Moving up the Stack–VMware and the Software World

Summary – Another long post so here’s a summary. Infrastructure will become commoditized and because of this VMware is moving up the stack. Ecosystem partners involved with CloudFoundry are best to partner with multiple PaaS vendors rather than CloudFoundry alone.

Recently there was a fairly heated discussion on Twitter specifically about VMware’s intentions with its CloudFoundry offering and more generally how much infrastructure vendors are having to look further up the stack in order to offset any potential commoditization of core infrastructure services. VMware are a fantastic case study to look to with regards this. Since their inception back in 1998 they have been focused almost exclusively on infrastructure enabling software, the “bare bones” of technology. At least they were until recently. It is informative to look at the investments that VMware has made in recent years in products that sit higher up in the technology stack from Wikipedia;

  • August 2009, VMware announced the acquisition of SpringSource, a leader in enterprise and web application development and management
  • January 2010, VMware acquired Zimbra, an open-source collaboration software tool
  • August 2010, VMware announced its intention to acquire TriCipher an identity federation provider
  • April 2011, VMware acquired SlideRocket a startup which developed a SaaS application for building business presentations that are stored online
  • May 2011, VMware acquired Socialcast, a group workstream service
  • June 2011, VMware acquired Digital Fuel, IT Financial and Business Management SaaS Company

That’s a whole heap of primarily application offerings for a company that ostensibly plays deep down in infrastructure. Witness also the release of CloudFoundry, an OpenSource PaaS offering that arguably has very few synergies with VMware’s core revenue stream. So where is this going, and what does it mean more broadly for the industry?

Others have suggested that CloudFoundry is very much a case of VMware securing its existing revenue streams. With the utmost respect to my colleagues who see it this way, I disagree. I believe that VMware is strongly following the strategy we’ve seen from Amazon Web Services, one of squarely aiming further up the stack. While clearly VMware has a history and a significant revenue stream from its hypervisor level business, recent developments have indicated a strong move upwards for the company.

In coming to this analysis I also need to refer to VMware CEO Paul Mortiz’s comments in a recent analyst call. When asked about the threat that OpenStack might pose to VMware’s business, Moritz said that;

we continue to believe that our greater, nearer-term challenge will probably come from Microsoft

In part I believe that this is a result of a focus on the higher levels of the technology stack, rather than infrastructure itself.

So what does this mean specifically for VMware and its ecosystem?

In terms of software, VMware would seem to be taking an approach that sees them determine a selection of different application types and delivering them under the VMware brand. When VMware first announced the acquisition of SlideRocket I was wondering how this would play out – over two years they have made a series of seemingly disconnected acquisitions that sees them now able to offer email, presentation and collaboration software. But it’s fair to say there is no real clarity in how this is going to look. After the SlideRocket announcement I expected to see VMware quickly back that up by acquiring a complementary set of office productivity applications. They haven’t done so and that leads me to wonder about the strategy behind the acquisitions. Add to that the fact that the trend is very much towards building ecosystems with third party ISVs (witness salesforce’s App Exchange, the Apple AppStore, the AWS Marketplace etc) and I’m not convinced about VMware’s software intentions long term.

Which leads us to CloudFoundry. I’ve spent a lot of time over the past few months talking to CloudFoundry partners – some already launched and some in stealth mode. Many of these partners are building point solutions on top of CloudFoundry in the hope and expectation that CloudFoundry will create a significant market opportunity. Indeed the level of buy-in that CloudFoundry has achieved in its short life is exceptional – vendors like AppFog, IronFoundry and many others hitching their wagons to CloudFoundry’s horse. While it is true that a small company can potentially make massive inroads by partnering with a large vendor, there are also risks involved in this. Recently I was talking to a company that has plans to create a service that essentially offers plug and play components within CloudFoundry. I opined to them that while I loved the idea, it was one which was at pretty high risk of being consumed by functionality introduced by CloudFoundry themselves. The age old question of what is core functionality, and what should be provided by ISVs yet again raises its head.

I reached out to a couple of CloudFoundry partners for their take on both the risk and the rewards of partnering with VMware. Wendy White, VP Marketing at Tier3, the creator of IronFoundry was particularly diplomatic saying that;

I can say that VMware is an important partner of ours on many levels and we do think they have taken the right steps with open-sourcing Cloud Foundry. We are confident that we have a commercially differentiated offering

I also spoke with Michael Surkan, a vendor building Microsoft-centric abstractions on top of CloudFoundry. He was a little more cautious in his approach, referring more to the fact that it is his view that partners need to think of a strategy that seems them build abstractions on top of multiple PaaS platforms, as he said;

it makes no sense to tie yourself to any one horse

CloudFoundry is a truly exciting initiative – both because it plays at a higher level than infrastructure and also because of the Open Source aspects of the offering. It’s an awesome opportunity but ecosystem partners need to be realistic about potential moves that VMware may take in the future.

1 Comment

Leave a Reply