The State of Hybrid Cloud – Futurum Tech Podcast Interview Series 004
We all know about the Cloud, but what about the Hybrid Cloud? In this Interview Series of the Futurum Tech Podcast, Daniel Newman sits down with Robert Christiansen, VP and Cloud Strategist at HPE. They discuss the recent Futurum/HPE survey on the State of Hybrid Cloud and how global enterprises are managing their migration to the cloud and the important role of hybrid public/private and on/off-premises networking and computing systems. From the importance of agility to understanding how to measure the value of cloud services, this discussion covers the key points around data and application portability that enterprises need to know to successfully development and deploy “native cloud” applications and business processes.
Daniel Newman: On this edition of the Futurum Tech Podcast Interview Series, we have Robert Christiansen, Vice President, Cloud Strategist at Hewlett-Packard Enterprise, to talk about the state of hybrid cloud. We’d like to thank Hewlett-Packard Enterprise for being our special sponsor of this episode, and we look forward to you enjoying the show.
Welcome to the Futurum Tech Podcast Interview Series. I’m Daniel Newman, your host, and we are excited today to be joined by HPE’s Vice President of Cloud Strategy, Robert Christiansen. Robert joined HPE through the acquisition of Cloud Technology Partners, but I’ll let him tell you all about that.
But on today’s show, we’re going to be talking about a recent survey done in conjunction with HP’s Pointnext on the state of hybrid cloud, a Pulse Survey, talking to 166 respondents. But before we jump into that, let’s go ahead and have Robert say hello and introduce himself. Robert, welcome to the show.
Robert Christiansen: Hey, thanks for having me, Daniel. I appreciate that. So, hey, little introduction about myself. I’ve been with Cloud Technology Partners since late 2013. I came on as the first public cloud knowledgeable person. We’re a born-in-the-cloud consulting company, so we built the global practice for onboarding clients into the public cloud. Then, when HPE bought us, they were introducing us to a hybrid strategy.
At first, I had to say I was a little taken back that there was such a thing, but with deeper investigations and conversations with clients and stuff, you get a broader perspective. I have to say the hybrid world is real, it’s a key part of what we’re doing today, and I’m really excited about this conversation. As a leader of our global practice for how to use and consume cloud, no matter where it is, HPE has been really, really headfirst into making that a reality and I’m just glad to be on here with you today.
Daniel Newman: Yeah, absolutely. Welcome. It’s been good. You and I have had a few chances to work together, we’ve done a webinar before. I think we shared a coffee talk in the past at a HPE Discover and I believe we might get the opportunity to do that again at the next one. If it’s not with you, it’ll be with one of comrades, but this topic is a very important one. It kind of bleeds all things together: public cloud, on-premises, manage services. There’s so many things going on, but let me just start by sharing with everybody about this survey that we did.
HPE came to us and said, “We’ve got a few different things we want to explore and one of these is we’d like to explore the state of hybrid cloud. We’d like to do a quick survey.” At Futurum, we do a lot of big studies that take months and months, but the interest here was more of, “Let’s just get a taste of where people are at. Let’s talk to them about how their cloud transitions are going, their use of different clouds.”
We talked to 166 respondents, primarily the great majority of them were decision makers, two-thirds in the C-suite. A lot of them in North America, but we had distribution across the globe. We talked to those that were highly involved as well in cloud services as influencers and we talked to almost 55 percent, which were the primary decision makers. We talked to folks from more than eight different industries, and that’s in a combination, so everything from banking and finance and retail to healthcare pharmacy, manufacturing and more.
Mostly talked to companies of greater size, so we didn’t talk to anything smaller than 1,000 employees because we were definitely trying to focus on companies that were most likely to be using multiple types of cloud. But we did talk to 42 percent of companies that were 1,000 to 5,000, 41 percent in that 5 to 50,000, and 17 percent of our respondents were at over 50,000 employees.
We asked them about what defined their cloud strategy, what percentage of their applications are in public cloud, how their adoption has been going, what have been the primary factors driving their interest in cloud, what are their biggest concerns, what does their cloud practice lack, and then, of course, we wanted to understand what could best define their cloud strategies.
There was a lot of really interesting finding, but rather than just going through all this, and by the way, we can share a link to the webinar where you and I did a lot deeper dive on the data itself. I’d like to use this as a little bit more of a conversation piece because I think the data, and for those of you that are interested, like I said, I really think there was a great webinar and you can link right in and see the deck and the data.
But Robert, you have a wealth of knowledge of what’s going on in cloud. You’re facing customers every day, you’re talking to enterprises that were represented in this survey. So, talk to me a little bit about what are the current trends that you are seeing as it pertains to cloud strategy and the enterprise.
Robert Christiansen: So, your timing is absolutely perfect for this podcast by the way, Daniel. I just got off the road this week. I did two separate strategy workshops for executive leaders of IT and business units around their hybrid cloud strategy, so two of them. The way we structure the workshops, they’re leadership forums. They tend to run anywhere between three hours to a day depending on what they want to get done.
This week, we did two. One was for a top-ranked bank in the Top 10. They’re stepping seriously into it and they have a lot of alignment issues. The other one was with a retailer in the grocery/retail space. So, two seriously different things. I thought the dynamic there was really interesting.
The way we start every workshop or forum is that we go to each person in the room on the client’s side and we ask them, “Hey, what’s your name and what’s your title,” and then, “What do you really do?” Their title says one thing, but what do you really do during the day? Then we say, “What do you hope to get out of this workshop?” So, we spend the first hour just writing down all their questions, and those questions reveal what’s happening. The first one, we had 28 questions just asking probably 12 or 13 people.
It was really interesting to see and it aligns with the data that you and your organization produced. That was, number one, “How do we decide where the application goes based on two issues?” One, “I want to get agility, I want to have the ability to move quickly and to respond to them.” Number two, “How do I know that I’m economically getting the best value and watching my cost, watching my spend?”
I think those were the two big takeaways we got around the data that you and I talked about. That was number one. Those were confirmed there. The third thing that was right up there at the top of the list was, “How do we enable our business units, our teams, the people who are asking for the resources now? Central teams have not been able to respond to us quickly. How do we get them to a space where they start developing true cloud native applications or refactor the ones they have? How do we bring them up that scale so that they can do that and mature them?”
That obstacle includes training, that includes right tool sets, what’s a secure DevOps pipeline look like, how do they build out with their CICD processes? All that stuff that’s necessary for development. Those three topics were just at the height of everything to do. So, you had all these spidered topics that would come off it.
I want to just double click on that, and that happened both on the retailer side and on the bank side. So, it was common issues.
Daniel Newman: So, they’re sort of starting with the principle, the basics, right? “How do we find value in it? How do we measure value in it? How do we deploy it? How do we make it simple?” The starting points, right? But then it sounds like they’re starting to drill deeper. “How do we secure it? How do we make sure the data can be encrypted, can be managed, can be secured, not only at rest but in flight? How do we move data between clouds and work loads?” It sounds like they’re getting deeper, the questions are getting more sophisticated.
Robert Christiansen: They are. So, Daniel, I want to touch on that last one you just said. They’re making assumptions on application portability so that they don’t have vendor lock in, and they’re asking questions about how do they achieve that. One of the things … And this is a high desire. They don’t want vendor lock in, they want to be abstract, they want to able to allow for SLAs that [inaudible] for fail over or some DR strategy or high available strategies that allow them to move from one platform to the other.
The realization that they all came to was data has gravity and that unless you have a data portability strategy, you don’t have an application portability strategy, because the apps are going to go where the data’s going to be.
Daniel Newman: And they’re so interdependent.
Robert Christiansen: Oh, they are. And data has physics, right? This is the part that I have to revisit with them. Latencies, speed of light. At today’s world, FedEx is still the fastest way to ship terabytes of data. It still is, there’s no getting around it. You got to put it on trucks, you got to put them in disks, you got to ship them off some place because over the wire doesn’t get it done. Yeah, so, I just want to confirm our findings that you and I talked about in that webinar. That was a really great webinar and it’s just coming up through every client conversation we’re having.
Daniel Newman: Yeah, it’s so interesting though that people don’t always make that connection. This is going to become a massively important issue with this movement to the edge though, because everyone’s applications and things to be running at the edge, and you talk about lighter weight. But the thing is is that it comes back data. So, when you’re talking AI and such, a vehicle making a decision, the compute has to be there but so does the storage and the data, because if those things aren’t all there concurrently like you said, you’re going to have a car calling into the cloud and being like, “Well, is that a person or is that a shopping cart? And should I stop?”
I mean, I know that’s the example everyone likes to use, but those are the big decisions. But even the simple decisions, right? As we try to entrust a vehicle to move human beings around, the data and compute and everything being close enough is going to matter. The timing is going to really matter, so I understand that’s not necessarily a hybrid cloud comment in itself. I think as applications continue to proliferate and the way we want to utilize things, companies are going to have to learn about this interdependence, so I think that’s a really, really great point.
I kind of just mentioned, I know it’s not a hybrid thing, but you kind of talked about some of these trends, but you started as a public cloud.
Robert Christiansen: We did, yes. That’s what we did. Yeah, public cloud.
Daniel Newman: And you were sort of saying when you first came in, you were like, “Huh?”
Robert Christiansen: Exactly.
Daniel Newman: “What? Why did anybody want to do that?”
Robert Christiansen: Yep.
Daniel Newman: But we’re seeing almost a renaissance, right? We’ve seen research come out in markets talking about repatriation. We’ve seen more and more … Obviously HPE’s GreenLake has been a winner with many customers, moving certain workloads on premises. The whole hybrid story OneSphere and helping companies in managing between the clouds. What are you seeing in terms specifically in related to the way companies are looking into public, private and hybrid?
Robert Christiansen: It’s confusing for them right now. They want to … Well, let’s go back to what they’re responding to. This is the whole point of on hybrid. The reason why the hyperscalers, AWS and Azure and GCP, Google’s platform, and others for that matter, have made such a dramatic impact is because they were serving a need, and the need was this: development community were not getting the resources they need, or the innovation they needed, to compete in the market.
So, they came up and they said, “Hey, if you’re not getting the right stuff from your internal IT, come and buy from us,” and that’s exactly what they did. They never sold to the internal system, the IT organizations, the central teams. They always sold to the developer community and as a result, they circled the whole central teams and they have forced these teams to change.
Given that, that force, that forcing function, change has to happen. Now, the realities though of that change have kind of peaked, meaning that the central teams are saying, “Hey, you’re going to need a bunch of data here to make that application you’re trying to write right now on top of Azure work.” Well, the only way that’s going to happen is if we replicate the data out and create a hybrid strategy so that we’re doing live live data between on-prem and in the public so that that application can work, or you bring that application back in, repatriate it and run it next to my data center.
The problem with that is that once they do that, they lose all the flexibility they were trying to get, thus there you have a standoff. Now you have a real problem. You want to be able to innovate and the platforms, the hyperscalers, are offering you that, but if you really look at where the data lives, the important data, what I call the “deep client data,” most of it’s all on-prem. It’s going to take years, if not a decade, to populate exterior to the public cloud in a safe way.
That means all the public clouds have become aware of this problem and you’re seeing them now change their stance on cloud first. It’s now a hybrid strategy, all of them. Number two, they have started to push their platforms back into the on-prem world, so it’s kind of like the camel sticking its nose under the tent, right, Daniel? You’ve got Outposts, which is exclusively a AWS hardware product that sits at the edge of the data center. You’ve got Azure Stack, which has been out for a while, with moderate success. Then you’ve got the new announcements from Google with GKE and Anthos on various vendor’s platforms, including HPE.
They’re all boomeranging back, so that’s what you’re talking about. Now you have a really interesting set of challenges with the hybrid strategy. Where [inaudible] live, where are we going to do it, how are we going to execute?
Daniel Newman: Yeah, it is a really, really big question. In our research, we basically saw they’re was big mix of what people were doing right now. Of all the companies, just a quick run through some data here, 7 percent were running kind of a cloud first strategy, and 15 percent consider themselves all in on public cloud with one vendor. Another 14 percent said they were all in on public cloud and going multicloud, but that’s still all those three categories together. The purest were only about a third.
Robert Christiansen: Mm-hmm (affirmative). That’s exactly right.
Daniel Newman: The other 11 percent that were all in on private cloud. Those were the people you’d start to think banks, healthcare, places that haven’t quite figured out how to not be on-prem. Then, what we saw was a much a larger, the middle of the pie, over 54 percent, were balanced public/private, were balanced with one leaning towards public, one leaning towards private, and then a very hybrid strategy. About 15 percent believed they had a really pure hybrid cloud with multiple public, private and on-off premises strategies, so you could see companies are all over the place.
Robert Christiansen: That’s an interesting state, that last one. First off, I think that clarity on … I think one of the things that you and I talked about was the definitions, right? There’s so much going on right now with just words being thrown out there and so much marketing stuff happening to the world as well. I think you and I agree that we’re saying, “Hey, okay, what’s the definition of some of this stuff?” It’s all of those spaces, number one.
The second one, I think what is interesting, Daniel, is there were outliers. They were all in on cloud or they were all in on private. Those were outliers. The majority had some form of hybrid, whether it’s hybrid multicloud, hybrid single cloud, hybrid data. It was just a mashup, and I think it’s accidental right now. It’s not intentional nor is it planned.
Daniel Newman: Yeah, well, I think a lot of that, too, right? Because of application consumption, because of operational buying centers that are outside of centralized IT where cloud is being purchased, SAS applications are being deployed, just because of a lack of continuity, strategy and speed. This really comes back to that digital transformation conversation, which, by the way, we don’t have enough time on this podcast. Boy, could we talk about how those two things are interdependent, right? What kind of IT setup you end up with based upon how your company is able to move and how flexibly your business units and IT departments are able to collaborate.
But I did want to talk about something else because as I said, timing is of the essence here and man, has this word been big in the last three months of my life. It’s been big over a year, but in the last three months, I’ve heard “multi cloud” more than I’ve ever heard it before and as an analyst … A year ago in Forbes, I said, “This’ll be a big thing next year.” Three months ago, I said, “This might be the biggest thing coming up as of right now.” It’s everywhere. Again, you went from the public cloud guy to the hybrid cloud guy. What’s your thoughts around multicloud? How is this impacting strategy workloads, etcetera?
Robert Christiansen: You’re seeing the development community really exercise their muscles. The people who produce revenue or impact for the company is the development community, not central IT teams. Central IT teams are there as a center to run and support the organization’s technologies.
Daniel Newman: They’re enablers.
Robert Christiansen: They’re enablers. Exactly, right? But the people who are writing mobile apps and edge apps and applications that keep people alive and banking, all that kind of stuff, all those value things, are software developers, and they’re going to get their way. They always get their way.
What ends up happening then is that they say, “Oh, look at Azure. They have a whole bunch of new features that came out.” For example, we’re talking this week to a retailer, and you can’t even talk about DWS with these folks because they won’t even discuss it. It’s only Azure or possibly GCP. So, they’re going to say, “Hey, here’s all the Azure AI and ML services that are available for us to use, some decent client behavior analytics.” So, they’re going to be off doing that, and some other group is going to say, “Hey, well, we found a whole bunch of nice stuff that we want to use on GCP that’s going to give us some different deep analytics and some behavior conversations that we want to do around machine learning and image processing to try to compete with what’s going on with AWS’s …” I forgot the name of that store that they just introduced where you just walk out with stuff, there’s no checkout. It’s all image based.
Daniel Newman: Just walkout technology.
Robert Christiansen: Right. So, not even that retail space, so how do you … They need to be able to do that, so they’re trying it on GCP, right? Think about what that does. Now immediately, you have two cloud vendors, multicloud, and they have amazingly different operating models, so the people in the center, the central IT teams, are going “Wtf,” right? They go, “What are we supposed to do?” We have two separate teams trying to manage it and they’re just stuck in the crossfire, so this is going to happen faster, Daniel, like you said. It’s accelerating, this multicloud strategy. And then how does that push back on to on-prem as well? So, now you get this hybrid.
Daniel Newman: Yeah, well, first of all, you were talking about Amazon Go, second of all, you bring up a really interesting point about retailers and their reluctance to work with Amazon. Of course, you should probably be reluctant. Although the companies are very disparate, they certainly … Amazon.com is certainly a significant contributor to challenges that other retail businesses have. It’s a decision each company has to make, and Azure has a great product.
As you mentioned though, each cloud does have some specific things, maybe it’s around security, maybe it’s around data portability, maybe it’s where their pops are located, maybe it’s their pricing for certain types of data workload portability, maybe it’s their integration with containers, which I didn’t even bring up because I don’t have time, but it was probably the next thing we could have talked about is how big the container business will become as a relation to the multicloud business. Things are going to happen in a big way.
It is kind of like this weird cycle between best of breed and everything from one location, right? Right now, multicloud is sort of the enabler of best of breed and then there’s these companies building technologies that really pull it all together. I know HPE is trying to participate in that ecosystem as well by making sure that the technology and the software really enables customers to pick and understand which cloud would be the optimal for each workload, as well as which on-premises solutions could be optimal.
By the way, who knows? Maybe one of those retailers wants to use a Cray Supercomputer [crosstalk] –
Robert Christiansen: Yeah, we kind of got that note today, didn’t it?
Daniel Newman: – Some deep analysis. So, yeah, I got an opportunity to do a briefing on that, but I had like 15 minutes to do it and I was busy, but I am really excited to learn more about it.
Robert Christiansen: Oh, me too.
Daniel Newman: I’ve already socialized it and shared more and I think, “Wow, what a great opportunity to get into the federal space and to academia.” And of course, as enterprises want to get more and more into analytics and AI, that type of compute power is going to be super important. Unfortunately, not the topic of our podcast, but heck, I don’t think anyone at HPE would mind that we mention that. And to the Futurum community, as analysts, it’s our responsibility to tell you guys what’s happening out there, so that is something that happened today that was a big thing.
So, my last question for you before I let you off the hook and let you get on with your day here, Robert, is what are some of the key recommendations that you’re giving to companies right now that are defining building and enhancing their cloud strategies?
Robert Christiansen: One of the big takeaways that we have been advising on is understand that where you park your data is where the apps are going to live and to de-emphasize app portability unless you have a strong data portability strategy. That means small pieces of data that can move with apps, or some way that the data can be pushed to a platform and then thrown away. We’re trying to educate people that where your data lives is where your apps are going to be and typically, we have to deconstruct their whole … We don’t want to be hooked in and have vendor lock in. That’s just not … Everybody asks for it, they never get it, so we kind of ground them into reality there. That’s one of the big, big takeaways, so that’s a principle around their whole strategy. That’s one of them.
The second one is that the more you have people developing multiple cloud applications that are going to go to production, this is important, you’re going to have a multiplier effect on the internal IT support organization, the IT operation organizations, that are going to cost significantly more to manage, and we try to ground them in that. We try to encourage them to force everybody, or at least get everybody into the same hyperscaler, whether it’s Azure, AWS or GCP. We try to get them to select one and go deep with that one on a basic operating model and then standardize on that. Then if they have other requirements coming in and it’s an exception base and then there’s going to be other things around that.
The third, the last one, is creating an organizational construct, a business office, we call it a cloud business office, by which they knock down the silos of all the various teams and serve the business with this new agile motion. That’s very important so that they can organize correctly and get people trained and get the motions necessary for driving this programs over the next couple of years. So, this is the big three that we’re doing right now.
Daniel Newman: It sounds like some great insight. So, for everyone out there, definitely … Robert Christiansen, a ton and wealth of experience and we really appreciate you joining us here today on the Futurum Tech Podcast Interview Series. Always love getting the outsiders to come in and talk a little bit with us about what you’re seeing out there. HPE, as a customer, as someone we work with on research and analysis, we value the opinions greatly. We hope everyone out there that’s listening to this show has a chance to hear what Robert’s saying. Definitely bringing you insights from the field.
For this episode of the Futurum Tech Podcast, the Interview Series, I’d like to thank everybody for stopping by, for joining us. Please be sure to subscribe to us on iTunes and stay with us each and every week as we bring more interviews and more shows from our weekly Futurum Tech Podcast. We look forward to hearing from you, seeing you here again, and talking to you soon. For now, I’m Daniel Newman, Principal Analyst, Futurum Research. I’ll see you later.