For Whom the (IT) Bell Tolls

I recently read a post by Jeff Kaplan, a SaaS expert with THINKStrategies, about being Nick Carr for a day. It seems that some IT folks are still stewing over Nick Carr's assertion that IT doesn't matter, and they directed their venom at Jeff when he piled on the Carr bandwagon by demonstrating that SaaS is a natural evolution in the commoditization of IT. Personally, I believe that IT matters, but the question is "for whom does IT matter?" and its corollary "for whom the (IT) bell tolls?"

Lou Gerstner recently gave a presentation to the Cognos sales team where he asked that team the following questions (or something very similar, as I was not there):

"How many of you have purchased an electric motor in the past year?"

Two people raised their hand.

"How many of you have purchased any one of the following in the last year: a refrigerator, a car, a fan, a dishwasher, a washing machine or a dryer, a laptop, PC, or a vacuum cleaner"

Half the audience raised their hand. The point is obvious. You do not care about buying motors, you care about the functionality they deliver when integrated into an application. A faster, cheaper, quieter, lower power motor matters VERY much to the appliance vendor, as it potentially gives them competitive advantage. The customer sees these benefits as longer batter life, a quieter appliance, cheaper electric bills, and so on. But the motor doesn't matter by itself, at least not to the consumer.

IT, therefore, matters very much to technology application providers. The ability to consume better components into their solutions so that customers get a better result matters a great deal. The ability to create unique combinations that deliver a new innovation matters a great deal. IT certainly matters to Apple, and Microsoft, and IBM, and salesforce.com, and Red Hat. If IT does not begin to matter a lot less to the customers of these companies, then the executives of these companies all deserve their ultimate low stock value fate.

But what about the IT staffer that has so long labored to overcome the inefficiencies of her technology suppliers? If her company's basic value is not achieved via the intellectual capital contained in the application technology that she produces, "never send to know for whom the bell tolls; it tolls for thee."

Oracle Ponders Linux Software Appliances

In an interview this past weekend with the Financial Times, Oracle CEO Larry Ellison pondered the notion of Oracle providing Linux as an integrated component of its software offerings. It makes perfect sense. Not because the world needs another Linux vendor, but instead because customers need a simplified approach for consuming complex software applications such as those provided by Oracle. With a single decision to ship Linux, Larry Ellison could transform all of the Oracle software portfolio into Linux software appliances that eliminate the customer hassles of integration and mismatched maintenance streams, delivering the value of software as a service (SaaS) for on-premise application deployments. Oracle could become the leader in SaaS overnight through the software appliance concept.

Linux and open source offer software application providers the historic opportunity to transform their business from selling licenses to delivering integrated solutions -- software appliances. Before Linux and open source became mainstream, delivering an integrated solution meant complex OEM agreements and being beholden to the technical and economic agenda of multiple third party component providers. Linux and open source, on the other hand, offer the perfect OEM license - freedom of distribution. And the technology is flexible to accomodate the technical agenda of the application provider. And customers get hardware vendor choice because Linux supports industry standard hardware instead of proprietary architectures. Customers get an integrated solution with "one throat to choke" -- the hallmark of SaaS, but for on-premise application deployments.

Larry's comments are less about Linux and more about customers. Linux doesn't matter. The customer experience with software applications does. Simplicity is good.

Red Hat Pays Big for JBoss

In a brilliant move to own the hearts and wallet of their datacenter customers, Red Hat yesterday announced that it would pay up to $420M for JBoss, the market leader for open source Java. After spending just less than 2 years attempting to build their own Java implementation with JOnAS, Red Hat decided it was time to go buy the leadership position in this critical technology area. With the purchase, Red Hat moves one step closer to the stated strategic goal of "owning the stack" for all infrastructure components their datacenter customers utilize to deploy applications.

What is the next item that Red Hat will add to the shopping cart in this buying spree? Well, several weeks ago Red Hat announced that they were going to create, maintain, and support "certified stacks" for their customers. If they are not successful building these stacks, perhaps they will go buy SpikeSource, mirroring the behavior in the Java space. And, given that a LAMP certified stack also includes PHP and MySQL, maybe Red Hat's next buy will be Zend or MySQL, or maybe Enterprise DB?

In any case, the acquisition is great validation of the value of open source, and it is now more certain than ever that Red Hat is going to be an even bigger force in the datacenter world.

Microsoft Embraces Software Appliance Market

At Linuxworld this week, Microsoft announced that the upcoming release of Virtual Server 2005 is going to provide a platform for software appliances. Well, they didn't actually use the term "software appliance," but they did say Linux would be supported on their platform. The underlying message, however, is consistent with the Microsoft objective of having all applications run on Windows - "If your application only runs on Linux, you can now run it on Windows with zero porting, installation, or configuration hassles."

Running Linux on Microsoft is not interesting at all. Giving application developers the flexibility to do a single port to a single platform without excluding any part of the available market due to OS incompatibilities is VERY interesting. Microsoft's announcement further cements the argument for application providers to choose Linux and open source as their preferred platform because they can eliminate the expense and time to market penalties associated with multi-platform support without sacraficing any market share associated with legacy "Windows Only" IT shops. Embracing the software appliance concept, with Linux and open source as the system software platform, just got easier.

Most legacy software vendors support multiple releases of multiple platforms in order to offer their product to the widest possible market. These companies typically spend about 14% of revenue on R&D expense. Salesforce.com only spends about 7.5%. Why the difference? Salesforce.com only supports a single platform, so they eliminate an enormous amount of engineering, QA, and test associated with installers, multi-platform ports, configuration scripts, etc. Imagine the benefits to shareholders and customers if software companies were able to drive 6 points on revenue to the bottom line or to better R&D (i.e. "core" issues instead of "context" issues).

Microsoft may have just helped the rest of the software world become more efficient by embracing the software appliance concept. Linux and open source becomes the universal platform and Microsoft provides the bridge for the legacy Windows users to avoid being left behind.