July 30, 2003
Web services traction and lack thereof
There's a lot of noise being made about web services. Trouble is the model suggested by many tech folks is something most business people would never accept, let alone use.
One idea suggested is that web services will allow software to interact with different providers in more automated (intelligent) ways. The idea is your side of the equation could 'pick and choose' from many different providers.
For folks that think this is a good idea, get a clue. Businesses don't engage in relationships with their various suppliers and customers based soley on performance metrics.
More often than not these relationships are based on trust. Trust of an esoteric nature not even remotely capable of being modeled reliably in software. Trust established based not necessarily on the idea that it's the most cost effective or fast. Trust many times based on understanding whether the other party will fuck it up or not. That's right, doing business with people not because you know they're the best but because you know they won't screw it up, and take you down with them.
Tell me how that would get coded in software! The sort of data needed to do this is highly inflammatory. Imagine the train wreck it'd cause if your ERP system's data about whether or not suppliers are idiots got leaked? We all have this sort of database but it's not the sort of thing we ever dare codify into hard data. Doing so just raises too many risks. It's also the reason we have people involved in the process. People that understand the esoteric rules of the game and the various conditions on the ground.
Until web services take this sort of thinking into account they're going to have a tough time gaining traction in the larger business community.
