Not really, usually you have some translation to ensure data from different systems is aligned/translated into a unified format. It’s expensive and cumbersome t

Not really, usually you have some translation to ensure data from different systems is aligned/translated into a unified format. It’s expensive and cumbersome to do this in every dependent downstream system over and over again. Especially if upstream systems change, you want to implement that change in one central place instead of all downstream systems. I think there’s a lot of value in the snowflake approach: put all your enterprise data in a single logical data layer, where you can easily share that data with different tenants. Transformation and unification can be handled by the data layer. Systems get customized a lot: sap installations have a lot of company specific fields and objects, same for most other enterprise systems. Having all those customizations effect a network of connected systems is hard to manage. Better to dump all data into an enterprise wide data layer, and then let every tenant take whatever it needs. Also will provide for more flexible access to the data (sorting, pagination, filtering) and better performance.