I’ve learned that when people ask me in a puzzled manner “How’s that work?” regarding open source I shouldn’t answer until I’ve let them reveal what aspect of the enterprise bewilders them. For example I had one SVP who’s key question was “How do you get along?” I had a guy who’s primary interest was how do we solve the distribution, or as he put it shipping, problem. After collecting a few dozen of these I’ve come to think it’s a little odd that people assume the greatest mystery about open source is the revealing secrets, or the ip rights issue, or the volunteerism.

The dynamics of open source at the level of firms is particularly interesting to me since it bleeds into another area I’m curious about. How do standards emerge, particularly industrial exchange standards?

The simple model for that stuff is that buyers and sellers rendezvous in markets and since standards make that easier there are network effects which will accelerate the adoption of a few standards. It is simpler if we all use the same weights and measures and it is safer drive on the smae side of the road.

Buyers, sellers, market makers (and their agents) that can consolidate enough power to push standards to emerge will do so. The realist will point out at this point those with that power will advocate choices that benefit them and possible disadvantage than the other players.

In the absence of extreme market power the player will find it advantagous to negotiate a standard and advocate it’s adoption by all parties. Both stages are key, in fact you really have to solve three problems. You have to find representatives of all parties that can bring the right talents to bear on the design problem. You have to muddle thru all the negotiation and coordination problems of getting the standard designed, implemented, and maintained. Finally you have have to solve the advocacy, adoption, distribution, customer support problems.

The good news is that we have a carrot and a stick to make this happen. The carrot is improved exchange efficency, in many cases exchanges become possible that were otherwise impossible. The stick is the fear that other players will abuse their market power to create standards that disadvantage us.

Open source provides a reasonably good framework for working on these problems. The Internet makes it a lot easier to find talent, in particular it gives you a huge sample space draw from and then and makes it easier for the talent to volunteer (no travel!). We have stumbled on some tricks for solving the coordination problem. Optimistic concurrency for example. The net also makes it easier to solve the propagation problem as do the open source licenses/pricing. Working with information goods in an age of vastly increasing communication makes all this a possible.

Any firm involved in any exchange need to think about this. For example if your one of a thousand firms processing phone bills you can a) build it yourself, b) buy it from a vendor, or c) join/create an open source project to do it. Which one is best isn’t obvious. There is a lot of risk in building it yourself. There is the danger of becoming locked into a vendor if you decide to buy it. Coordinating an open source project, or any standards setting exercise, is a huge pain.

If you think of the problem as a game with moves it gets even more interesting. Assume, for example, you have built the solution yourself. In that situation you might find it advantageous to open source it first. It could reduce your development costs. It could force vendors to lower their prices. It could help to assure the industry “does it your way” instead of some other way that would be costly to switch to.

If your a vendor of such software you might want to move toward a more open version of the software for different reasons. If your customers are afraid of lock in this addresses their pain point. If you competitor goes first you could be toast as they become the default answer to the problem. If your customers start improving the software you can capture the value they create.

While open source or standards are dominate strategies in lots of information markets they doesn’t win all the time because the problems (talent, coordination, adoption) are still hard. We are getting better at all three though. It’s only just getting started.