From AI Hype to Enterprise Value: A Fireside Chat with Cisco’s Chief Strategy Officer Mark Patterson

Patrick Moorhead: The Six Five is On the Road here with Cisco in Los Angeles. Dan, it has been a great show. AI, the conversations are very similar to ones we’ve had as we’ve gone across the globe talking about AI.

Daniel Newman: It’s been a really interesting couple of years. I think we’re starting to see a lot of the technology discussions materializing though into enterprise use cases and people starting to get a lot more focused on measurement and a little bit less caught up in just what the tech is and how it works and really how is it powering companies. And that’s a lot of what I’ve been hearing here as we’ve been spending time with the folks at Cisco and their partners here in Los Angeles.

Patrick Moorhead: It’s interesting, I think we always said here on the show and in our research that this was a marathon, not a sprint. And if we can look on the map and say we are here today, AI with enterprises is just moving from experimentation to POCs. And with POCs you’re going to have multiple POCs and then moving into mass adoption and scale. And there’s a lot of planning that’s involved here, it requires vision, strategy, objectives. And the rate at this is changing, it seems like it’d be super difficult to be able to plan a business about this. And here we have, Mark, great to see you. It’s been a while, I think seven months here we saw you at MWC to talk about strategy. You do run strategy here at the company.

Mark Patterson: I do.

Patrick Moorhead: And I’m sure you are busy mapping out short-term, long-term, do we build it, do we partner, do we acquire, and all that fun stuff.

Mark Patterson: That’s right. It’s been a fun seven months. I’ve worked now with the ELT though, for, gosh, almost seven years, so it’s a lot of fun. One thing for sure is that things are moving fast, and I feel like hopefully partners have seen it this week at Partner Summit that we are moving at a different pace as Jeetu said on stage, tempo matters. And it really does, particularly for a big company like Cisco, we got to move like a startup, but still have scale.

Patrick Moorhead: It struck me that final comment after we got up there, hey, here’s my philosophy, but here’s what Chuck says, which is speed. And I’m like, classic Chuck.

Mark Patterson: And then Chuck yelled to him on that. He said, “Hurry up.”

Patrick Moorhead: That’s good.

Daniel Newman: There were some pretty good quips too. We love you by the way.

Mark Patterson: You’re the first to say it in that way.

Daniel Newman: Well, thank you. I will also be the last, we’ll get it on the way out. But you started to talk about tempo. You talk about speed, you talk about execution. This is a race, and I know people-

Patrick Moorhead: I just called it a marathon but okay.

Daniel Newman: It’s a race. And by the way, Pat, marathon-

Patrick Moorhead: It is a race.

Daniel Newman: … is a race. But the point is I think we’re also trying to figure it out as we go. There’s all these good anecdotes and analogies. Oh, it’s the build the plane while it’s flying thing, but we know it’s all in. We’re hearing these hundreds of billions of dollars of CapEx investment that’s being made right now. Cisco of course, is a huge beneficiary. The company’s also made some of the major pivots that it needed to make over the last few years. I’ve talked countless times with Scott Herren about this ARR pivot, and you’ve now seen 50% of more than $50 billion of revenue move to ARR, so you’ve turned that crank and you’ve now turned over that big moment. The next moment is becoming a company that’s really synonymous with AI. That’s the thing. Of course you’re there, of course you’re in it, but leading the strategy, how do you do that? What’s the vision for making Cisco an AI company and what are you doing? Tell us a little more about the strategy behind your AI.

Mark Patterson: For sure. At Cisco we believe we have a significant role to play, obviously. And connecting and protecting the AI era is the high level, if you will. We see really three areas of opportunity in particular. There’s tons of opportunities that are out there. The first is in the hyperscale space and the major cloud players. And that’s certainly been where a lot of the discussion has happened. The money that’s going into the GPUs and servers and the LLMs and training the models, et cetera. We’ve got a presence in every one of the major cloud players. We’ve taken over a billion dollars in orders to date already in that space. The second is really as that moves from hyperscale or cloud down into the enterprise, and I think this is the year of the enterprise actually, 2025 will be. And so what we’re really trying to do is take what we’ve done at the highest level and bring it into an integrated stack. Very simple plug and play for the enterprise.

And so that’s where you’ve seen a lot of the announcements from us come. And then the third area is frankly just preparing for these enterprise applications. We do a readiness survey every year, and it won’t surprise you that I think 95% of the enterprises say, “Hey, we understand that the load that’s going to be on the networks, that’s coming is going to be so much greater than it’s been, and they’re going to have to have so much more scale, et cetera.” And there’s only about 17% of them that say they have the networks that can scale to that today. There’s a lot of preparation in just preparing the networks, preparing the data centers, re-architecting them, bringing your security posture up to snuff for these applications that are to come. And so those three are the big areas of opportunity for us. Obviously we’re investing in a big, big way here. We’re redeploying hundreds of millions of dollars of operating expenses within the company in terms of organic expenses to AI. We also announced here in July actually, or June, it was at Cisco Live, our billion dollar investment fund that’s all around AI, so we feel things are moving fast and you need to invest to win.

Patrick Moorhead: Two years ago when Gen AI kicked off and the fireworks started going off, my company made… I don’t think it was controversial then because I agree it’s panning out, but enterprise generative AI is going to be very different from hyperscaler generative AI and fundamentally with 80% of the enterprise data on-prem, I don’t see them behaving differently. We’re not going to magically move all of this data to be activated up there. And then for the enterprise, you need utmost of security. You need to have fast networking, and you have to have a handle on your data and it’s data, so I like the way that the Cisco AI strategy is coming together. But I’m curious, a year from now, I made the statement maybe you agree or disagree, hey, we’re moving from a lot of experimentation to POCs, we’re not at scale yet, but what do you foresee for the next year for the enterprise with AI?

Mark Patterson: I think that one thing is clear, and we’re hearing this from customers all over the place, and I’m hearing it from my own CEO, is there’s a lot of pressure on companies to move, so they know they need to do something that’s like, what are we doing in AI? What are we doing in AI? What are we doing? And so there’s still, I think we’re early innings, you called it a marathon. We will traverse across many sports here. But I think that companies are still, they’re figuring out, what’s my use case? What’s the ROI? How many GPUs do I need? What data do I need to actually bring together? How do I need to re-architect my network? How do I need to re-architect my data center, et cetera. They’re thinking through all those things but I think this is the year that you’re really going to start to see the use cases get deployed and really go into full deployment within the enterprise.

Daniel Newman: The challenges though, Mark, are substantial. First of all, the entire IT stack is changing. The AI infrastructure is creating… It’s birthing a new era. We’re hearing about AI data centers, AI factories. There’s different names that are being tossed around, but companies… And by the way, the speed of the hyperscaler versus the speed of the enterprise are diametrically opposed. Enterprises cannot move it nearly the same pace. Of course cloud gives a certain amount of leverage to moving in that direction but at the same time, as Pat said, you still have most of the data of enterprises that sitting on-prem. You’ve got legacy apps that haven’t been ported or lifted over. The challenges are significant. I’m just curious about you… Cisco is so deeply entrenched in enterprise, what are you hearing as you’re out on the road talking to companies about strategy? What are the challenges? Because if you were to listen to the advertising that’s out there, you can just deploy a container with a bunch of AI in it and everything will run and you’ll discover drugs and you’ll be able to-

Patrick Moorhead: Instant efficiency.

Daniel Newman: … build instant cities and engineer. But it’s not like that. I have to imagine you’re hearing the real challenges and wonder how you’re addressing them.

Mark Patterson: And I mentioned a few of them earlier. I think they’re focused on… Many of them are just focused on, how do I get started? Particularly a lot of the CIOs, how do I get… What’s a use case I can really drive and begin to learn from, if you will. A number of them are really wrapped around how much money am I going to have to spend on this and what’s the return going to look like? And I am also hearing many people say, you can’t focus on ROI in this space. You just got to go do it. This is one of those transformations-

Patrick Moorhead: Transforming the way the CapEx is rolling out right now.

Mark Patterson: It is.

Daniel Newman: Tell me the look that you see on the CFO’s face when you hear that in a room.

Mark Patterson: Well, you guys talk to Scott, Herren.

Daniel Newman: Do you see the blood just drain out of their face and they just turn ghost white?

Mark Patterson: Exactly. There’s data sovereignty issues, there’s security issues. Security keeps coming up over and over. I’m going to bring some of the highest value data that we have as a company together for training a model, using it for inferencing and what have you, and how do I know that that data’s going to be actually protected? How do we know that that’s going to be secure? How am I going to monitor this model? And so I’d say those are just a few of the challenges that companies are facing these days.

Patrick Moorhead: It’s good. And simplification, by the way, we heard a lot of talk about with pods we’re simplifying this, security-based, baked into the network from the start. And now the overlay of all the data that you have in there is, I think, super key here in figuring that out. As we see with every major inflection, you have companies who start it and who want to take it all. And enterprises are like, I want the best of both worlds here. And what it really takes to move something forward as an industry is multiple companies working together in some collaboration. Maybe after a while they decide to… They’re competing at first and then they shake hands on stage and they do both. And I’m curious, what is Cisco’s role in the broader AI ecosystem? What are you bringing to the table to move the industry forward?

Mark Patterson: That’s a great question. I think there’s a couple of different areas where we’re going to need to see some really tight partnerships and collaboration happen. Certainly in the GPUs, that’s a space that we’re partnering with AMD and Nvidia, and there’s a number of companies that we have invested in through the billion dollar investment fund that we set up as well that are all up and down the AI technology stack because we recognize that customers are going to need more than just a point product, if you will. They want integrated stacks. They want things that are going to be very custom for their vertical or their industry that they’re in, et cetera.

And they need us to work together. And I also think it’s going to be important for us to look to some of the industry players as well. For instance, in oil and gas, I was just at a customer that is a major oil refinery customer that has upwards of 10, 11% of the capacity for the entire world’s oil and gas pipelines. And what they’re doing is they’re using large language models to actually listen to a lot of the equipment that they use and the pumps and the drilling operations, et cetera. And it turns out that just by listening to those drills and the operations that you can actually identify when a problem is about to happen and do predictive maintenance. I think that working with some of the vertical players is going to be a key thing for us to do going forward as well.

Patrick Moorhead: Side note, I’ve been tracking your build by invest and you even do seed funding. That was a new thing that you rolled out a few years back, but you have a very robust pipeline to get that tech in. It looks like you’re continuing that with AI and not just on a horizontal basis, but it sounds like a vertical solutions’ basis as well.

Mark Patterson: A big reason why we set up this investment fund, if you will… And so that is just purely for investments, just to clarify as well. We’ve also made acquisitions and we’ll continue to make those. And 2024 alone, I think we’ve made 11 different investments and acquisitions that are in the AI space, if you will. But one of the things that we look to do when we make these investments is first of all, just learn. Stay close to these young growing companies that are in this emerging space. Many of them, we have a board observer seat or a board seat itself where we participate in and really learn a lot from that company. Many of them, we have actual agreements that we’ll enter into that are commercial arrangements whereby we’re either going to use their product in our solutions or we’ll jointly go to market together, et cetera. But the bottom line is we see working with the ecosystem overall will allow us to bring a much more robust and better solution to our customers in the end.

Patrick Moorhead: I appreciate you outlining that. Not everybody has been tracking it for the last 10 years but I have.

Mark Patterson: Yeah.

Daniel Newman: Mark, it’s a very competitive landscape, and throughout this conversation, we’ve embedded the spirit of sport many times. We’re all Baylor fans in this group and in football-

Mark Patterson: Bears.

Daniel Newman: … we’re winning the Big 12 is a battle for the Bears. That’s been problematic for a long time. Occasionally they pop up to the top, but here’s another battle that is going to be fought, is becoming synonymous on the tongue of people with AI. I heard a little bit of that throughout presentations is like, how does Cisco… Because obviously you’ve been powering network, you’ve been powering… You’re there, but at the same time, some companies have leaped into the forefront of being that name with AI. I think Cisco wants to be there. As the head of strategy, what are you telling everybody out there about how does Cisco get there from here?

Mark Patterson: Well, I think that while a lot of the money has been made early on in GPUs and servers and things like that, and a lot of buzzword technologies, if you really look at it, it’s the infrastructure and the network that really powers AI that if you talk to some of our largest customers that run the most complicated complex networks on the planet with some of the big cloud players, they’ll tell you that the single biggest defining factor in whether a AI project is successful or not is the network. I think that the way we can bring infrastructure, and we can bring data, we can bring security together along with, we’re here at Partner Summit this week, along with our go-to-market and our enterprise presence and the trust that we have with our customers that we’ve built up, I think arguably Cisco’s got the most complete and robust AI offering in the industry, and so we’re pretty excited about the opportunity going forward.

Daniel Newman: Well, as the arbiters of the industry, Pat, and I’ll be tracking that closely and be sharing our thoughts. Of course, we’ve got a lot of data intelligence, and we’ll be keeping in touch with you behind the scenes on what we’re seeing in the market, of course. Mark Patterson, thank you so much for joining us here on The Six Five. Thank you.

Mark Patterson: Thank you guys. Appreciate it.

Daniel Newman: And thank you to everyone for tuning in. Hit that subscribe button, be part of The Six Five community. Join us for all of these conversations here with Cisco, and of course, everything here on The Six Five. For this episode though, for Patrick and myself, time to say goodbye. See you all later.