Let me take another try at this problem of why firms might need to commoditize their own markets.
It appears to me that the high value is found in solving hard problems. The way we solve problems is that we bring together a collection of component bits that were not previously integrated. As we do this we complain: the design should be more elegant, the parts more modular, guiding principles of the design less random, the coordination problems less tedious. These complaints are the symptoms of the hard high value problem.
As times passes the solution to the hard problem becomes better understood. Any numbers of processes drive this transition. Design knowledge emerges and becomes widely known. Workable modularity is distilled out of tangled spaghetti like systems. The existence of exchange/interaction bottlenecks encourages standards to emerge. Rituals that lower coordination costs become better known.
This shift in a given problem domain from hard or tangled problem solving toward easier more modular problem solving causes two complementary shifts. The modularity enables a substantial increase in the size of the market along with an associated increase in total value generated. At the same time it knocks down the garden walls around the firms that built highly integrated solutions. It wreaks havoc on their discriminatory pricing.
All this is consistent with the insight that discriminatory pricing is likely when a firm