Cloud Computing is not a Business Model

I’m at the Dagstuhl Seminar “Information Management in the Cloud” where I keynoted about cloud computing businesses models. Given that I’m hardly a cloud computing expert this may seem like a stretch, however, the organizers had asked me to talk about my open source experience and relate this to cloud computing. This perspective turned out to be surprisingly fruitful. By realizing that both open source and cloud computing are disruptive innovations that enable a new generation of business models, I believe I was able to draw reasonable conclusions on the future of cloud computing from the history of open source. I reason by analogy, and here are the main conclusions:

  1. Cloud computing, like open source, is not a business model in itself, but an enabler of business models
  2. Cloud computing is not a business model but a distribution (read: sales and marketing) strategy
  3. Cloud computing, like open source, will have a novel type of business model built solely from commodities (distributors and utility computing, respectively)
  4. Cloud computing, like open source, will have a novel type of business model using proprietary software (single-vendor/open core and single-source clouds, respectively)
  5. Truly new businesses built using cloud computing need to educate their customers, i.e. rapidly grow the market; while doing that it is a landgrab
  6. Cloud computing, like open source, will be commoditized over time, where a commoditization frontier drives an innovation frontier to keep expanding
  7. Open source and cloud computing work synergistically, helping each other, as examples like SugarCRM show

I expect 2. above to be most controversial. That’s because many cloud experts talk about cost of providing the cloud service first before they talk about customer value, implying that customer value is a consequence of cost. Which is obviously getting it backwards. The core cloud computing customer values of try-before-you-buy, pay-as-you-go, higher quality of service, etc. are enabled by novel technology, which can also come with a lower cost structure.

Cloud computing is a sales and distribution strategy because the fine-grain provision and releasing of resources and the matching fine-grain pricing schedule drive adoption of cloud services through the line-of-business rather than the IT department. Open source strategy, anyone?

The slides + notes from the talk are available as PDFs (slides, slides + notes). I recommend you read the slides + notes version.

7 Replies to “Cloud Computing is not a Business Model”

  1. Hi Dirk,
    This sounds very interesting. So cloud computing will fade into being a commodity unless it acquires community involvement? So how would a cloud computing company get a community? What would that mean?

    1. Hey Martin! Good to hear from you. What part of the talk are you referring to? The commoditization frontier? IMO, community involvement creates commoditization. In a strictly commercial setting, you only get those communities going, however, if commoditization is the desired goal, e.g. because you want to avoid that a dominant player is reaping all the benefits.
      The way I view it, there is an ever growing bottom-layer of commoditized services, trying to catch up with more innovative non-commoditized services on top. Commoditized services may have smaller margins than non-commoditized ones, but there will still be plenty of room for differentiation around SLAs etc. I wouldn’t exactly call that “fading away”!

  2. Hi Dirk,
    i have never seen the attempt to compare open source and cloud computing history before. However i read each of your analogies. While your analogies seem convincing to me i would like to challenge one thing.
    After trying to convince myself for years i am not sure anymore that open source is a real disruptive technology, nor am i that cloud computing will be one.
    I do admit and fully agree that both technologies really change the way in which the market constitutes today and business models will be shaped tomorrow.
    And i do also agree that both technologies tend to cannibalize a large chunk of their predecessors market over time. But do they drive their predecessors obsolete over time ?
    After some time of investigating the relation between open source and closed source i really doubt that. I think there will be a market for closed software vendors even in 50 years from now. This convinces me at least that open source isn’t a real disruptive technology.
    I also believe there will be a scenario in which cloud computing doesn’t match the requirements of a specific customer. For example i do hardly believe that the intelligence services will migrate their data analysis and anti terror investigation
    to the cloud anywhen in the future.
    Greets Frederik

    1. Hi Frederik, I don’t think you are making a case against open source or cloud computing being disruptive 🙂 If there remains a niche for closed source, so be it. I don’t think being disruptive means completely destroying the old paradigm, just making it mostly irrelevant. As to cloud computing, I agree that many companies will want to run their own private cloud, but so be it. It is a reflection of the trend to cloud computing, adapted to specific requirements. I’d still consider this disruptive. I don’t think that even an intelligence agency will have everything in-house; just not cost-effective. Maybe one of the subtle drivers of cloud computing hasn’t received enough attention yet: In-house IT is usually much worse than service provider companies. So I imagine an outside service managing an in-house cloud. Now, isn’t that disruptive cloud computing, even for intelligence agencies? 🙂 –Dirk

Leave a Reply