In an exclusive interview with CRN US, the 19-year Cisco CEO offers up candid views on everything from Cisco’s rivalry with VMware, his succession plan, the Intercloud strategy and why Satya Nadella is “perfect for Microsoft”.
How are you getting over the perception that Cisco is too hardware-focused?
Results. First you outline vision and strategy. The majority of our engineers – 80 to 85 percent of them – are already software engineers. And if you watch the areas that are growing, it’s software.
Now what is [Application-Specific Integrated Circuits]?
ASICs is hardware and software combined. You put in the hardware, something that doesn’t change [and] you put in the software, something that changes all the time. ASICs is something that will be a combination of the two. So we have our software and hardware teams sitting side-by-side, developing ASICs. And before it even comes out, we are already starting on a second generation of custom ASICs where you put more software into it to get the speed.
Watch our scale and how quickly [our Application Centric Infrastructure] takes off versus our competitors. In fact, I would actually argue our competitors in the SDN area have very few true, paying customers and very few examples of scale.
When you say ‘software networks’ can’t scale, are you talking about VMware NSX?
I’m talking about players like that, yes. And so, what we will do is take those concepts and introduce new concepts like self-learning networks. It was something we started about 15 months ago; you would have never heard about it if it didn’t work. We took a couple handfuls of people and said, ‘How do you really do self-learning networks for the Internet of Everything, for optimisation of the network and for security reasons?’
So you will see us embrace this software and then play it out with merchant ASICs, which is how other people do it, but also with our own custom ASICs where we embed it even deeper and it goes faster. We let the customer [decide] on openness, whichever way they want – and hardware. That leads to the Internet of Everything, which is really going to be the driver of business productivity for the next decade.
There’s a lot of partner interest in Arista. Are they now one of Cisco’s top competitors?
Basically, Arista is one of several dozen competitors we face and one of several dozen that are in the top peers in a product category that we compete against. We compete against them in architectures. We compete against them in terms of how you combine software and custom ASICs and merchant ASICs and hardware. And we compete against them in terms of embracing SDN in an implementation that is able to scale and able to do it with security.
Now let’s deal with the numbers. The product was just shipping – the Cisco Nexus 9000 – two quarters ago. Twenty-five customers were [adopting] it. Last quarter, [it was] 175. I can’t share with you what this quarter is… But just watch our number of accounts. And watch how quickly we blow right by Arista – either revenue-wise or number of customers.
VMware said recently they have 150 NSX customers. That sounds pretty on par with ACI adoption…
First, we already have 175 customers, two quarters in, with what we are doing. That’s two quarters in, not five years in. Secondly, are those [NSX] customers really in scale and are they paying for it? In other words, are they large-scale implementations and are they paying for it or are they just taking a segment of their enterprise licensing agreement and allocating it to NSX? I can’t tell you that. You can all decide.
It’s a nice way of saying ‘I love our hand versus VMware’ and ‘I love our hand versus Arista’. And while two to three years ago you could say we were slow in embracing in SDN, now we are going to lead. We know how to scale. We know how to tie back to Application Centric Infrastructure when it isn’t about the data centre but about the WAN and the edge. This is how we compete against them.
Will those 175 Nexus 9000 customers definitely deploy ACI?
In fact, they have put [the Nexus 9000 switches] in first, in standalone [mode]. And then you will see a number of them move to ACI. So we will track those ACI customers.
But remember, ACI just ships this quarter. So we are in the middle of just starting at the end of the summer, and then it’s the implementation done through it.
But that is where our selling mission has changed. We don’t sell SDN. We sell programmability, investment protection, lower cost of ownership, and quicker outcomes, and we combine all these.
What’s the biggest transition partners should be undertaking?
The way they deliver outcomes. The way we deliver outcomes. So it starts with how do you talk to customers in discovery mode, and then it’s how do you make this relatively transparent from discovery to outcomes? And how do you bring each of Cisco’s franchises or, if you will, architectures to life to be able to deliver on this?
It won’t change. In my view, I clearly understand that services is a huge part of making this work. But my ratio is 5:1. For every [dollar] of revenue we get for services, I want my partners to get five, and say, ‘How do we together deliver outcomes?’ That’s the transition that I think we have to make as an industry, and this is how you compete against bare metal, white-box, new startups and individual players.
Some wonder if you are going to be CEO long enough to see these transitions through…
I think you are going to see Cisco maintain a very crisp vision, strategy [and] execution model. We will adjust as we go through it, but if you watch these transitions we have just made, we have really set our strategy for the next three to five years. And if you watch, as a company, we have changed every player at Cisco during my time here at least three to maybe even seven times and partners didn’t even notice. We are a partner-centric company, period.
We changed services [leads], we changed CFOs, we changed marketing leads, and multiple engineering leads. We do this very well. And this has allowed Cisco to stay on top for two decades. Who else has stayed on top and moved into new areas with anywhere near the leadership that we have? I’d say no one.
Partners are a huge part of making that happen. Now, each time we move, they have to move with us. And, if you remember, a decade ago most partners paid no attention to services. Today, it’s probably 50 percent of their average revenue and maybe as much as 70 percent of their profits.
Talk about steps you’ve taken to build a smooth succession plan.
If we do this transition right, you won’t even notice. My goal is to make this one of the best transitions there has ever been in high tech, and I know a number of them haven’t gone very well over the years. But this isn’t something we have just been focused on in the past year. We have been focusing on this for a decade, on how we build the relationship and a strategy that the company buys into and is implemented by the company. I have the broadest base of leaders and the strongest leaders I have ever had. And I have moved people out of the company who were passive aggressive and not strong team players.
How big of a threat is the white-box trend to Cisco’s core business?
If you are selling a standalone router or a standalone switch, especially if you are doing that with merchant ASICs, your differentiation – whether it’s a server white-box or a storage white-box or network white-box – is very little. When we saw this coming, we moved to architectures so you combine the technology of the network, the custom and merchant ASICs with the software, with the hardware, with the storage and do it throughout the network, and you focus on outcomes.
Customers will pay a premium. If you are buying the cheapest product, that’s probably not where we are going to win. If you are buying the best total cost of ownership and highest probability of outcome, we’re tough to beat.
What percentage of Cisco’s R&D budget would you say is for hardware versus software?
Well, because we bring them together, it’s a hard number to break out… but just hardware engineers are probably less than 15-17 percent.
With Intercloud, some are saying Cisco is late to the public cloud game. How would you respond?
The key takeaway here is Cisco is the disrupter. We are going on the offence. We are not playing defence. We are not going to respond to people who are challenging. We are going to say, ‘Here is what the customers want and we are going to deliver it with faster speed with our partners than anybody else can.’
What are the partner economics around Intercloud?
Intercloud is a step beyond a public cloud. Think of Intercloud as Cisco services delivered from a cloud: hosted communication, security, collaboration. Think of it as our customers’ private cloud capabilities. Think of it as our partners’ and Cisco’s architectural implementation of this, if you will, in terms of our partner clouds. And think of it in terms of a public cloud.
Our role is we see all four of them growing well, and we are going to play in all four. So Amazon is a customer, a partner and, at times, a competitor. But it’s a public cloud. Then our customers’ own private cloud is what we help build out.
Now, if we can move around workloads based upon guaranteed response times, security issues, etc, to meet the customer outcomes that they want – that’s what Intercloud really does.
You can maintain, if you wanted, data sovereignty in a country. And, at the same time, that offers a huge competitive advantage.
So this is classic Cisco. It’s not that we are saying, ‘Here is what we are doing with our own cloud and here is what we are doing with our customers’ private clouds and here is what we are doing with our partners’ architectural clouds built on Cisco, and here is what we are doing in public.’ It’s how we bring the hybrid, the private and the public together and how we do it in a way that really accomplishes what customers want.
Some partners fear Intercloud competes with their own cloud services. What do you say
First, as you all know, I poll [partners] at each of the partner conferences constantly. A year ago – in fact, it’s been almost a year-and-a-half since that partner conference – I asked [partners] how many thought we should move into this market. My team thought everybody would say ‘yes’ but I knew they wouldn’t. When I asked back then, it was 50-50. And when I asked this year, it was 95 percent. We, of course, looked at the five percent who didn’t want to do it, and we immediately went to them. I won’t say who they are, but suffice to say we went to them.
So watch how partnerships like the one we have with Dimension Data evolves to see where we’re going. And the one with partners like IBM. So I think all partners just need to understand how they are going to evolve here and why this is in their best interest and ours.
You’ve said the NSA revelations have impacted Cisco’s business abroad. Are they still?
Let’s move way beyond the NSA. Let’s talk about what’s so important. Customers have a right to know that their data is secure and so do the citizens and so do the companies. And this is not an issue unique to the US. Every country is involved
in espionage. That’s what they do. They have done it for a long time.
What needs to occur is a standard of conduct and rules of the road that will guarantee to customers the supply chain and data integrity that they need. And I think that’s where countries have to step up and begin to provide these guidelines and an outline to that.
As you know, we don’t give anybody our core software even though a number of peers do, because if you get the software you can eventually, with processing power, figure out how to break it. We don’t provide backdoors. And we focus on outcomes. And if we find anyone – no matter if it’s hackers or governments – involved in any of our customer environments anywhere in the world, we tell our customers, period. And we do that in the US, in Europe and China and India. And we have done it.
So NSA spying has had a direct impact on Cisco?
It does have. And it’s mainly in a couple of countries, but it’s the overall security issue where, instead of saying this is a problem, you reserve it. We are going to become the No.1 security player globally. We’ll use concepts such as Intercloud for data sovereignty. You begin to say, ‘Here is the problem we can solve’ and then, at the same time, you work with governments to say to the leadership in Europe and Germany, ‘Here is what we can do together.’
Think of how much of the supply chain comes out of China or out of Asia. What has made our industry so powerful is the complete globalisation of the supply chain and the ability to move, and for every citizen and every country in the world to potentially benefit from it.
So this is where I think a lot of people tend to look in the rear-vision mirror. In my letter to [President Obama] I said, ‘It’s interesting how we got here, but it’s even more important how we go forward. How do we solve these problems together?’ Part of them we will solve on our own. We are going to be very direct in terms of protecting our customers’ data, information, etc. and very direct in terms of if we find any issues, regardless of where they come from, to turn them over to customers.
Has the US government given any assurance it’s not tampering with Cisco gear?
Well, remember, this could be any gear by any company anywhere in the supply chain. And I don’t know of any government that has given those assurances. So this is where I would like to see countries come together to say these are the rules of the road or the business conduct. Now, having said that, we are going to do the same. We are going to make the equipment very difficult to tamper with. We are going to ship it with a lot of information on it, and we are going to ask ‘How do we do this better than anyone else?’ But, remember, this is just one of 100 different ways people can gain access to information. That’s why I think you need some rules of the road and that’s in every country’s best interest.
What are your thoughts on investors pushing for EMC to spin out VMware?
Well, the last thing I would ever try to do is give one of my peers and a good friend, [EMC chief executive] Joe Tucci, advice. So I think looking at how we approached our business and how multiple players in the industry, not just EMC, but HP, IBM and others approach theirs, we chose a different path.
If you look at what we just outlined here, every one of our product areas is now completely coming together in a total architecture. So routing isn’t separate from switching, isn’t separate from collaboration, isn’t separate from services, isn’t separate from mobility, isn’t separate from security, isn’t separate from the Internet of Everything, isn’t separate from Intercloud, isn’t separate from ACI, isn’t separate from compute with storage with the network. They are all tied together so that our customers get the benefit and so that our products have a premium. Our ability then to continue to transform the organisation gets results that companies who lead in these silos or are loosely coupled together do not get.
So we took a different approach than almost everybody else in the industry in terms of [having] 18 major product families and then said, ‘How do you bring them together in architectures that allow you to get a premium, a quicker response from your customers, and a partner community who will standardise on you, and outcomes?’ We chose a different approach here and I would argue we are getting the benefits of that. I don’t know anybody who would look at Cisco and who really understands our business – especially the moves we have made in the last few years – who would argue that breaking the company into pieces makes sense.
You personally congratulated Satya Nadella when he became Microsoft’s CEO. What do you think he brings to Microsoft?
I normally don’t comment about my peers, but I am going to make an exception. The industry has not done a good job with CEO transitions. If you watch – often when CEO transitions have occurred in high tech over the past 30 years, especially if the CEO had been the person who grew the company or the founder for a number of years – they have been rough. And you can probably count on one hand those that even went OK versus really good.
Now, Satya, I think, is perfect for Microsoft. If you watch, and watch what he has already done, he comes from a very strong technical background. He understands their culture. It would be very different to bring in somebody from outside to run Microsoft, even though people have different views, because what goes on is so complex. Yet, you had to have somebody who was a change agent and somebody who, given their culture, is pretty technical in terms of the direction. I found him to be a breath of fresh air in terms of his willingness to make changes. And while some people might second-guess him on his strategy or his implementation or his decision to size the organisation right, those are decisions that leaders have to make. You can’t lead if you don’t make the tough decisions.
Talk about how Nadella might impact or change the Cisco-Microsoft partnership.
We have probably done more with Microsoft, with Satya, in the last year than we had in multiple years before that. And it wasn’t that we didn’t get along well with the [previous] leadership at Microsoft; we got along very well. But watch our data centre announcements together. Watch what we do in terms of the market. They were on stage with us at our ACI launch. Sayta was. So it’s a hard job and I don’t underestimate the challenges that Sayta will face, but if I were betting on whether he’s going to be successful, I think he absolutely will be.
You noted how tough CEO transitions can be. Why do you think Cisco can pull it off so seamlessly?
I think you have watched how our company has evolved and we have been very much focused on this. We knew that this would be a hard issue 10 years ago. And if you watch the breadth and depth of our team, it’s the best it’s ever been. I am empowering my team more, and sometimes it gets tremendous results and sometimes it makes mistakes. But if you don’t give them a chance to learn, then their ability to lead as a team won’t be there.
So we have been very much focused on how to make these transitions smooth and have been working with our board and others over the last decade and especially over the last three to five years.