Geoffrey Moore asserts in his seminal book, Crossing the Chasm, that markets are segmented in terms of how rapidly customers adopt technologies. Early adopters eagerly jump on new offerings, while others wait for technologies to have plenty
That's certainly been the case in the Linux and open source world, and it's a problematic scenario. Most technology companies have to grow up to sell to those later market participants; grow up being defined as offering support, consulting, comprehensive marketing materials, partnerships, and so on. As many of us have found, growing up is painful and not much fun.
In the technology world, many companies are not able to grow up (or cross the chasm, as Moore puts it) because of the cultural and economic challenges the process entails.
I've been pondering Moore's theory for some time. I must confess that I find his concept of crossing the chasm so compelling that I adapted and extended it in my book on open source, Succeeding with Open Source. In the book, I address the impedance mismatch of a situation in which low-cost open source cannot economically support the sophisticated requirements that mainstream users expect, while mainstream users must find ways to locate and assess resources to create the complete product they require. To enable these organizations to perform that assessment process, the book describes the Open Source Maturity Model. This model assesses any given open source product to determine how mainstream-ready it is.
With the foregoing, you can understand how eager I was to hear Moore's presentation at the recent Open Source Business Conference in San Francisco. He did not disappoint. In his wide-ranging talk, he covered a topic dear to my heart: the role open source can play in helping a company execute its strategy.
According to Moore, a company's strategic differentiation is its core; while all the ancillary activities it must engage in to deliver core are context. Context is important, but provides no advantage to the company. Another way you might think about this is that core is what your customers think they're paying you for, while context is all the stuff they expect you to do at no extra cost.
The problem is, as a company matures, the balance between core and context begins to swing inexorably toward context. The company has more and more stuff it has to do, but the tasks bring no joy upon completion. Inevitably, the company is vulnerable to attack from new market entrants that have little context sucking up management time and corporate capital.
The solution to this is clear: Find a way to reduce or eliminate context. Moore outlines five steps to do so: centralize administration, standardize, reengineer, automate, and outsource. These steps, taken in sequence, allow a company to accomplish context tasks, while reducing investment of both time and money.
Where does open source fit into this? Moore proposed two initiatives to take advantage of open source's strengths and to apply them to reduce context.
Moore's first tip is to use open source as a pure cost reduction strategy. If you have to run IT systems as part of your context, figure out a way to do it as cheaply as possible. Open source costs less, so use as widely as you can.
Moore's second recommendation was a bit more complex, but might ultimately pay off even better. Since there is a lot of similar context in many companies (human resources, recruiting systems, lease tracking, etc.), why not band together with other companies to jointly build these applications on an open source basis, thereby reducing your individual spending?
This latter recommendation is being implemented by the Avalanche Technology Cooperative. While not a pure open source organization, members of the Cooperative can submit and withdraw applications and benefit from everyone's investment. It's worth taking a look at the cooperative.
According to Geoffrey Moore, open source is crossing the chasm. Every company should take a look at how they can use open source software, particularly for "context" applications or where cost is a significant issue. Give it some thought.