International Business Machines Corporation (IBM) Earnings Call Transcript & Summary

My name is Wamsi Mohan, I cover IT hardware for BofA. Delighted to welcome IBM here today. We have Rob Thomas, SVP, Software and Chief Commercial Officer. Rob, thank you so much for joining us here today.

Great to be here. Nice to see you.

It's super exciting to have you here because I know that you've been in the takeoff where all the excitement is. And so there is just a whole lot of things that we want to touch on. And unfortunately, we only have 30 minutes. But Look, we've had a few opportunities actually to host you over the last several years here at our tech conference and now your roles kind of expanded. So do you want to just maybe share a couple of thoughts on your expanded role.

Sure. 2 years ago, I dug into our go-to-market. And so I've been in software before that and spent the last 2 years on how we're going to modernize our go-to-market, which I think we've made really good progress on, but there's always more to do. In January, I took on the software role in addition. So I spent roughly 50% of my time on software products, innovation, M&A, what we're building, productivity, that type of thing. And the other 50% is go-to-market, working with clients, hearing what clients are doing, working with partners. So I kind of split time between those 2.

Excellent. So maybe just to dive right at, right, I think there's been a lot of discussion about the macro environment now for some time. And given your work with clients, what are you hearing and seeing with respect to demand and what the outlook looks like here?

I would say it's consistent to what we talked about in April. And the fundamental item is technology is more important than ever before for companies. So even if companies are looking at the environment and trying to figure out what's going to happen, maybe in the past, their answer was reduce the IT budget, get some quick savings. That's not how I see people approaching this. They're thinking of technology as much more core to our competitive advantage at this point. And yes, we're focused on productivity. One of the biggest enablers of productivity is probably going to be technology. So it's really the approach to technology that I see. I was in Europe last week, I was in Asia 2 months ago. That's a pretty consistent discussion as I travel around. We talked about in April how we've seen some softness in consulting in the Americas. That's true. And -- but it doesn't change, clients are thinking about how do I take advantage of technology. Generative AI foundation models, I'm sure we're going to talk about that. That's driven a lot of interest. But the tone is for many companies is productivity. How do we do more, faster. That's what I see.

Okay. That's helpful. So -- maybe just for those who are less familiar with the current scope and strategy of IBM, can you just talk about where the leadership focus is. And as part of that, I know you've made some very exciting announcements I think recently. Maybe you want to highlight a couple of other things from there as well.

We've been pretty focused on our strategy going back to 2020 when Arvind started. It's first day, he said hybrid cloud AI. That's what we're going to focus on as a company. And I think we've made a lot of progress there. In 2020, nobody was really talking about hybrid cloud. I think we were still in the era of everything is going to go to a single public cloud. I do think the markets come in our direction on that, where I don't talk to any client now that's not saying our strategy is hybrid. Now how do we take advantage of that? So a lot of time on hybrid cloud across both technology and consulting. AI is, I'd say, the newer front here. And in addition to those core strategic areas, the business we're focused a lot on go-to-market, how do we bring more technical skills into our go-to-market, changing the dimension of how we work with clients. Like I said, I think we've made good progress on that. Our consulting team is aligned around hybrid cloud services and business transformation. So we have a very focused strategy, and we're executing. We've got to keep up the momentum. You asked about Think. We made some significant announcements at Think, namely watsonx, which is our platform for generative AI and foundation models. We started this work in 2020. So we've been investing for 3 years. And if you think about the arc of Watson, I think the first instantiation, which was really around machine learning, deep learning. We had some successes, we had some failures. It was really hard. You think about those projects, a lot of it was data labeling, data annotation. It was so much manual effort that you had some success, you had some failures. I think what's different this time is generative AI, kind of the core technology around transformers. It gives you a lot greater ability to get value out of your data more quickly. We -- using transformer architectures, we can train on unlabeled data. And we're starting to see customer impact. We have a lot of clients using watsonx now as we're in preview, and we're looking forward to where this goes from here.

And what about watsonx? Would you say is differentiated relative to maybe other LLM offerings out there?

We're not trying to play in the consumer space. So we are not going to have the largest parameter LLM, it's not part of our strategy. Think of 3 dimensions of watsonx. First is watsonx.ai. This is a studio for building foundation models. In addition to that, we have built foundation models in IBM that we make available as part of watsonx.ai. And we're really focusing on more narrow domains, things that we have the data to train models on. So things like cybersecurity, IT automation, digital labor, customer care. These are narrower use cases where we can build and train a model using IBM data. And then we go to a client, and we say, you bring your proprietary data, then that becomes your model. We think these will be very competitive. And also very accurate. And I think the most important distinction between sometimes you get hallucination in consumer AI, there's really no appetite for hallucination in the enterprise. So accuracy is really important. Our approach around use cases will drive accuracy. I think that's important. So first piece is watsonx.ai. Next piece is watsonx.data. This is a data architecture, a lake house for making new data ready for AI. The biggest challenge that we learned through the last decade is proprietary data architectures make it really hard to access data and train models. So we've built on iceberg, which is an open source table format using Presto, which is an open source query engine. So a lot of innovation happening in open source. This is part of the transformation, I think that in part having Red Hat as part of IBM has brought, we're starting with open source here. We're seeing incredible performance metrics, price and performance on watsonx.data. Third piece is watsonx.governance. Think of this as a nutrition label for your AI. How is it performing? Where is the data coming from? Are managing the life cycle of the models? Are you able to understand how decisions are being made? So many in the market, their approach is we have one model, use our model. That's acceptable for some use cases. Our approach is businesses will adopt differently. They're going to think about building models. They're going to think about data, they're going to think about governance. That's why we've done this with watsonx. We also partnered with Hugging Face to bring open source models into watsonx. So we're indifferent where clients start. We think there's a lot of value in the IBM models. But if clients want to work with a Hugging Face open source model, they can do that right inside of watsonx.

And how does that accelerate? Or does it accelerate the development time frame and lower the cost when you've got some of these foundation models?

We've made a lot of the fixed capital investment upfront. It goes back to 2020. We built a large GPU cluster. We've been doing the training. So that initial hurdle for training, we've kind of absorbed that. That's really the value that we provide in part is we've absorbed the big upfront so that clients can get started much faster. Now it's not an immediate thing. You still have to bring in your data, you still have to do the training, the tuning. But they don't have to face the investment curve that others would have to face. So I think that's a big part of the differentiating story of bringing models to solve specific problems. And right now, every CEO I talk to is thinking about what am I going to do? How do I do it? And then quickly gets to, can we trust the decisions that are being made, which goes beyond the model into things like governance. So I sense that we have the right approach in terms of the capabilities that we're working on. Now we just have to engage with clients and pay it off.

So there's a lot of discussion around AI and -- where does AI become a threat versus an opportunity? What's IBM's overall view on it? I know Arvind has kind of spoken about 30% of back office can be replaced with AI. So when you put together sort of what's the puts and takes are, how do you think about this from an IBM framework?

We've been an aggressive adopter internally, which is what you've heard Arvind alluded to. One of the products that runs on watsonx is called Watson Orchestrate, which is really about training skills to automate repetitive tasks. So the example that you alluded to is we're using Watson Orchestrate to automate our HR functions. And we've learned a ton in this process. It's a great way to kind of pressure test products as we bring them to market. And I think it's really important for us to be customer #1 of anything that we're doing in technology because it becomes a good proof point that what we're talking about is possible. So I feel really good about what we're learning through that process and how it's actually transforming IBM. In terms of disruptors, I think you could also look at what's happening in the market. The first big partnership we announced was with SAP, where SAP has adopted watsonx as their AI platform. And I give them a lot of credit. They are a very discerning engineering culture. So trust me, they put us through all the steps of are we sure this will work? Can it do what we need to do? AI can be disruptive to business applications. But they're kind of going on and off, and saying, we're going to integrate watsonx. We're going to change customer experience. We're going to make it easier for our clients to use our products, I give them a lot of credit for having the foresight to do it. I think that's an example of -- if they don't act, then they could be disruptive, but to their credit, they're acting. And so I think every software company is going to face that decision pretty quickly. Are you going to invest to do this on your own? Or are you going to find a partner? Doing nothing is probably not a good option.

Yes. Yes. Well, that makes a lot of sense. Maybe switching gears from -- can you just double-click on the elements of your software business? I mean, starting with Red Hat. What's driving the performance there? How do you view the opportunity going forward?

We're really pleased with Red Hat. I think it's achieved everything that we had in mind and more. And the momentum around hybrid cloud is obvious, as I said before. Every client is now hybrid cloud. The other thesis behind Red Hat related to that was, are we going to see a transition and architecture from J2EE towards containers. And I would say we also see that playing out. We talked about in the recent call Red Hat's now -- Red Hat OpenShift is now $1 billion ARR, growing 40%. I think that's a pretty good testament to the momentum of what's happening in containers. Red Hat is also Ansible, an automation platform, another announcement, I think, was around Watson Code Assistant, which is first use cases around Ansible, doing code completion for Ansible developers. We think that will create a lot of momentum in Ansible. And then we've got Red Hat Enterprise Linux. Another big announcement we made, I guess, maybe 3 months ago was, again, SAP announced that they're moving away from SUSE Linux towards Red Hat Enterprise Linux. That's a significant move in the market for a company that all has been on SUSE. So Red Hat has momentum in all 3 dimensions, I would say.

What do you think precipitated that move?

The value of Red Hat Linux, the core value proposition ultimately is security and upgrades because any client knows that the biggest potential security threat is something that happens at the operating system level. And they're also constantly upgrading applications. So you need a way to keep the operating system current while you do that. SUSE has been through a lot of integrations. That creates uncertainty for people. We think with Linux, we provide great certainty on -- it's going to be the most secure operating system platform. You're going to be able to do upgrades of your applications seamlessly. There's going to be no business disruption. It is now extended even in things like General Motors that's adopting Linux for an onboard computer. And seeing Linux go from the data center to edge, I think, represents another significant market opportunity.

That's really interesting. How are you thinking about other areas of software, if you think about automation, data and AI security? And do these really benefit from a hybrid cloud/AI focus?

I think one of the big decisions we made if you go back to 2020 was we started actively partnering with hyperscalers, namely Microsoft and AWS. First step was building consulting practices around them. That's created a lot of opportunity and tailwinds in our consulting business. Next step was to start to move software properties onto those clouds. And this is really about reacting to client demand. We have clients that are using our products. They want to run them in a hybrid fashion. And to me, it was an easy decision to play into the demand of where clients were. As I think about the business, so -- at its core, it's about hybrid cloud. Next you have watsonx, which will be the platform for AI and data. If you think about the software properties on top of that, digital labor is an emerging space. I talked about Watson Orchestrate, Code Assistant, Watson Assistant. These are products that make it easy to automate tasks. I think there's a lot of upside in that. Every company is asking about that. Next is IT automation. We made an organic development bet with Cloud Pak for Watson AIOps, that's done very well. And then we did acquisitions with things like Instana and Turbonomic. We have, I think, the most complete portfolio for IT automation. There's a lot of demand there. Next is cybersecurity. And we're kind of making a different play. We're focusing a lot on data security. I think there's so much interest in threat and threat is very important. The part of the market that has maybe been ignored, it has been data security. We have a great property with Guardium, which we're now extending for multi-cloud hybrid cloud deployments. So those are kind of the big areas I think about on top of the strategy around AI and data and hybrid cloud.

Okay. That's helpful. There's been a lot of focus on transaction processing. We've seen several years of decline. It seems as though the trajectory is kind of resetting to a higher level. Can you talk about what's driving that?

There's a fundamental shift that happened, and it's hard to pinpoint an exact time frame. But I'll say it's in the 2020 area. When clients were myopically focused on -- we're moving everything to a single public cloud, by definition, mainframe became a little less important to them because they're thinking my mandate is I'm moving everything to public cloud. That means I have to move some stuff off the mainframe. I'm not sure what I'll move. It kind of became, I'd say, contained for a bit. The fundamental shift that happened is the moment that a client says, hybrid cloud is my strategy. Suddenly, the mainframe was critical again, because then it becomes about economics and performance. What's the best place to run this? So suddenly they had the flexibility to do that. I was -- as I mentioned, I was in Europe last week, and I had 2 clients that opened the meeting and say, I know we were saying a lot of stuff a couple of years ago. We weren't sure about the mainframe. I just want to be clear. The mainframe is critical to our strategy. So it's a complete change. And that is -- should not be understated, I'd say, in terms of the shift for how people are thinking about the importance of that. And -- so we've seen workload growing. We've seen MIPS growth. That's positive. We did take some pricing action. And I think that proved that we do have pricing power here, which is what we thought. But we're optimistic on continuing to kind of see this low single digit, mid-single-digit growth on the transaction processing.

And just -- so one thing you just add was, Rob, was during maybe starting in 2020, this realization that hybrid cloud is the right strategy and not everything is going to go to the public cloud. Prior to that or during that time frame, what happened to MIPS and MIPS growth? And as you think about the installed base of MIPS, how has that been changing?

I think -- well, there's a few factors here. One is that I'd say that just overall fundamental trend. Transactions, meaning when there's more transactions volumes, by definition, you get more MIPS growth over time. So there was a period in 2019, 2020, where things did change in the world, right? But then we saw a lot of transaction growth as we got to the second half of 2020, I would say, and continued. So I think, though, the biggest distinction is mainframe is now part of the architecture. We still have clients that want to discuss, how do I modernize the mainframe? How do I get better usage of the capabilities in mainframe? Our consulting colleagues and many of our global system integrator partners will talk about how does the mainframe work in context with your container strategy with your public cloud strategy. So it's not just about MIPS growth, it's not just about price. It's not just about the architecture. I'd say it's a mix of all these factors that has clients thinking about how they better use the platform. I think one great example is Citigroup took their MongoDB instances that were running on a couple of hundred x86 systems and consolidated out of zLinux. I don't think you would have seen that 5 years ago. But again, once mainframe is part of their architecture, it's actually a pretty easy decision. They already are using the floor space for mainframe. It's 30% more energy efficient. If I don't have to keep buying x86 machines and I could just run it in a zLinux partition, that's a no-brainer in terms of CapEx, OpEx, you name it. So I think that's a good example of clients [Audio Gap]

How much of those need to interact in a hybrid cloud world with what's sitting on maybe public cloud instances? Like what's the -- like is -- are we thinking about it the right way if you think that Red Hat has kind of OpenShift as a bridge between mainframe versus all the cloud instances out there.

I can't really give a percentage because every client is different. What I'll say is, when you look at the mainframe, there's applications that economics performance, you will never duplicate what you have on the mainframe. So that's good. There's some that containers might be the right answer because you want some more flexibility for how you can deploy, how you can burst, if they're more bursty type applications. So that could be different based on the application. We do a lot of work now with OpenShift on the mainframe as part of zLinux, which is one example of doing that. But I think the important thing is just clients are thinking of this as part of their architecture. And containers is another option alongside that, to your point. And it's just giving -- it's opening up a new dialogue that we think is very [ helpful ].

Okay. That's helpful. Can't let a discussion go over here without talking about M&A. And curious to hear how you and Arvind are they're thinking about M&A as it pertains to the overall landscape and particularly in software?

We've -- I think since Arvind became CEO, we've done over 30 acquisitions, 60% or so software, 40% consulting. So we've been aggressive but also disciplined. And so number one is we know the spaces that we want to do acquisitions in. It's back to the areas that I talked about in software. It will be automation. It will be data AI -- data and AI. It will be cybersecurity. It will be hybrid cloud. Like those are the areas that we'll focus on. Given the, I'd say, the macro environment over the last year where there's been a little bit of a disconnect on valuations, we've had to be patient and opportunistic as things become possible. We've done 5 deals this year in software, which are probably below the radar for what many of you think about. But I think it shows that we have different muscles, different archetypes. We've done smaller deals that accelerate organic innovation that we're doing. We did one in data security just recently. So we've been active, but for the moment, we're looking at what are the ones that make the most sense. We definitely are looking for things that are more sizable than some of the smaller tuck-in ones. And -- but that will be about the economics, finding the right fit to the portfolio. And so we'll keep our eyes open here.

I think Arvind said in the past that you guys could do something very large. I think, but we haven't seen post Red Hat anything that's very large. And so is there an appetite for doing something that's in the zone of $15 billion, $20 billion?

I'm not sure the regulatory environment today is conducive to that because I think what we've seen at that scale, you will be tied up for 2 years. That's a pretty significant opportunity cost in technology. So I don't know that that's attractive because even if you're confident, you can get it done, it's 2 years. That's a long time. There's a lot of other things that we could be doing. So I think for the current environment, that's probably not the most attractive zone. That could change, but that's probably not the top of the list at the moment.

Yes. And in Consulting, would you say that -- I know you've done a whole bunch of M&A in Consulting that all be relatively small. Is there any reason to think that, that would change in any material fashion?

The focus in Consulting early on was around supporting the moves that we've made in ecosystem and our partnerships. So we did some AWS partners. We looked at different software partners, Adobe, Salesforce that kind of thing. So we did things in that space. The deal we did last year for Okta was sizable in the U.S. federal space. And that was kind of a longer-term thesis that we have around governments and investments and what's going on there. So that's sizable. So we're open-minded here. But we've been -- Consulting is really largely focused in 2 areas: hybrid cloud transformation, which includes getting people onto public clouds and business transformation, which is largely partnerships with the likes of SAP, Adobe, Salesforce, that type of thing. We're going to kind of stay largely in those areas where we have confidence we can execute.

And just one thing on Consulting, if I could, on AI. I know when you made these announcements on watsonx, you also were doing a significant amount of training and ramp-up of Consulting expertise around this. What -- is this going to be specifically on watsonx? Or is there a broader approach to it as -- from a consulting arm standpoint as you think about leveraging Consulting, which is going to be very much needed to -- for enterprises to really -- enterprises are still trying to figure out ML and how to use it. So now we've got a whole new layer of things that -- and complexity that they probably need help with. So what's IBM's broader approach to it?

A few things. Number one, watsonx is a product strategy. Should be clear on that. I think the first iteration of Watson, there was too much custom services that happened early on. We've learned from that. We want to build a platform, a product. That's the focus on watsonx. I think the monetization of watsonx in the next year or 2 will largely be in Consulting for the reason that you state, because you go to clients, you talk about use cases, they immediately want help to do this. So I do think there's a monetization opportunity around Consulting. We did launch the Center of Excellence for generative AI. Is it specific to watsonx? There's a nuanced answer there. Watsonx is as much open source as it is IBM models. And my view is generative AI is going to play out more in open source than for any one company or one model. I really believe that. So consulting is basically has no limitations on what we can do. If you look at watsonx in terms of the IBM models as well as what's happening in open source. If something else gains a lot of steam, we're not going to constrain Consulting. We were consulting to meet clients with where they are. But I believe the winner for generative AI and foundation models is going to be multi-model, multi-cloud, embracing open source. So I think there's a play here that's really right aligned with what we're doing in watsonx.

Okay. That makes a lot of sense. Rob, unfortunately, we're almost at 30 minutes, which is really not enough time to really dig into a lot of stuff. But what are your top priorities as we look out over the next, call it, 3 to 5 years?

Continue our momentum in go-to-market. We've really transformed our go-to-market, built customer success teams, more technical teams. I think we have had positive revenue performance because of that. So it's keep going on the -- that's number one. Two is we're going to unleash a new era of innovation in IBM, spending a lot of time now on productivity, our development sites, our technology road maps. We have the opportunity. We certainly have the technology for IBM. [Audio Gap] In other cases, it's other technology, but that's the business model. We'll continue that. Fourth is kind of like we talked about M&A. We'll obviously continue to be opportunistic. Keep our eyes open for that and fit this with watsonx. AI and data platform is a critical inflection point in the industry. I think we've learned a lot through the years. We can win with watsonx. That would be a big focus.

Awesome. Well, unfortunately, we've got to wrap it up over here. Rob, thank you so much for being here. Really appreciate for all the thoughts.

Thank you Wamsi.

Thank you.