wrote a book on this called The Anticipant. The premise of the book, I wrote it with Mark Barrenechea, our former CEO, and General David Fraser, who commanded troops in Afghanistan. The trick here in building apps is the users have to anticipate because you can't react. Think of Flash Crash all those years ago. The humans didn't know that the market was going down for about 15 - 20 minutes. That's a lifetime for nanosecond trading.
Right.
The dilemma that we have is as we build these apps, the humans have to manage the bots, but the bots are so much faster. There is a degree in these applications that you have to start to anticipate. We're seeing that today in cybersecurity. We're averaging, I think if I got this right, we're 200,000 breach attempts per second. Think about that.
200,000 breach attempts per second. You have to architect your bots so that they're anticipating the other bots, and you cannot participate at that nanosecond level. You can only reach yourself back. I'm sure you're sorry you asked that question, but that's the kind of stuff we have to think through as we go through this because we have a wide range of use cases.
Yeah, no, I mean it's interesting times, right?
Yeah.
How quickly the tech is evolving here. Maybe we can connect it back a little bit to the financial side of the equation. I think Aviators have been out for two and a half, three years, something at this point. What contribution is that having today to revenue or to bookings? Maybe how do you kind of think about what that means for the go-forward financial model for OpenText?
Yeah, it's a great question because we debated how this was going to, and now Aviator itself, although it was released in its first version about a year and a half ago, it was really last year's Aviator that you could really deploy. This year's version will really, really, I think, please our customers. Here's the thing, just like so many other disruptive operating system level technologies, they themselves are not the drivers of revenue. I remember in my era, we came up with search engines and we thought, oh, search engine, this is great. They were free within 24 months. I think that's part of the dilemma that vendors have right now, that they've invested billions into this. Yet customers perceive the utility as table stakes of what they're doing within the application.
In fact, the way we look at Aviator revenue is the revenue that it's enabling in the content servers. The content archives, the records management, all of those utilities that come with increased use of content, that's actually where we're making our money. When you track us, management's been giving you all the content cloud revenue bookings growth. That's where the money is.
Okay.
That's where they've had double-digit growth. That's how you can tell. It's actually, Aviator is an enabler. It's actually the core products that are growing. Without Aviator, you cannot access them to curate the data to then train agentic AI.
Sure, it's a second order effect.
It's a second order effect.
Okay, that makes sense. I have a couple of minutes left here. I just want to see if there's any last remaining questions in the room there.
It's also obvious.
Yeah, right. I think one of the questions we've gotten from investors is just, you know, think about the past few years. What can OpenText do to just kind of improve execution and just kind of get back on kind of a, you know, steady foot and, you know, improve that moving forward?
The execution operationally has actually been pretty good.
Okay.
Here we are, two years later, having done an acquisition which was of the equal size. You took a $1 billion multinational and added a second $1 billion multinational. Two years later, you're clipping along about flat growth and 35% EBITDA. From an operations execution point of view, we're pretty happy, actually. Now it's to go into that growth cycle. I think the big message for investors is I think all of us wish we had gone faster, but it was a pretty complex integration. What investors have to look for now is do we pare down to get to that core fast in this next year and grow from there? I think that's really the watchword for us: can we rationalize that portfolio and grow from that core? When you see the segment analysis, you'll see the core is growing. We want to make that really easy to track.
Okay, that makes sense. In the last minute here, let's think out, you know, five years. We're kind of through all the, you know, kind of what the core OpenText is at that point. What does the business look like? Maybe what are we kind of, you know, talking about what OpenText is moving forward as we get to that time period?
I hope for investors that it becomes a boring name.
Yeah.
You know, that it's mid-single digits in its growth. It does tuck-unders for another mid-single digits. It keeps to its operating discipline of 35%-40%. One of the great things for investors is can OpenText, like other tech companies, do the AI-first strategy, move away from rule of 40%, and maybe start to see rule of 45%, rule of 50%? That remains to be seen whether the industry can do that because we're a price-competitive industry. We'll see. I would say that's what investors should look for. Back to basics for OpenText. We're not doing any more big transformative M&A. We'll do tuck-under. Expect more of the same what we did for 30 years.
All right. That makes sense. I think we can leave it there. We're at time. Tom, I want to thank you so much for being here today. I want to thank everybody in the room for joining us as well.
Okay. Thanks.
Awesome.