 From theCUBE Studios in Palo Alto in Boston, bringing you data-driven insights from theCUBE and ETR. This is Breaking Analysis with Dave Vellante. Well, it looks like hybrid cloud is finally here. We've seen a decade of posturing, architecture, slideware, and narrow examples of hybrid cloud, but there's a little question that the definition of cloud is expanding to include on-premises workloads in hybrid models. Now, depending on which numbers you choose to represent IT spending, public cloud only accounts for actually less than 5% of the total pie. So the big question is, how will this now evolve? Customers want control, they want governance, they want security, flexibility, and a feature-rich set of services to build their digital businesses. It's unlikely that they can buy all that, so they're going to have to build it with partners, specifically vendors, SIs, consultancies, and their own developers. The tug of war to win the new cloud day has finally started in earnest, between the hyperscalers and the largest enterprise tech companies in the world. Hello, and welcome to this week's Wikibon Cube Insights, powered by ETR. And this Breaking Analysis will walk you through how we see the battle for hybrid cloud, how we got here, where we are, and where it's headed. First, I want to go back to 2009 in a blog post by a man named Chuck Hollis. Chuck Hollis at the time was a CTO and marketing guru inside of EMC, who remember owned VMware. Chuck was kind of this hybrid multi-tool player, pun intended. EMC at the time had a big stake, a lot at stake, as the ascendancy of AWS was threatening the historical models which had defined enterprise IT. Now around that time, NIST published its first draft of a cloud computing definition, which as I recall included language, something to the effect of accessing remote services over the public network, i.e., public IP networks. Now NIST has essentially or since evolved that definition, but the original draft was very favorable to the public cloud. And the vendor community, the traditional vendor community said, hang on, we're in this game too. So that was 2009 when Chuck Hollis published this slide. He termed it private cloud, a term which he saw buried inside of a Gartner research post or research note that was not really fleshed out and defined. The idea was pretty compelling. The definition of cloud centered on control where you as the customer had on-prem workloads that could span public and on-prem clouds, if you will, with federated security and a data plane that spanned the states. Essentially, you had an internal and an external cloud with a single point of control. This is basically what the hybrid cloud vision has become, an abstraction layer that spans on-prem and public clouds. And we can extend that across clouds and out to the edge where a customer has a single point of control and federated governance and security. Now we know this is still aspirational, but we're now seeing vendor offerings that put forth this promise and a roadmap to get there from different points of view that we're going to talk about today. The NIST definition now reads cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources, e.g. network server storage applications and services that can be rapidly provisioned and released with minimal management effort or service provider interaction. So there you have it, that is inclusive of on-prem. But it took the industry a decade plus to actually get where we are today. And they did so by essentially going to school with the public cloud offerings. Now in 2018, AWS announced outposts, and that was another wake-up call to the on-prem community. Externally, they pointed to the validation that hybrid cloud was real. Hey, AWS is doing it, so clearly they've capitulated. But most on-prem vendors at the time didn't have a coherent offering for hybrid. But the point is the on-prem vendors responded as they saw AWS moving past the demilitarized zone into enemy lines. And here's what the competitive landscape of hybrid offerings looks like today. All three US-based hyperscalers have an offering or multiple offerings in various forms, outposts from Amazon and other services that they offer, Google Anthos and Azure Arc, they're also prominent, but the real action today is coming from the on-prem vendors. Every major company has an offering. Now, most of these stemmed from services-led and finance-led initiatives, but they're evolving to true as-a-service models. HPE GreenLake is prominent, and the company's CEO, Antonio Neri, is putting the whole company behind as-a-service. HPE claims to be the first, it uses that in its marketing with such an as-a-service offering, but actually Oracle was there first with clouded customer. You know, possibly Microsoft could make a claim to being early as well, but it really doesn't matter. Let's see, Dell has responded with Apex and is going hard after this opportunity. Cisco has Cisco Plus and Lenovo has TrueScale. IBM also has a long services and finance-led history and has announced pockets of as-a-service in areas like storage. At Pure Storage is an example that we chose of a segment player, of course within storage, that has a strong as-a-service offering and there are others like that. So the landscape is getting very busy. Well, so let's break this down a bit. AWS is bringing its programmable infrastructure model and its own hardware to what it calls the edge and it looks at on-prem data centers as just another edge node. So that's how they're depositioning the on-prem crowd. But the fact is when you really look at what outposts can do today, it's limited. But AWS will move quickly, so expect a continued rapid evolution of their model and the services that are supported on outposts. Azure gets its hardware from partners and has relationships with virtually everyone that matters. Anthos is as well a software layer and Google created Kubernetes as the great equalizer in cloud and it was a nice open-source gift to the industry and has obviously taken off. So the cloud guys have the advantage of owning a cloud. The pure on-prem players, they don't. But the on-prem cloud or crowd has rich stacks, much richer and more mature in a lot of areas as it relates to supporting on-premises workloads and much more so than the cloud players. But they don't have mature cloud stacks. They're kind of just getting started with things like subscription billing and API-based microservices offerings. They got to figure out Salesforce compensation and just the overall as-a-service mentality versus the historical product box mentality and that takes time. And they're each coming at this from their respective different points of view and points of strength. HPE is doing a very good job of marketing and go-to-market. It probably has the cleanest model enabled by the company split from HP. But it has some gaps that it's needed to fill and it's doing so through acquisitions. Esmeral, for example, is a new data play. It just bought Zerto to facilitate backup as a service and it's expanded partnerships to fill gaps in the portfolio, some partnerships which they couldn't do before because it created conflicts inside of HPE or HP. Dell is all about the portfolio, the breadth of the portfolio, the go-to-market prowess and its supply chain advantage. It's very serious about as-a-service with Apex and it's driving hard to win that day. Cisco comes at this from a huge portfolio and of course a point of strength and networking which maybe is a bit tougher to offer as-a-service but Cisco has a large and fast-growing subscription business in collaboration, security and other areas so it's cloud-like in that regard. And Oracle of course has a huge advantage of an extremely rich functional stack and it owns a cloud which has dramatically improved in the past few years but Oracle is narrow to the red stack, at least today. Oracle if it wanted to we think could dominate the database cloud, it could be the database cloud especially if it decided to open its cloud to competitive database offerings and run them in the Oracle cloud. Wonder if Oracle will ever move in that direction. Now, a big part of this shift is the appeal of Apex versus CapEx. Let's take a look at some ETR data that digs a bit deeper into this topic. This data is from an August ETR drill down asking CIOs and IT buyers how their budgets are split between Apex and CapEx. The midpoint of the yellow line shows where we are today, 57% Apex expecting to grow to 63% one year from now. That's not a huge difference. There's not a huge difference when you drill into Global 2000 which kind of surprised me. I thought Global 2000 would be heavier CapEx but they seem to be accelerating the shift to Apex slightly faster than the overall base but it's not really in a meaningful way so I didn't really discern big differences there. Now, when you dig further into industries and look at subscription versus consumption models for Apex, you see about 60, 40 favoring subscription models with most industry slowly moving toward consumption or usage based models over time. There are a couple outliers but generally speaking that's the trend. What's perhaps more interesting is when you drill into subscription versus usage based models by product area and that's what this chart shows. It shows by tech segment, the percent subscription that's the blue versus consumption or usage base that's the gray bars. Yellow being indifferent or maybe it's I don't know. What stands out are two areas that are more usage heavy, consumption heavy. That's database, data warehousing and IAS. So database is surely weighted by companies like Snowflake and offerings like Redshift and other cloud databases from Azure and Google and other managed services but the IAS piece, well not surprising is we think relevant because most of the legacy vendor as a service offerings are borrowing from a SaaS oriented subscription model with a hardware twist. In other words, as a customer you're committing to a term and a minimum spend over the life of that term. You're locked in for a year or three years, whatever it is. To account for the hardware and headroom the vendor has to install because they want to allow you to access to increase your usage. So that's the usage based model. See, you're then paying by the drink for that consumption above that minimum threshold. So it's a hybrid subscription consumption model which is actually quite interesting. And we've been saying what would really be cool is if one of the on-prem penguins on the iceberg would actually jump in and offer a true consumption model right out of the box as a disruptive move to the industry and to the cloud players and take that risk. And I think that might happen once they feel comfortable with the financial model and they have nailed the product market fit. But right now the model is what it is. And even AWS without post requires a threshold and a minimum commitment. So we'd love to see someone take that chance and offer true cloud consumption pricing to facilitate more experimentation and lower risk for the customer entry points. Now let's take a look at some of these players and see what kind of spending momentum they have. This is our popular XY chart view at Plotsnet score or spending velocity on the X axis and market share pervasiveness in the data set on the, on the, sorry, net score or spending momentum on the Y axis and pervasiveness or market share on the X axis. Now this is cut by cloud computing vendors as defined by the customer's responding. There were nearly 1500 respondents in the ETR survey. So a couple of points here. Note the red line is the elevated line. In other words, anything above that is considered really robust momentum. And no surprise, Azure, AWS and Google are above that line. Azure and AWS always battle it out for top share of voice in the, in the X axis in this survey. Now this remembers the July survey, but ETR, they gave me a sneak peek at the October results that they're going to be releasing in the coming week. And Dell cloud and VMware cloud, which is VCF and maybe some other components, not VMware cloud and AWS, that's a separate piece. But those two are moving up in the Y axis. So they're demonstrating spending momentum. IBM is moving down and Oracle is at a respectable 20% on the Y axis. Now interestingly, HPE and Lenovo don't show up in the cloud taxonomy and that cloud cut and neither does Cisco. I believe I'm correct in that this is an open-ended question, i.e. who are your cloud suppliers. So the customers are not resonating with that messaging yet, but I'm going to double check on that. Now to widen the aperture a bit, we said let's do a cut of the on-prem in cloud players within cloud accounts so we can include HPE and Cisco and see how they're doing inside of cloud accounts. So that's what this chart does. It's a filter on 975 customers who identify themselves as cloud accounts. So here we're able to add in Cisco and HPE. Now Lenovo still doesn't show up in the data. It shows up in laptops and desktops but not as prominent in the enterprise, not prominent at all. But HPE Esmeral did show up and it's moving forward in the October survey. Again, part of the sneak peek. Esmeral is HPE's data platform that they've introduced combining the assets of MapR, blue data and some other organic development. Now as you can see HPE and Cisco, they show up on the chart as I said and you can see the rope in the tug of war is starting to get a little bit more taught. The cloud guys have momentum and big account presence but the on-prem folks also have big footprints, rich stacks and many have strong services arms and a lot of customer affinity. So let's wrap with some comments about how this will shake out and what some of the markers we can watch. Now the first thing I'll say is we're starting to hear the right language come out of the vendor community. The idea that they're investing in a layer to abstract the underlying complexity of the clouds and on-prem infrastructure and turning the world into essentially a programmable interface to resources. The question is what about giving access through that layer to underlying primitives in the public cloud? VMware has been very clear on this. They will facilitate that access. Believe Red Hat as well. So watch to the degree in which the large on-prem players are enabling that access for developers. We believe this is the right direction overall but it's also very hard and it's going to require lots of resources in R&D. I would say at this point that each company has its respective strengths and weaknesses. I see HPE mostly focused today on making its on-prem offerings work like a cloud whereas some of the others, VMware, Dell and Cisco are stressing to a greater degree in my view enabling multi-cloud and edge connections, cross connections. Not that HPE isn't open to that when you ask them about it but its marketing is more on-prem leaning in my opinion. Now all of the traditional vendors in my view are still defensive about the cloud although I would say much less so each day. Increasingly, they look at the public cloud as an opportunity to build value on top of that abstraction layer if you will. As I said earlier, these on-prem guys they all have a ways to go. They're in the early stages of figuring out what a cloud operating model looks like, how it works, what services to offer, how to pay sellers and partners but the public cloud vendors, they're miles ahead in that regard but at the same time they're navigating into on-prem territory and they're very immature in most cases. So how do they service all this stuff? How do they establish partnerships and so forth and how do they build stacks on-prem that are as rich as they are in the cloud and what's their motivation to do that? Are they getting pulled with digging their heels in or are they really serious about it? Now in some respects Oracle is in the best position here in terms of hybrid maturity but again it's narrowly focused on the red stack. I would say the same for pure storage, more mature in as a service but narrowly focused of course on storage. Let's talk marketplace and ecosystems. One of the hallmarks of public clouds is optionality of tooling. Just all you do is go to the AWS marketplace and you'll see what I mean. It's got this endless bevy of choices. It's got one of everything in there and you can buy directly from your AWS console. So watch how the hybrid cloud plays out in terms of partner inclusion and ease of doing business. It's another sign of maturity. Let's talk developers and edge. This is by far the most important and biggest hole in the hybrid portfolios outside the public cloud players. If you're going to build infrastructure as code, who do you expect to code it? How are the on-prem players cultivating developer communities? IBM paid 34 billion to buy its way in. Actually in today's valuation terms you might say that's looking like a good play but still that cash outlay is equal to one third of IBM's revenue. So big, big bet on open shift. But IBM's infrastructure strategy is fragmented and it's cloud business as IBM reports in its financial statements is a services heavy kitchen sink set of offerings. It's very confusing. So they got to still do some cleanup there but they're serious about the architectural battle for hybrid cloud as Arvind Krishna calls it. Now VMware by cobbling together the misfit developer toys of the remnants from the EMC Federation including Pivotal is trying to get there. But when you talk to customers they're still not all in on VMware's developer affinity. Now Cisco has DevNet but that's basically CCIEs and other trained networking engineers learning to code in languages like Python. And it's not necessarily true devs although they're upskilling. It's a start and they're investing Cisco that is investing in a community leveraging their champions and I would say Dell could do the same with for example the numerous EMC storage admins that are out there. Now Oracle bought Sun to get Java and that's a large community of developers but even so when you compare AWS and Microsoft ecosystem to the others it's not even close in terms of developer affinity. So lots of work to be done there. One of the point is pure acquisition of Portworx again while narrowly focused is a good move and instructive of the changes going on in infrastructure. Now how does this all relate to the edge? Well I'm not going to talk much about that today but suffice to say developers in our view will win the edge and right now they're coding in the cloud. Now they're often coding in the cloud and moving work on prem, you know wrapping them in containers. But watch how sticky that model is for the respective players. The other thing to watch is cadence of offerings. Another hallmark of cloud is a rapid expansion of features. The public cloud players don't appear to be slowing down and the on-prem folks seem to be accelerating. I've been watching HPE and GreenLake and their cadence of offerings and watch how quickly the newbies of as-a-service can add functionality, I have no doubt. Dell is going to be right there as well as is Cisco and others. Also pay attention to financial metrics. Watch how as-a-service impacts the income statements and how the companies deal with that. Because as you shift to deferred revenue models, it's going to hurt profitability. And I'm not worried about that at all because it won't hurt cash flow. Or at least it shouldn't. As long as the companies communicate to the Wall Street and they're transparent, i.e. they don't shift reporting definitions every year and a half or two years, but watch for metrics around retention and churn RPO or remaining performance obligations, billing versus bookings, increased average contract values, cohort selling, the impact on both gross margin and operating margin. These are the things you watch with SaaS companies and essentially these big hardware players are becoming as-a-service slash SaaS companies. These are going to be the key indicators of success and the proof in the pudding of the transition to as-a-service. This should be positive for these companies assuming they get the product market fit right and can create a flywheel effect with their respective ecosystems and partner channels. Now I'm sure you can think of other important factors to watch, but I'm going to leave it here for now. Remember these episodes, they're all available as podcasts wherever you listen. All you got to do is search breaking analysis podcast and please subscribe. Check out ETR's website at ETR.plus. We also publish a full report every week on wikibon.com and siliconangle.com. You can get in touch with me with email david.volante at siliconangle.com You can DM me at dvolante. You can comment on our LinkedIn posts. This is Dave Vellante for theCUBE Insights powered by ETR. Have a great week, everybody. Stay safe, be well, and we'll see you next time.