Lots of people are blogging about open source. I am a bit hesitant to add anything to the fray as I think most of the really important points have been made. One line of logic that hasn’t gotten as much press, in my opinion, is the theory of complements.
For those of you with a background in economics, you’ll remember that theory of complements talks about the relationship between complementary goods such as razors and razor blades, operating systems and desktop productivity applications, and bread and butter. General theory teaches us that when we are selling in a complementary goods situation, we want the price of our complement to be as low as possible (and they likewise want the price of our good to be as low as possible. How does this relate to software?
Well, this has a lot to do with software. Most of software is a complex ecosystem of complementary goods. N-tier applications — that’s an example of complementary goods. ERP systems and databases? That’s an example of a complementary goods relationship. Unix on Solaris? There are lots of other examples, too. I really do like Jonathan Schwart’s piece on how Sun monetizes Java as it talks to some of these (but not all of these) points.
So much of the discussion about open source has been about how JBoss will sink WebLogic and WebSphere, how MySQL will take on Oralce and IBM DB2, and how Linux will eventually vanquish Unix. Sure, some of this might happen. But what is even more interesting is what these developments mean for complementary goods relationships. Say you are SAP. Because your customers need a very reliable database tier to make your applications run, you have partnered with Oracle and IBM to provide that functionality. On some level, you recognize that’s revenue you could capture if you could provide a relibale AND supported database tier. Alas, before MySQL, you had a lot of open source options, none of which had commercial or developer support communities. But now, if MySQL can become what it looks like it might grow to be, partnering with a low-cost, supported, reliable database means that you don’t need to rely on an expensive database tier anymore. The amount of value on the table (which arguably hasn’t changed — substituting vendors on the database tier doesn’t really change the amount of value that the customer captures) hasn’t changed and you are now positioned to get more of it.
If you look at the software world from this lens, it has some important implications. First, those companies who only play in one segment of software are the most vulnerable. For example, if you are only a database or application server company, your complementary goods partners might look to cut you out of the picture by partnering with open source alternatives. Second, if you are providing a particularly valuable component (such as ERP or business intelligence) in the IT infrastructure, your eyes should be lighting up with the prospect of capturing more value as you have the option to support and integrate lower-cost complementary goods.
Is it any wonder that SAP and IBM have been so quick to reposition themselves as friendly toward the world of open source (not all parts, mind you, but many)? In this mythical (and often mentioned) world where all software can be acquired from an established open source community, there is one element that cannot be commoditized; professional services. That leaves IBM in a great position to become the de facto integrator of open source solutions. Leaving aside the tremendous economies of scale that accrue to IBM in becoming the master of all open source integration (very few corporations can employ a staff to stay on top of how all of these packages interoperate). As long as ERP continues to be a complex, expensive, and mission-critical application, open source ERP applications will have a challenge in unseating SAP.
Comments? Email me at blog @ charleshudson.net