With all the attention of late that PaaS providers are getting, it seems a good time to reflect on the stark contrast between two types of SaaS provider: those who do their own infrastructure, and those who farm it out.

Everyone knows that salesforce.com is the grandaddy of SaaS vendors, and it has gone down the only path that was open to it when it was conceived, that of creating and hosting its own infrastructure. As the somewhat acidic FSJ commented;

Benioff, ironically, has built his business around a bloated, overly expensive, outdated business model, a model that comes straight out of the late Nineties — he’s running his own data center, and he’s using Sun servers and Oracle software. It’s like “Back to the Future.” Meanwhile the rest of the world has leapt ahead onto Intel architecture and Linux. For Benioff to survive into the era of the cloud he’ll have to rip up his entire architecture and rebuild it. Yeah. It’s like that. He’s stuck. And he knows it. He’s not doing cloud computing. He’s doing what we all already recognize was a precursor to the cloud.

Already the CEO of SaaS vendor Sonian Networks uses the term “Legacy SaaS” to refer to those player of old who actual do their own hosting and serving.

I think it’s too early to entirely discount the self-hosting strategy, but it does seem, with ubiquitous, scalable and economically priced PaaS solutions now available from a number of vendors, that proprietary infrastructure will go the way of greenscreens.

A post over on ZDNet that included this estimated price for hosting a CRM type app on BungeeConnect;

At $3.60 per month it would seem something of a no-brainer to avoid the hassle, scaling issues and non-core business removal of focus that self provided infrastructure would cause.

So, over to the readers to vote;[poll=7]

Ben Kepes

Ben Kepes is a technology evangelist, an investor, a commentator and a business adviser. Ben covers the convergence of technology, mobile, ubiquity and agility, all enabled by the Cloud. His areas of interest extend to enterprise software, software integration, financial/accounting software, platforms and infrastructure as well as articulating technology simply for everyday users.

  • I’ve enjoyed reading your posts over the past few months, very interesting topics.

    My response to this question would be….no. My own belief of what we’re seeing is a normal evolutionary cycle. I work with lots of ISVs with SaaS offerings, some provide their own hosting (which initially surprised me) and some go outside to “professional” hosting companies. The main reasons they give for their decision are cost and control. They feel as though they have more control over the experience they deliver to their customers as well as the overall cost of the service.

    I think they will evolve over time as well, especially as their customer base gets more sophisticated and pushes them to provide services they can’t afford to deliver on their own. Until then, their customers are happy, and they control their costs and experience better than they would get if they handed the keys to a hoster.

    As far as the PaaS argument goes, I get the value proposition, which is nothing to sneeze at, but where I’m still not sold is on the sales value to the ISV for the commitment they make to the PaaS vendor. If I’m making a technical and product commitment to a platform, I’d like some help selling it, and thus far I haven’t seen any PaaS vendor out there with a model that does that. Seems fair if I’m essentially locked into the platform.

    Just a wrinkle to the evolution argument.

  • I think Bob’s right and wrong here.

    To my mind there has always been a dichotomy in SaaS providers. They push the benefits but (sometimes) don’t always subscribe to it. How many SaaS providers use hosted email and apps for instance?

    The same is true of PaaS for SaaS. I think this is a natural evolution, if Phil @ ZDnet is right is hard to argue with the economic benefits of doing this.

    I guess the counter to this is ‘special sauce or customisation’ (which again is similar to the end user). Would you gake CAD saas? problably not, how about that speciality billing app that you had wrote? Same is true for SaaS on PaaS. Some things are commodities (whole software sub – industries and functions like processing power, back up, etc).

    To that end, I do agree that the initial approach to saas (1.0) is being superceded or even disrupted by this newer approach. Benioff will hate it because it both validates and obliviates his business model!

    Time will tell

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.