My thoughts on why there are so few ERP startups among thousands of new web 2.0 startups:
Young founders don't understand enterprise: Most startups are founded by young kids who are in college or right of out college. Problems faced by companies are not their problems. The best startups are built around problems that the founders have faced themselves.
An ERP is a very complex app: An ERP has so many entities that are tied up in numerous layers of relations to various other entities and the complexity of the interactions is very high. It takes months if not years to understand many of these interactions.
Accounting is hard: Apart from the hierarchical structure (that can neatly be modeled using the nested set model) of the chart of accounts, there are numerous reporting requirements, constraints, conditions and dimensions.
Older devs are happy: People who "can" build ERPs, the older devs, already have steady jobs and clients. Risk appetite is inversely proportional to age. It is highly unlikely they are going to risk their well paying jobs and build a new system from scratch
ERPs have long gestation period: The functionality is atleast an order of magnitude more (if not two) than a usual social app, hence the period of time the application will take to mature will be much longer
Established players have no incentive: Established ERP players are building new web based ERPs but they are too expensive. The are used to a particular cost structure and they will never compete with "startups" in building low cost ERPs because its against their DNA.
