In this podcast, Heidrick & Struggles’ Nick DeMagistris speaks to Jim Miller, board member and advisor at TheRealReal, Brambles, and Deep North, former CTO and board member of Wayfair, and former VP of worldwide data center operations at Google, about the innovation and leadership trends in the data center and public cloud industry. Miller discusses what the next five years could look like in the space and shares some advice for the next generation of leaders in digital infrastructure companies. He also offers his thoughts on how his experience, perspective, and relationships as a board member of Wayfair helped prepare him for his role as CTO.

Some key questions answered in this podcast include:

• (0:57) Would you share your overall take on the data center industry as a whole? What are the trends that you've seen? What's next? What is the next five years? What are the headwinds?
• (4:20) Do you see a saturation point? A point where AWS just can't grow anymore? What do you think slows them?
• (7:03) Intel is starting to build chips for cloud computing, and you’ve got start-ups doing the same thing. Do you feel like that will be a competitive differentiator for the hyperscalers, where if they can build their own chips and build it well, it will set them apart?
• (8:38) What advice would you give to the next generation of leaders of cloud computing or digital infrastructure companies?
• (14:53) Do you think your board experience, both on the Wayfair board and other boards, helped to better prepare you for the CTO role? Please tell us a little bit about the different perspectives you had to have in a full-time operating role versus what you have to focus on in a public board role.

Below is a full transcript of the episode, which has been edited for clarity.

Welcome to the Heidrick & Struggles Leadership Podcast. Heidrick is the premier global provider of senior-level executive search and leadership consulting services. Diversity and inclusion, leading through tumultuous times, and building thriving teams and organizations are among the core issues we talk with leaders about every day, including in our podcasts. Thank you for joining the conversation.

Nick DeMagistris: Hi, Nick DeMagistris here, a partner in charge of Heidrick & Struggles’ Boston office and co-leader of our global digital infrastructure practice. On today's podcast, I'm talking to Jim Miller, board member and advisor at TheRealReal, Brambles, and Deep North, and former CTO and board member of Wayfair and VP of worldwide data center operations at Google.

Jim, Welcome. Thank you for taking the time to chat today.

Jim Miller: Thank you, Nick. It's great to be here.

Nick DeMagistris: So, Jim, I wanted to kick the discussion off and ask for your overall take on the data center industry as a whole. I mean, you've seen it from both sides. You've seen it from the hyperscaler provider side at Google, the user side from Wayfair and all the cloud compute needs there. I'm kind of curious—what are the trends that you've seen? What's next? What are the next five years? What are the headwinds? I'll turn it over to you. I've got some more detailed questions I can pepper in, but I'd love your take at a high level to start.

Jim Miller: Sure. You know, it's amazing to me. I'm now approaching 60. I've spent my entire adult life in data center and computing technology. And we're really seeing this generational shift to hyperscale computing. You really have to give kudos to Amazon who started this all out. And now, you've got this ecosystem that's being built with the likes of both Azure and Google GCP and a number of other players.

The thing that I find most interesting is that first of all; it's not game-over by any stretch. You know, I'm a former Googler, but I'm neutral in most of this. There is still plenty of room for innovation and evolution. We're nowhere near seeing where this will ultimately end up.

Maybe the first wave is to do a lift and shift. Wayfair certainly went through this when they went to GCP. They basically took their applications and largely lifted them into a hyperscale environment to take advantage of rapid capacity utilization and ramping up, among other things, moving fixed costs to a variable cost and then taking advantage of many of the resources that the hyperscalers have, particularly in areas like machine learning and artificial intelligence.

Now, what you're seeing is this whole ecosystem being built up around people building bespoke solutions for all areas of the stack. Snowflake, Datadog—we can go on and on and on about everything that is being built right now. Obviously, the whole content delivery ecosystem and network.

Then you've got a bunch of tier colocation providers and things like that. We’re seeing more of the world talk about hybrid multi-cloud. That's going to have its own shakeout and evolution as well, but I think one of the most interesting things happening right now is a re-architecture around cloud-native apps. I think everybody's doing that right now.

And then—and this is interesting because I talk to a lot of my counterparts across the industry—I think many of them are still struggling with the fact that they've taken these large compute jobs and applications. They move them into the cloud, but they still have a very difficult time doing things around capacity planning, cost planning, budgeting because they really don't understand fully how to actually inflate and deflate and then be able to plan capacity performance.

And then of course there’s the cost around all that. I think that that is a huge opportunity across all the hyperscalers and probably all the cloud infrastructure because that tends to be very homegrown. And again, whether you're on Azure, AWS, or GCP, everybody struggles with really understanding what the true cost is to compute and the cost to serve.

I think there are many, many areas that will continue to evolve. It's going to be interesting to see how the whole ecosystem evolves into a true native cloud architecture and infrastructure and applications.

Nick DeMagistris: You mentioned AWS. I saw recently they had a 37% growth year over year. They beat analyst predictions.

They're growing like crazy at a scale, $30 billion, $40 billion. It just continues to blow my mind. Do you see a saturation point there, Jim? Is there a point where AWS just can't grow anymore?

Whether it's security concerns, latency concerns from government or healthcare clients—what do you think slows them?

Jim Miller: Well, you know, it's funny: when I joined Google, I worked for one of the founders, and his selling point to me to come to Google was, “Well, we're spending a billion dollars on CapEx every year, and in a few years we'll be spending north of 10 billion.” I laughed at his face and said, “If we get to half that number, that'll be a compelling job.”

Well, little did I know that through my tenure there, we would spend almost 70 billion on CapEx and eclipse that $10 billion CapEx mark annually—a number of times. To answer your question, if you think about where we are in cloud today, we still have a large number of laggards that are still looking at cloud and cloud technology.

So, you know, I think we're far away still from what we would call a saturation point, but I think there are really two compelling factors: right now, the ability just to get semiconductor and core processing technology. As everybody knows, there’s the worldwide semiconductor shortage. That's impacting many of the players right now.

In fact, I just saw something in The Wall Street Journal about how Cisco is struggling, and people are going to alternatives. I think that's going to continue to play out for the next year to a year and a half, at minimum.

Then I think there's this whole notion that (and I think this will come in waves of evolution) where people are largely still doing a lift and shift into the cloud. Either they've partially done a re-architecture into cloud-native app, or they're doing that as a second phase of their cloud migration (and that arguably is even more difficult). That's going to take time to do because it's largely a change management and a technology management problem, but also it will further enable this ecosystem that you build around cloud-native apps.

I think you're going to continue to see the growth for the foreseeable future. You've got a lot of companies—and I'm on the board of one of these companies—that have really just started their cloud migration. I wouldn't call them a laggard, but they're a more mature industry, and they've been slow to adopt cloud technology. But as all their systems, whether they're HR or finance systems, are going cloud, they're in some ways being forced to go to cloud. But they're also seeing the advantage.

So, I think you'll see this happen in waves. But I don't think we're anywhere near a saturation point in cloud technology yet.

Nick DeMagistris: You mentioned the chip piece. Amazon came out with the press release a couple months ago about Graviton. They're going to build their own. I think that's a fascinating component to this ecosystem, where now Intel is starting to build just chips for cloud computing. You've got startups like Ampere doing the same thing. Do you feel like that will be a competitive differentiator for the hyperscaler? If they can build their own chips and build it well, does that set them apart? If you think about the big three or Oracle as a big four.

Jim Miller: It's interesting. I think absolutely that's the case. I think about even Google, where we designed our own machine learning chips for both inference and learning. We really needed that because—and again, I think what Nvidia has done is amazing—the capability to go build a custom chip that is custom tailored to a specific workload or job type is critical because of this scale that you're operating at.

If you look at any of the hyperscalers, they have a growth book value well in excess of 10 billion of capital. The ability to optimize workloads in that type of environment goes straight to the bottom line. So, the ability to go and build hardware that's enabled for a particular job or workload and improve the utilization of that entire tech stack and ultimately that entire capital investment at that scale? You would be crazy not to do that.

I think this train's left the station and it will continue. You'll see more and more startups and more and more native investment going in with these hyperscalers to go build their own architectures and their own hardware.

Nick DeMagistris: When you think about leadership in this space—for today's leaders of these cloud computing or digital infrastructure companies, and then the next generation—based on your experience, what advice would you give the next generation of leaders?

Jim Miller: It's interesting. You know, I was probably a bit of a unicorn at Google, but I think it served me very well. I'm a technologist at heart, but at the same time, I've got a business school degree. I worked very closely with Google’s CFO because, for the amount of money we were spending on capital and OPEX built around cloud infrastructure, I really had to bring together the technical and the P&L.

I think in today's cloud world, you have to be very savvy with technology. And by the way, there are going to be pure play technologists that go build the next generation storage stack or the next generation ML hardware. And I'm not disparaging those people at all, but I think the leaders of this cloud industry really need to think holistically around not only the technology, the customer, and the applications that they're providing and the value that they're creating, but also the P&L. Because again, the technology and how it's implemented has profound implications on OPEX and capital utilization for a company.

Nick DeMagistris: You stepped down off the Wayfair board into the CTO role in late fall 2019, just pre-COVID. You’re in the role a few months when COVID happens, and you've got to deal with all of that. Can you talk a little bit about that experience from a leadership standpoint? As it became more of a remote environment, did the cloud computing needs change? How did the infrastructure needs change in that role going through that sort of transition?

Jim Miller: It was really a fascinating experience. The moral of the story is, if you're ever on a board and your board asked you to write a position paper on what the technology organization should probably do, you should probably refrain. That put a big target on my back: “Oh, we love your solutions and your proposals! You should be our CTO.”  That led to this circuitous path to becoming the CTO.

But in all seriousness, it was an amazing experience, and I would not have traded it for anything. At the time that I took over in the summer of 2019 as interim CTO, we were going through a large lift and shift into the Google cloud. We had made that decision. I was part of that decision as a board member. It was fascinating to go from the perspective of being a provider of that infrastructure at Google to being a board member to being one of the decision-makers to go to GCP and then to actually inherit the project of going to GCP.

And then, of course, a rapid global pandemic around it all. Everybody knows the dynamics that COVID had on eCommerce. It was amazing; it was probably the highlight of my career and a great capstone to go from being an advisor to an operator. We could not have done it—and this is not an advertisement at all for Google or GCP—but we could not have done it without being in the cloud. We literally completed the last piece of the lift and shift, the storefront move, just weeks before the meteoric rise of demand for eCommerce and for Wayfair, when everybody was starting to work from home.

Our ability to scale the business and storefront along with all the backend infrastructure and applications—had we not been on GCP and we'd been trying to do that on PRIM? I'll be bold enough to say I don't think we would've been successful. We got extremely lucky in that almost asymptotic growth to eCommerce adoption. We had just completed our storefront migration literally weeks before that happened. So, you know, sometimes it is better to be lucky than good, so to speak.

People are going to write books about the pandemic, and we still probably haven't seen how this all shakes out, but leading a large-scale transformation (both from an organizational and technology standpoint) during a global pandemic when primarily everybody was working from home was pretty fascinating.

The two major takeaways for me were, first and foremost, the resiliency that people demonstrated. Think about it: most companies like Wayfair went en masse to work from home in a matter of days, if not weeks. People did it, fairly transparently, under crazy circumstances: kids being at home, parents working, raising kids, all the concerns that people had about health and everything going on.
And then, in addition to that, I learned a lot about change management and communicating, even over-communicating and being radically transparent with people. It's funny: people expected their leadership teams to have all the answers, and I think there was some frustration in the beginning around the fact that we didn't. I remember having a town hall one day with a good chunk of the tech organization at Wayfair. I just said, “Look, let's be perfectly clear. We're making this up as we go along. This is so unprecedented. You all think we have the answers? You know, you think that we have this crystal ball? We don't. We're like you, making it up as we go along.”

I think they were probably equally shocked by that admission and pleasantly surprised that we were in completely new territory and making it up as we went along. I think there was a collective sigh of relief, like “Wow, if they don't have it figured out, that means that I don't have to have all this figured out either.” I think that took a lot of weight off the organization.

But we're still living in this endemic COVID world right now. I live in the San Francisco Bay area, and we're still not back to what I would consider to be normal. Monday and Friday, you know, where commutes would be hours long, there is no traffic. Clearly, people are not back to what probably is a new normal quite yet, so this will continue to play out.

Nick DeMagistris: Do you think—and I know San Francisco's been in the news a lot compared to places like New York for that reason—do you think it ever does get back to close to what it was pre-pandemic, or is it that that is the new normal in places like San Francisco?

Jim Miller: That's the new normal. It’s going to be interesting to see the impact that it has on residential and commercial real estate. Companies have been very public (and we know who they are) about saying, “We're a permanent work-from-home culture. We're allowing you to work everywhere.” I have real mixed emotions about that.

Nick DeMagistris: Do you think your board experience, both on the Wayfair board and other boards, helped to better prepare you for the CTO role? Talk a little bit about the different perspectives you have to have in a full-time operating role and then the perspectives and what you have to focus on in a public board role.

Jim Miller: That's a great question. It's interesting because when I took the CTO role, I knew for a fact that June 30th, 2022 would be the date that I would be done. I made that perfectly clear and that in many ways lit a fuse because I had a large transformation to do.

On being CTO from being a board member: I knew the company well. It's a founder-led company. I knew the culture well, and I knew the players well. But, I have to admit, I didn't understand all the nuances and the details as you would expect, being a board member. I was not an operating executive. It was a unique experience because going from being a board member to the CTO (even the interim CTO role, which I first started with) was really seamless.

It's interesting because, in many ways, it gave me a clinical detachment. Arguably, it was probably the most successful role that I've been in. And it was largely because, first, I knew when it was going to end and what I had to do by that end date. Literally, I would wake up every day and think, “Okay, it's t-minus one day to that date. Where are we relative to that transformation plan?”

There was no kicking the can down the road because I was adamant that whomever I handed my role off to, I was going to give them that role largely with that transformation complete. Second, I went in with (and maybe this is a combination of maybe being older and more mature) really no emotional attachment to the job, so to speak.

I knew that it would not personify my career. I didn't necessarily look at it and think, “Well, what's going be after Wayfair?” For me, it was like, “I've got a job to do. It may take two or three years to get done. It's a big job, and I’ve got to do it with clinical precision and emotional detachment.”

That's not to say that I wasn't passionate about it. I felt very strongly about what we were trying to complete, but it was a unique experience, and it was largely very successful. Going from a board member to an operating executive in the company that you're on, it's a very interesting and unique perspective.

I chose not to go back to the board because, first, I felt like it was the right thing to enable my successor to do the job. I'm not going to second guess you. Secondly, as a board member, you want to be independent from a governance perspective, and I could not go from being a board member to an operating member to back to being a board member and have any modicum of independence. For all my talk about emotional detachment, I probably was and am emotionally attached to it. So that just wouldn't work out.

Nick DeMagistris: Would you say the relationships you built both with fellow board members (when you were on the board) and later members of the management team—did that help smooth some of the ramp-up when you stepped into the CTO role?

Jim Miller: Very much so. I had the confidence of both the board and the executive team who I knew very well—almost too much so. I've got a couple of funny stories about this: I would go to the board and present this very large transformation plan. I told them, “Look, don't take my word for this. I want pushback. I want you to second guess me. You all think that I have all the answers? I don't.” I think they respected me too much at times and they deferred to me too much. I wish they would've pushed back a little bit more, but it ended up being perfectly fine.

Nick DeMagistris: So Jim, as we bring the conversation to a close here: Looking ahead, what specific leadership skill sets and capabilities will be most important for you in your board member and advisor roles?

Jim Miller: I think the best way to prepare is to really understand and talk to board members. I mean, first, have a skill that is absolutely critical for a board right now. If I look at boards, we've got a diversity problem with boards right now, but we also have a skills gap.

I think we're still living in a legacy where we've gone through a massive digital transformation and digital disruption. Many of the boards that I see have a dearth of digital and analytical skills and capabilities. I always say, “Hey, look, if you want to get on a board, make yourself useful and indispensable.” When you build a board (you know this better than I do), you're looking for an amalgam of skills and capabilities and an amalgam of culture.

You have to understand that (and this is probably the hardest thing—in fact, I was coaching a new board member last week about this) you're no longer an operator, and many first board members are still operators. They have to really develop that empathy and detachment needed to be responsible for, first and foremost, the governance of the company, being an advisor, a mentor, an observer, but knowing also when to back off being an operator.

You have to look at things and understand. I find myself understanding or looking at the culture, the speed that the company operates at, and then being able to say, “Wow, I understand the destination. I don't necessarily understand the route that they're taking or the speed at which they're moving. And if I were doing it, I would do it a different way.” Well, okay. As a board member, it's not effective to walk into your company and say, “Hey, I think you're doing it all wrong! This is how I would do it.”

You have to develop the empathy, the relationship, the trust, where you can ask questions, even obtuse questions. “Hey, why are you doing it that way? If I were to propose something, would you be open to it? What do you think about this? Hey, if I were doing that, I would do it differently, but I'm not telling you how to do it.”

I've grown up in cultures that were very direct. For me, I had to hone a different set of influence skills, where it was, “Okay. Let me step back and understand the culture, appreciate why they're doing it that way, and develop some empathy and objectivity.” I think that that's difficult for a first-time board member who walks in, guns blazing, saying “Hey! What are you doing?” I've seen that happen and it doesn't go very well.

But again, I think that that's part of the path and the journey that every new board member takes.

Nick DeMagistris: That’s such a good point, Jim. Thanks so much.

Jim Miller: My pleasure, Nick.

Nick DeMagistris: In my role, leading this digital infrastructure practice, it's an area of such unbelievable disruption. Just take the hyperscalers—they’re growing like crazy. They're their own category.

You talked about the hybrid cloud piece, right?

Jim Miller: Yep.

Nick DeMagistris: VMware, HPE, Red Hat—there's a whole subsection there. You've got the traditional DRTs and Equinixes, and how that all fits together. Companies’ cloud computing needs are different. We're excited about it; this is going to be a launch around, “Hey, this is what we're doing,” or “This is how we're thinking.”

Jim Miller: I have this group of classmates from MIT that all went off to do interesting things in their careers and largely are now kind of winding down, but we still talk quite a bit—Wilkie, the CEO of Amazon, and that group, and others. And you know, it's interesting because you think we could talk about ML and AI and data science. We didn't even touch on IOT.

It's amazing to me—this is going to take decades. I'll be long gone, probably dead, by the time this is all done. But you know, you just kind of look at it and think—I don't know. My father worked at IBM and was a senior executive there and really lived in the whole mainframe era. He's still alive, and he's saying, “Wow, I can't believe like seeing this whole thing and how it's evolved.” This is going to be still decades in the making. it's largely still in its infancy, I would argue.

Nick DeMagistris: Well Jim, thank you so much for your time here. You've had such an unbelievable career with so many different experiences in different areas and sharing your perspective on these topics is incredibly valuable. Thanks again for the time; I really appreciate it.

Jim Miller: Thank you. It was a pleasure, and hopefully, this will be helpful to somebody.

Thanks for listening to the Heidrick & Struggles Leadership Podcast. To make sure you don’t miss more future-shaping ideas and conversations, please subscribe to our channel on the podcast app. And if you’re listening via LinkedIn, Twitter, or YouTube, why not share this with your connections? Until next time.

About the interviewer

Nick DeMagistris (ndemagistris@heidrick.com) is the partner in charge of Heidrick & Struggles’ Boston office and a member of the global Technology and Data, Analytics & Artificial Intelligence practices.