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

Hey. Good afternoon, everyone. Thank you for joining our Virtual Tech Conference. I'm Wamsi Mohan, IT, hardware and supply chain analyst at B of A. We're delighted to have Rob Thomas, who's SVP of IBM Cloud and Data Platform with us today. Rob has previously held roles as GM for IBM Data and Watson AI and has had a long career at IBM. So we're delighted that, Rob, you are able to take some time and join us here today. Welcome.

Wamsi, great to be here. Thank you for the opportunity. I hope everybody is safe and well on your side.

Yes, absolutely. Same to you.

So let's go ahead and dive in, Rob. So in April, you were actually announced as SVP of Cloud and Data Platforms within the IBM Software segment. Can you share a bit of your new responsibilities?

Sure. Arvind, when he took over the CEO role, described IBM strategy around hybrid cloud and AI. So think of my role as I'm focused on the software components of that, and that is the segment that we call Cloud and Data Platforms. So I operationally manage the majority of IBM's software portfolio, including things like data and AI, security, cloud integration, IBM design, the Red Hat synergy work that we do and our offerings across Multicloud. And I work closely with the different software application areas, so things like health and IoT and supply chain because the hybrid cloud architecture that those applications run on, that's what I'm responsible for delivering. So for Cloud and Data Platforms, I lead unit strategy, product design, offering development, marketing, field operations, that type of thing. As you mentioned, or as you alluded to, I joined IBM in '99, started in consulting, did that for a few years, was in IBM Microelectronics for a good bit as we were building out the high-tech services business in Microelectronics. And then I moved to Software in 2006. And I've been in data and analytics since then until this new role. I did different roles, led business development, led product engineering, and then I had general management responsibility for data and Watson AI. And now the new role, which began in April, so excited about that.

No. That's perfect. It's great that you bring so many perspectives to the table. So Rob, Jim, your CFO, talked a lot on the call in April about the dynamics in the business, including specifics for software. How has the demand environment changed? And how is IBM helping clients deal with this pandemic?

Well, if you look at the first quarter, through February, as Jim had alluded to, we were tracking right in line with our expectations. And then as we got into March, you could see a noticeable change in client priorities as clients were reacting to the situation. So there was a bit of a pause that occurred then as clients were trying to figure out what does this mean for them, focused on resiliency, those types of things. And so many of the software transactions that would typically close in the last couple of weeks of the quarter were delayed a bit. As we saw COVID-19 become a reality, the shift in client behaviors were to things like automation, things like security, new ways of working. And these are spaces that we're pretty familiar with. So we immediately operationalized a new set of initiatives to align to that. One example would be Watson Assistant. So that's one of the applications that we have inside of Watson. And we took some CDC data, and we were able to train Watson into an offering that we called Watson Assistant for Citizens that could basically help any organization deal with COVID-19-related questions and guide people. And just to give you a sense for the rate and pace, I mean, we've already rolled this out now to more than 58 clients in 26 countries. There's another 100-plus in process. And this is all demand that did not exist before COVID-19. So it gives you a sense for how fast technology can be applied when organizations are motivated. And it's a range of -- from people like City of Austin, who launched this to Children's Healthcare of Atlanta, and even public health agencies around the world like the Polish Ministry of Health. And so to me, it was really encouraging just seeing clients be able to go live in as short as a couple of days and really respond to the situation. So that was kind of around virtual agents, automation, caring for their customers. In terms of the security space, we saw much heightened interest in specific applications related to work-from-home, which I know every company dealt with. So just to give you a sense, in April, we saw a 14,000% increase in COVID-19-related attacks in just a 2-week period. And that included things like ransomware in hospitals. And so we quickly mobilized and say, how do we make our technology available, things like IBM Security MaaS360, which includes Watson, our cloud identity. We started to make those available to clients at no charge just to get up and running. And again, we're able to see clients go live very quickly. One that stuck out to me was a bank in Egypt, Attijariwafa bank. They launched a solution which was end point management, MaaS360 with Watson, and they were able to safely give their hundreds of remote workers instant access to resources that they could use to serve clients. So it just gives you a sense for when client behaviors change, obviously, you have to react quickly, which I think we were able to do. But you also have to have offerings that are fit for that environment. And so we felt pretty good about how we were able to mobilize at that time.

No. Those are some very impressive stats. So just to follow up on that, Rob, when you say that some of these offerings were offered for free, what sort of parameters are around that? Like is that free for several months? Like how should we think about the duration for which you're helping your clients at least kick-start their initiatives?

Yes. Look, it's a mix. I mean we -- specifically related to COVID-19, we launched a number of offerings that were free for 90 days, basically just to eliminate any barrier for a company getting started. And obviously, unrelated to COVID, we have a lot of offerings that have a freemium model. And it's a way just to give clients an experience with products and get comfortable, and then they can go through their normal buying processes. So specific to that, it was around 90 days. We also did stuff for universities, like we made Watson Studio, SPSS, data science tools available to academia who were suddenly dealing with remote student situations.

Okay. Great. So IBM has been talking about the hybrid cloud as a destination now for many years. How do you think the strategy will evolve under this environment? And can you talk to why, for clients, IBM should be the destination of choice?

Yes. So first of all, hybrid cloud is going to be the destination for enterprises. I would say we're very confident in that. And Arvind talked about this a bit in April when he talked about seeing companies go through the shift to remote work, automation, what that means for their applications and how they modernize them. And hybrid cloud is obviously coming to the forefront. And I would say this kind of starts with our overarching thesis, which I would describe this way. If you think about technology workloads, they've continually evolved from centralized to decentralized architectures over a long period of time. Like think about it, mainframe was very much a centralized architecture when it was first launched. As we got into the '90s, everything decentralized with the client server. That continued with things like mobile and the Internet. If you look at public cloud in its purest form, that's actually a return to -- that's a move away from decentralized back to a centralized architecture because many of those are walled gardens, proprietary capabilities. And look, I think the seesaw will always occur in technology, but we don't believe the right answer for clients is going to be a walled garden centralized architecture. So when we talk about hybrid cloud, it is about moving away from that centralized architecture of a stand-alone public cloud back to capabilities that can run on your private cloud, across multiple public clouds. I think over the arc of IT, clients have always preferred open architectures, decentralized IT, integrated data and applications, and that's exactly what we're doing when we talk about hybrid cloud. And if you kind of think about where we are today, given the acquisition of Red Hat and how we've integrated that technology across, even how we're building, we've covered the full stack for hybrid cloud, from the operating system with Linux, through Kubernetes containers with OpenShift, to middlewares and applications. And so we are the only company that can provide, I would say, the capabilities across security compliance for that entire stack. And we think that's a great advantage. Just maybe to spend a minute on containers, I mentioned that, and it's a key part of that. Let me just define that a bit. So think of containers this way. This is about how you simplify software development, software deployment, and the benefit for a client is just the ability to make new software available faster. And the analogy I would use, which is actually where the term comes from, is you think about shipping containers, physical ships moving containers around the world. ISO set a standard back in 1968, I think it was, where packaging goods became standard from the home, to the truck, to the port, to the ship. And by standardizing on one way to do this, the impact of the outcome was commerce, meaning global commerce could occur at a scale that it couldn't before. Software containers is a similar idea, meaning virtualization was about virtualizing the hardware stack and containers take that to the next step for the entire stack of software and hardware. It breaks the software to smaller micro services with a shared operating system. It just gives clients a lot of flexibility for how they think about managing their applications and their data. IBM now has over 100 products that we've containerized, and we'll be doing more in the future. And just to put it in a strategic context, there's a quote from Ben Thompson that I love where he talks about making a platform worthwhile is about the applications that run on it. And that's exactly what we're doing with Cloud Paks, which I think will turn a bit. But our strategy, to be clear, it's hybrid cloud and AI. OpenShift is our platform for hybrid cloud. And then Cloud Paks is what enables clients to bring things like AI, security and other capabilities onto OpenShift.

Okay. That's really great perspective. So if we dissect this a little bit, can you share a little bit about Red Hat? And how have you and your team been working with Red Hat? And where are you seeing the progress? And what progress have you seen so far?

For sure. The short way I would describe it is I prefer Red Hat in how I build products, but Red Hat does not prefer me. And that's really how we're operating. When we announced the acquisition, we said, look, this is not an integration, this is about synergy between the 2 companies. Because we want Red Hat to continue to operate independently and work with the whole ecosystem of partners they work with. So when I say I prefer Red Hat, that means we are optimizing the IBM software portfolio on OpenShift. And so therefore, I'm a really good partner of Red Hat. And it works well because they can maintain independence. And over time, I can start to benefit as well from the broader ecosystem. In some cases, we're merging capabilities. IBM Cloud Private, that's what we were doing on containers before the acquisition. We've merged that with OpenShift. So in some cases, we merge capabilities. And I think it's important to think about the broader context. Red Hat's focus has always, I would say, largely been around applications. You think of JBoss, automation tools. So something like the data world, where obviously we have a big presence, is new to them. And so anything that we do in our database portfolio, AI portfolio, data governance, that drives cores over utilization of OpenShift, which is a great point of synergy. And the same is true for things like security, where we're providing security to clients on the Red Hat platform, things like identity, data security, threat management, where we've integrated with Ansible. And it also extends to things like RPA, business process management. So these are all capabilities from IBM that get layered on top of OpenShift to provide clients with a great experience. One of the products we've announced recently, the newest Cloud Pak, if you will, is called Cloud Pak for Security, which is about how a company can gain greater insights across the fragmented landscape of all their security tools. So Cloud Pak Security deploys, leveraging OpenShift and Ansible, and suddenly, if a client is using 1,000 different security tools, they can integrate that into one tool, one capability, which we think adds a lot of customer value. And you can see the momentum picking up. One of the metrics I just look at is how are we doing when we run customer events and attendance. And we just held our Think Conference, which was the largest we've ever had. And historically, we've seen about a 5% overlap in terms of the audiences that attend the Red Hat Summit and IBM Think. And this year, that was closer to 25%. So whether it's partners or clients, I would say, many are hungry to understand how they can play a part of what we're doing around how these pieces come together because it is, again, back to where I started, it's all about that journey with hybrid cloud and AI.

I see. Okay. No, that's helpful. So just to touch on Cloud Paks, Rob. You announced this around a year ago or so. Should we be thinking that these solutions are modernized versions of your middleware? And will we continue to see more solutions launched in the form of Cloud Paks?

So I would think of it this way. This is organic development, so it's new development. And it is modernized hybrid cloud. And we do show any of our existing middleware clients, we show them the path to easily on-ramp to our Cloud Paks. And in some cases, it's containerizing those capabilities. In other cases, it's just bringing that capability into the Cloud Pak as part of a modernized deployment. So I would describe it this way, think as if Cloud Paks are pre-integrated IBM middleware that's containerized and where the value really comes in is when you get to a common standard around logging, monitoring, management because it really becomes, let's say, all the flexibility and integration that you would expect in a cloud deployment, but it can run anywhere, meaning public or private. Today, we've got Cloud Pak for Data, for Multicloud Management, for Applications, Automation, Integration, Security, and all of these are those capabilities delivered on OpenShift. Just to give you a sense of the flywheel that gets degraded here is -- let me give you an example. If a client is using Db2 as their database, when they adopt Db2 running on Cloud Pak for Data, they immediately get a containerized modern version of Db2, but they also get access to all the capability that's inside the Cloud Pak, which is things like Watson Studio, Watson Machine Learning, Watson OpenScale. So they're getting a whole different set of capability, a data governance catalog beyond just the database. And so we continue to see strength in our database business. And I think the flywheel is part of what creates that interest because clients get a whole new level of capability. I saw in your recent survey, customers were positive in terms of their reaction to how we're making our middleware software available. I think it was something like 40% were expecting a slight or a significant increase in spending. I think it's because people are genuinely excited about the direction that we're taking, where they get the best of what they love about our middleware, but then bringing that forward to containers and these open architectures. Just to give you some examples. Sorry, go ahead, Wamsi.

No. I was just going to say, Rob, that, yes, you're right, like we picked up definitely some improved mindshare in our survey results more broadly even across IBM, even on the Consulting side. But as you point out, there were some much more positive sort of uptake, both on the database side as well as interest around Cloud Pak and solutions that are -- customers who are either customers of Red Hat or IBM before, wanting to do more on software-combined basis. Sorry, finish it, Rob, I don't mean to interrupt.

No, that makes sense, and I appreciate that. I was just going to give you a couple of examples to bring it to life. Like Sprint is a client that's adopted Cloud Pak for Data. They had hundreds of different repository storing data, and they wanted to bring that together into a single view to support their aftermarket service. And using Cloud Pak for Data, they could collect data from all those sources, virtualize data and then build AI models using Watson to run across all those data sources. So it just completely changes how they can deliver on their analytic needs. And then there's companies like Talktalk Group in the U.K., a telco service provider, adopting things like the resilient capability that's inside a Cloud Pak for Security, basically just to give them a better view of what's happening in their security architecture. ENN is an energy company in Asia, adopting Cloud Pak for Automation. That's for RPA, so robotic process automation, plus Watson, to automate now 3,000 different tasks that before were done manually. So it's bringing to us a new set of clients in many cases. In many cases, it's clients that are using our core middleware that are now adopting more. So we're really optimistic on this.

No. That's great. So IBM made a number of announcements around Think, and one of those that, in my mind, stood up was AI for Ops. Can you share a bit of the evolution you saw that led to this newest offering? And are there other trends that are being reprioritized as well?

Look, we're always monitoring what's happening to figure out where we drive investment. This actually started in IBM Research. And the observation was the biggest cost to many clients is issues in their IT infrastructure. And I saw an estimate of something like $26 billion of lost revenue, in some cases, due to IT downtime. And part of the reason for that is every company has many disparate tools. And so the thesis we had was can we use AI to bring together a -- think of it as a semantic layer across all the IT tools in an organization to enable them to understand what's happening in their IT environment, to predict problems before they happen and then to solve problems if they do happen. And so Watson AIOps, which is what you're alluding to, we announced at Think, which is basically an application that provides that semantic layer across all of the IT systems in an organization. And just to give you a sense, in terms of the experience that it drives, is what used to take 5 hours to deal with, you can now accomplish in 10 minutes because you're not having to go into every system and figure out what every system is saying differently. You can basically do it in 1 place. What would take 17 steps now takes 4. What took 10 people to respond to, you can now do with 1. And so we've been able to drive down the cost per incident from a couple hundred thousand down to $5,000 with -- in the early going with Watson AIOps. We launched this with partnerships, which is, I would say, part of the whole mentality for how we're building products. So we launched this with Slack and with Box. And just to give you a sense, the UI for Watson AIOps is actually in Slack, meaning you're just in that environment. And if something happens, Watson creates an alert that shows up in Slack, tells you how to fix that problem. And so I would say it's a really unique way to deliver a product, but everything we're doing is focused on open standards, integrating with the ecosystem. And so we're really positive on this one.

No. That's great. So regarding AI and Watson, right, you guys have done a ton of work. I remember there was a Project Debater that was sort of really interesting in terms of addressing some of the most difficult topics in AI. I mean to have -- be able to go in, pull up and understand not just what someone is saying, but to comprehend it truly and then to come back with a formulated response is actually not just sort of a program chatbot, it's something much more sophisticated. What sort of technology innovation has helped do that? And how are you guys thinking of leveraging that technology in other AI offerings?

So to step back for 30 seconds. So Watson today is 2 things. One is it's a set of tools for any company that wants to build their own AI. And they can kind of pick and choose from a menu of which capabilities they want. We've decomposed the capabilities as micro services to make it really easy for anybody that wants to build AI to use those capabilities. Watson is also applications where when we've seen a common use case, things like call centers, we built Watson Assistant. For things like data discovery, we built Watson Discovery. For financial planning, we took our planning analytics product to put Watson inside there. So think of it as tools and applications. And so when we look at something like Debater, the process we went through working with IBM Research is what is the best way to expose this powerful technology that was built around a debating use case but for a business. And what we realized is there was really good technology in there on Natural Language Understanding, understanding PDFs. So what's actually happening in a document, not just the words on the document, but the context of those words, how they're being used. And so we decided to make these capabilities available in Watson Discovery, which is one of the applications, and is also available in Watson Natural Language Understanding, which is one of the tools that I referenced. So when we have great technology like this, the objective we have is to make it as accessible as possible to both builders and to the consumers of AI. And so that's kind of the difference between tooling and applications. So if you look at the stuff in Debater, it's really powerful for being able to summarize a huge corpus of documents or data. And if you think about businesses, businesses are basically a summation of all their communication, whether that's text, chat, e-mail or documents. And we think this is really powerful.

No. That's great. There have been a lot of client questions around how has Arvind's appointment change things around IBM. And I was wondering if you could address both sort of Arvind's role as CEO as well as Jim Whitehurst as President. How are you seeing the culture of IBM change or evolve?

Culture, I would say, is defined by behaviors over time and can certainly be heightened in a crisis like the current health and economic crisis that we're all facing. I would say it this way. Ginni really set the tone for us always around client centricity. The way that she would engage with clients, it's something that will stick with IBM forever and it taught us a lot. And Arvind is obviously a different person, but it's building off of that, meaning he starts with the client, just like Ginni kind of taught us, then he brings the candor, transparency. I mean his first communication to all the IBM employees was on LinkedIn, which is different. He's been actively engaging across the ecosystem, joined the earnings call. I mean he does -- he gets on Slack and does Ask Me Anythings with the employees. And it's really -- it's opened up the dialogue in IBM. And look, Arvind is very technical. He can go 5 meetings deep on any technical topic, but his strategic intent is also off the charts. And so employees love hearing from him and engaging in that kind of dialogue. And then to kind of add to that, you bring the other dimension of Jim Whitehurst. I mean he coined the notion of the open organization in the book that he wrote. And so he's engaging a lot with the teams around culture and innovation and how we bring agility to how we're building products and engaging with clients. So I would say it's a combination across all 3 that is really building a great culture here, which we've always had, and just building on that.

No. That's great. And I can definitely attest to the 5-layer deep conversation that you just said about Arvind. We've spoken to him in the past from what is the right architecture around AI and you need floating point processors to do fast matrix multiplication or can the algorithms be written for doing fixed point processing. And he can definitely go 10-layers deep on something like that. So it's really impressive. Well, we're almost out of time, Rob. So I just wanted to ask you, just thinking more long-term, how do you think the technology landscape, including for software specifically, will be changed by the recent events here in 2020?

Look, for recent events in 2020, it's -- to me, it's just about acceleration, meaning what clients were going to probably do anyway, now maybe they're considering looking at that on a compressed time frame, which is obviously a great opportunity for IBM. Our focus is on innovation. I gave you a couple of examples where we're talking about technology coming from IBM Research, getting into products and getting into the clients' hands on a very compressed time frame, which I think is great. And as I look out at the market, there's a lot of different areas that we're investing in. Just a few I'll mention just to kind of give you a forward-looking view, things like homomorphic encryption, which is analytics on data without having to decrypt it first. That's a big area of interest, something that we're looking at in research. You've got things like neuro-symbolic AI, which is kind of a new frontier around combining learning and logic and being able to answer questions with AI where you may not have all the data upfront or first. That's an interesting area. We're doing a research topic that we call analog AI, which is we're combining compute and memory on a single device to eliminate the volume and bottleneck between compute and memory. We just open-sourced some technology in IBM Research called SysFlow, which is a new system telemetry format for monitoring system behavior for security compliance and performance analytics. So these are all new frontiers and you're going to continue to see IBM bringing technology through research, into products and then surrounding that with services capabilities in the market, which we think is a recipe for success. In the shorter term, the phrase we've been throwing around, it's about how do we help clients work smart and work safe. Everybody is starting to have the discussion around what it looks like to return to the office and maybe being in a hybrid work environment. And so a lot of that's about adaptability. And to give you an example, we just adapted our TRIRIGA offering to help employees prepare for return to the office. So we can actually monitor occupancy in a building, help them create temporary space plans. So it's about building off of what we do really well today and -- but at the same time, investing in new frontiers and research that can have a big impact in products.

No. That's great. Unfortunately, we're out of time. But Rob, we really appreciate you taking the time to be here today. I appreciate all the thoughts and look forward to catching up soon, hopefully in person, and best of luck in the new role.

Thanks, Wamsi. Great talking to you.

Thank you.