How Cloud Computing Is Slowing Cloud Adoption

Last Updated Sep 24, 2009 2:25 PM EDT

There's the cloud, and then there's the cloud. The first cloud everyone talked about was really software-as-a-service (SaaS), a method for delivering applications over the Internet (the cloud) more effectively and cheaply than traditional implementations installed behind corporate firewalls, as exemplified by the likes of Salesforce.com, Successfactors, NetSuite and many others.

Then along came this other cloud, the infrastructure that you could rent by the processor, by the gigabyte of storage, and by the month, and which would expand and contract dynamically according to your needs, which Amazon, Microsoft, IBM and many other vendors offer.

Both have come to be known as cloud computing, but as you can see, they're very different. You can, of course, run cloud-based software (let's call it SaaS) on a cloud infrastructure, but why should the cloud have a negative impact on SaaS adoption? To understand this, you have to go back to the reason most customers chose to run SaaS applications to begin with.

It's not because on-premise software is evil, as Salesforce.com CEO Marc Benioff enjoyed implying during his dog-and-pony shows, and it's not even because traditional software companies treat their customers extremely poorly in many cases (a state of affairs allowing former Forrester analyst Ray Wang to create a business around the idea of an enterprise software user's bill of rights). In the end, large corporate customers began using SaaS because it allowed them to offload maintenance and administration of critical business applications to an external entity.

As Mark Lewis, president of EMC's content management and archiving division, noted over coffee this morning, most enterprise IT shops are consumed with having to buy and configure new hardware, and deal with integration and security issues -- not with optimizing software delivery methods. But SaaS gives them a way of integrating new applications without having to add more hardware. With SaaS, "you're getting a very efficient infrastructure for running your applications," he said.

In other words, the main motivation wasn't, "Geez, let's get rid of SAP and go with these friendlier guys;" it was more like, "ok, if we have to run this application on top of SAP, why don't we just go ahead and let someone run it for us in the cloud."

Now, however, there's another option for enterprise IT, which is to run applications in the cloud but continuing to use the applications that have already been customized for your purposes and enjoy widespread adoption within the organization. As Lewis put it, cloud infrastructure "allows you to take your custom applications that are sticky within your organization and put them into a cloud environment."

This doesn't mean the end of SaaS by any means, but it does give enterprise IT administrators one less reason to start using SaaS. Cloud computing fulfills one of the main functions of SaaS, which is to allow customers to add applications without having to purchase, configure and maintain new hardware.

That might not have been a primary motivation for Microsoft to start offering Azure, its cloud infrastructure play, but I'm sure that staving off threats to its enterprise applications business went into its thinking.

SaaS vendors still give customers more flexibility of deployment because their employees can access applications and data remotely by plugging into the Internet, as well as a more logical, usage-based cost structure. Other, secondary advantages, like seamless and continuous upgrades, also come into play.
But the availability of cloud infrastructure gives enterprise IT managers another option when it comes to managing their IT resources, and that will come at the expense of SaaS vendors.

[Image source: maniskka via Flickr]
  • Michael Hickins

    Michael Hickins has written about technology and business for BNET, InformationWeek, InternetNews.com, eWEEK -- where he was executive editor from 2007-2008 -- The Curator, Pseudo.com, Multex Investor, Reuters, and Conde Nast's WWD.com. Hickins is the author of The Actual Adventures of Michael Missing, a collection of short stories published by Alfred A. Knopf in 1991. He also published Blomqvist, a picaresque novel set in 11th century Europe, in 2006. Hickins remains passionately interested in the intersections of business, technology, politics and culture, and endures a life-long obsession with baseball. He is married with two children and lives in Manhattan.