Again I am participating in a project that must decide to use a Canonical Data Model or not. In theory this is generally a good idea as already described by Wolf's Enterprise Integration Patterns . Does this also hold for a SOA? Yes when we must believe Thomas Erl, in fact it is one of his SOA pattern ( Canonical Schema ) We all know the advantages * Less transformation needed (this only holds when you have NxM relations) * Data reuse between Services within the same Service Inventory What do others say? Steve Jones describes three ways in his blog to setup a CDM . In Common demoninator (1) the minimum set of data that can be used to effectively communicate between areas on a consistent basis. The goal here isn't that this should be used on 100% of occasions but that it represents 70-80% of the interactions. In a Superset (2) the goal is to capture a canonical form that represents 100% of the possible interactions between services. Thus if a service might need 25 fields of pr