Matt points to an interesting article about using an open business model.

Not sure if the author, Henry Chesbrough mentioned that IBM made significantly more than $100m/yr in revenue as a result of its Linux investments, whether through increased hardware or services sales. And, we continued to drive revenue to AIX. This last point isn’t often covered, but the AIX business didn’t die.

Very few customers take a one-size-fits-all approach to their IT needs. So, yes, you may need AIX or Solaris or a System z for some of your applications. But you’ll also need some RHEL, SLES or Windows for other parts of your infrastructure.

Customers want choice. So when a new option in a given software category becomes available that saves money, time, is easier to use, etc., customers will pay attention. If IBM had tried to ignore Linux, customers would have satisfied their curiosity (initially) and desire for Linux with another vendor. At the same time, these customers wanted to know how Linux fits into their current infrastructure (i.e. technology integration & skills reuse). Having an answer which included AIX and other parts of IBM’s offerings helped. Far from being a hindrance (i.e. related products that needed to be protected at all costs), these related products gave IBM the ability to speak to customer needs for the given project and other projects with different needs.

The same thing is happening in the application server space. When JBoss started popping up on customer’s radars, the traditional app server vendors didn’t pay attention (maybe to the degree they should have – depending on who you ask). But, customers satisfied their needs none the less and JBoss downloads and revenue grew.

Lucky for IBM, JBoss started out by targeting customers from other application server vendors. That’s when we made the move to purchase Gluecode Software Co. (a key member of the Geronimo community) and throw our support behind the Apache Geronimo open source application server. We delivered a paid support offering for Geronimo. We also came out with WAS Community Edition (WAS CE), a free product built on Geronimo that developers, customers, and partners could use without support or with paid support.

At the end of the day, customers want choice. And while the WebSphere Application Server (WAS) portfolio is broad, it didn’t give customers a free, light-weight and open source offering, some of the things that were attracting attention to JBoss. We filled that portfolio gap and our revenue continues to surpass market rates. Ensuring a strong tie between these offerings and the rest of the WebSphere portfolio has been a tremendous selling point because customers have different needs for different projects and want to take advantage of the skills they have.

Do we see JBoss in more deals? Yep. Do we lose some? Sure, that’s the software business for you. Do we win a lot more than we would have if we didn’t have the Geronimo Support or WAS CE? Absolutely.

Do you have a point that summarizes this long post? Yeah, thanks for asking: Don’t view open source competition as something to ignore. Get into the ‘game’, give customers the choice they seek and explore how your higher-value offerings (the ones you think you need to protect) can be an asset to your business.

Should we have made the move to support Geronimo earlier? Yeah. But our WAS revenue was growing so fast and our sales teams weren’t running into open source competition, so maybe we became a little complacent. It happens. I have one possible explanation for why, but that’s for another post. BTW, please read my disclaimer here.

[The pic is from Flickr user jawcey]