According to a Street publication, Google’s cloud isn’t really gaining market share. That’s obviously not a good thing, and the truth is that Google’s cloud business has been consistently ranked fourth in the space, behind even IBM. I wondered early in January why Google wasn’t gaining, given that frankly it may well have the best cloud technology and infrastructure of the whole public cloud space. That raised the broader question of just what enterprises thought about public cloud, and what public cloud providers think about their customers, so I’ve been gathering views. This blog will present them.
First, though, I want to make a point, which is that enterprises make up less than half of the public cloud market. The majority is made up of OTTs and startups, and Amazon’s cloud business is likely Number One because they’ve gotten the largest share of this bigger chunk of opportunity. I didn’t try to survey this group because 1) I don’t have the breadth of contacts, 2) I’d have NDA issues with clients, and 3) I don’t think I’d get a truthful response in many cases. Suffice it to say that this group of cloud users tend to fear Google and Microsoft as direct competitors, so they gravitate to Amazon.
Enterprises have their own emotional reactions to the cloud players, and I can get a taste of their views for Amazon, Google, and Microsoft. I’m less comfortable about the IBM cloud data; my contacts suggest that most of the IBM cloud business comes from relatively few larger customers, which makes accurate data difficult to come by. For the three, I think there’s enough comment to draw some insights.
The first and most obvious question is whether enterprises believe they know the offerings of those three cloud providers, and almost all of them say they do. Similarly, almost all the enterprises think that the services offered by the providers are “substantially the same”, and just short of three-quarters said that pricing differences among the providers was “minimal”. This would suggest that something other than features and price drove the decisions, or that enterprises weren’t representing their process accurately.
The next question might tell the tale. If you asked enterprises what cloud provider they engaged with first, the tally of responses by cloud provider match quite closely to the market shares of each. As a follow-up point, if you asked whether their the currently dominant provider in their business was the first provider they engaged with, just under 70% said that it was.
This, to me, suggests that cloud success is tied closely to cloud provider sales engagement with enterprises. That’s not surprising; incumbency is a powerful force. It does beg the question of what happened in the remaining 30-plus percent, though, and here is where some “issues” or “differences” emerge, and there were two highly related things that drove enterprises.
The first of these issues was hybrid cloud, named as a priority by over two-thirds of enterprises (some named more than one issue). It was particularly interesting to me that half this group said that hybrid cloud was an “emerging” issue for them, but at the same time indicated that the applications that created their focus on hybrid cloud were in fact options all along. That suggests to me that enterprises didn’t at first see the nature of their own future cloud usage.
That seems to be supported by the second priority, which was creating new front-end technologies for their applications to accommodate changing business needs. This got support by almost exactly the same percentage of users, and it was universally cited by the half of “hybrid cloud” users who said their applications of hybrid cloud had been options all along.
What cloud providers were seen as best in their support of one or both of these? Among the three, Microsoft was named by the largest number, over two-thirds of enterprises. That means that Microsoft’s Azure was seen as “top cloud” by two-thirds of the enterprises who changed their primary providers. Clearly, it’s this factor that has been driving Microsoft’s cloud growth. Amazon was second with about 25%, and Google was third with about 8%.
Google’s failure to gain market share, then, could be linked to the fact that it is not seen as the premier hybrid cloud player. The importance of hybrid/front-end cloud in wrestling enterprises away from their legacy cloud provider is likely the reason IBM decided to emphasize it in their own positioning. Both these points are validated by comments on how the cloud providers engaged those enterprises who did abandon their incumbent provider for another.
On the hybrid/front-end cloud issue, every enterprise who switched cloud providers at all said that Microsoft’s Azure sales team engaged them fully on the issue. A bit under half said that Amazon had, and the number who said that Google had was less than a sixth. Microsoft’s presentation was considered “highly useful” by almost all the enterprises, Amazon’s by a quarter, and Google’s (significantly) by only two enterprises, below the level of statistical significance.
In some interactive chats with these enterprises, they often noted that the initial contact they had with their cloud provider options (including the incumbent) when considering a change was unsatisfactory. The biggest complaints were things like “didn’t listen to my needs” or “gave me a canned pitch”. These all boiled down to being non-responsive, and the complaint was levied most often against Google, then Amazon, and then Microsoft. If we broaden the universe of enterprises to all cloud users I had data on, rather than just those who changed from their incumbent provider, the incumbent was less likely to be considered non-responsive (not a surprise), whoever they were. That may help explain why incumbency is so valuable; it teaches the provider to listen better.
Amazon’s biggest problem came from its channel partners. Enterprises liked them the least of all the channel partners, citing them as being less responsive to their needs. Microsoft’s channel partners got the best comments, which might mean Microsoft is doing a better job of qualifying them or training them.
Better, but still not great. None of the cloud providers are prepared to tell enterprises that most of their current applications won’t ever move to the cloud, or even suggest it could be a very long time before that happens. As a result, many enterprises don’t really see the hybrid/front-end strategy holistically, and that’s an important prerequisite for effective use of public clouds.
Hybrid cloud technology, if it is indeed the key to the cloud future, is both simple and complicated. On the simple side, the coupling between workflows in the cloud and in the data center is simple, with any number of options available. Enterprises could use any helpful set of cloud features and have little or no impact on how they link back to their data center applications. On the complicated side, if hybrid cloud is “multi-cloudized”, then it could be smarter to use portable platform software (from HPE, IBM/Red Hat, or VMware, for example) to build the cloud piece, which could then be run anywhere with minimal changes. This would also facilitate redeployment and scaling across cloud and data center boundaries.
What I hear from enterprises suggests that the cloud providers are trapped in the “move to the cloud myth” to some degree at least, and that their success with the enterprise relates to the extent to which they’re breaking out. Microsoft has an advantage here because they’re an enterprise company first and a cloud company second. IBM is in the same position, but is hampered strategically by the fact that their influence is limited to big mainframe accounts, and they’ve not effectively exploited Red Hat in broadening their position. In any event, Red Hat and OpenShift depend on being cloud-neutral for a big part of their potential success. IBM, then, should have embraced a cloud-neutral strategy to maximize hybrid cloud positioning, but that would then limit the benefit to their own cloud business.
It’s hard to say whether cloud providers are driving misconceptions about “moving to the cloud”, whether they’re victims of media hype as much as buyers are, or whether buyers still believe that’s where things are heading. Whichever is the case, it would seem that both enterprises and cloud providers would benefit from examining the current cloud world, and positioning based on what’s really happening.