Since all the information associated with the executed command is included as part of the produced event, there is no need to ask for more information about the source application. Consequently, there are fewer requests, reducing the load retrieved by the source application.

The relationship between the produced events and the requests retrieved once the event is emitted is 1:1 in the best-case scenarios. In other words, one request will produce one event, but this could be even worse depending on how many dependent systems need to ask for additional information when the event is retrieved.

To avoid this extra load, all downstream systems often have their own data storage in which the events information is persisted, as shown by the following diagram:

When using this approach, each one of the downstream systems only stores the data that is relevant to itself, and the rest of the information provided is ignored because it is useless for the system, and won't be used at all.