Blog post

The Evolution of the Cloud Computing Market

By Tom Bittman | November 03, 2008 | 6 Comments


I’ve had a number of posts on the structure of the cloud computing market, (Can A Cloud Computing Provider Be Too Massive?, Is Google the Mainframe of Cloud Computing?, Partly or Mostly Cloudy?) and I’m getting more and more comfortable talking about three major phases of market evolution. Very simply, I think they look like this:

image Phase 1: Monolithic (Early). Early cloud computing services will be based on proprietary/internal architectures – islands of cloud services delivered by megaproviders. This is what Google, Salesforce and Microsoft look like today.

image Phase 2: Vertical Supply Chain (2+ Years). Over time, some cloud providers will leverage cloud services from other providers (for example, ISVs moving into SaaS on top of Microsoft’s Azure Services Platform, use of, use of Google App Engine). Still proprietary islands, but ecosystems starting to build.

imagePhase 3: Horizontal Federation (4+ Years). Smaller providers will federate horizontally to gain economies of scale (and efficient use of assets) – also, enterprises will leverage horizontal federation for peak capacity (overdraft protection, cloudbursting). There will be more choices at each layer of cloud computing, and standards will gain momentum.

Again, very similar to the development of the overall server market.

The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.

Leave a Comment


  • Interesting idea. Why do you think that the entrants in phase one will be limited to proprietary/internal architectures – islands of cloud services delivered by megaproviders. Seems like there’s a great opportunity for a number of providers right now with variations on what is being offered by the big providers.

    Also, why do you think this will take so long to play out. This market is moving quite quickly and four years might be a very long time. Are we already in phase 2?


  • Tom Bittman says:

    The point is the center of gravity of cloud, and how it shifts – not hard barriers in the market. Agree it will move quickly – but by talking with vendors and users constantly, two years between cycles seems pragmatic.

    In phase one, the center of gravity is in large, vertically-integrated, proprietary service providers. It is the exception that cloud providers use cloud providers in phase one – they tend to build their own stuff. The center of gravity in phase one is the highly vertical and integrated, proprietary stacks – vendors trying to create their own standards. Sure, already there are lots of small vendors trying to find a role in the market. Another way of saying it – in phase one, the hot startups will be those going after phase two. It isn’t be Google or Salesforce anymore.

    In phase two, the center of gravity shifts to variety (of QoS, focus on industries, regions, etc.), and ecosystems (management services on top of Amazon, MS, etc.). The new Salesforces in phase two wouldn’t consider building their own data centers. Sure, some of that happens now, but it is not the center of gravity. Considering Amazon just came out of beta last week, and Microsoft’s Azure won’t be “officially” launched until late 2009, saying that the center of gravity doesn’t shift beyond them for another year seems realistic to me. But I agree – it is and will move fast.

    I put phase three out at 4+ years because it requires vendors desiring to work together horizontally, and it requires standards. I didn’t say it, but I also believe this is where the service broker becomes the center of gravity, not the “ecosystem” provider. Service brokers will shift workloads from one provider to another without a user knowing. Again, not that there won’t be some of this before 4 years, but it will not be the center of gravity. I’m saying in only four years from now, the center of gravity will be vendors like Microsoft subcontracting data center servers from third parties on the fly to handle loads.


  • Tom

    Enjoying your blog and wondered if you’d seen and his notion of types of platforms (arguably cloud platforms)?

    just thougth you may find it interesting


  • So,

    Phase 1 – Cloudware.

    Phase 2 – Cloud apps.

    Phase 3 – Cloud middleware.

    if I can drastically over-simplify? This seems like a reasonable expected evolution. From general plumbing and infrastructure to value built on that to, “finally,” beginning to abstract that value further away from the plumbing and infrastructure and to reduce the risk of having all your eggs in one basket, so to speak.

  • Paul Miller says:


    that’s an interesting way to characterise the evolution in Cloud Computing, and one that I’ve cited in a post over at


  • Phase1: I agree that phase 1 is today made up of the bigger players; namely Amazon, Google, Microsoft (soon), (soon, not big enough yet). However, there are already many smaller players sprouting (e.g. RackSpace, etc.). Hence, before the market evolves to phase 2, there’s going to be a proliferation of more cloud infrastructure (albeit smaller than the big boys) or platform players, esp. cloud storage.

    Phase 2: this phase is already beginning to form. We have players like RightScale and CohesiveFT forming around the eco-system of Amazon EC2. This is kinda like smaller fish swarming around the larger whales.

    Phase 3: definitely going to happen within 2 years. Many of the smaller players in phase 1 are going to leverage on their eco-system partners in phase 2 to create federated cloud standards and frameworks to make phase 3 happen. This will be accelerated by larger enterprises crying out for private clouds.