What type of join is used in data blending?

A blend behaves like a left outer join from the primary data source, though it is not literally a join — each source is queried and aggregated separately, then the results are matched on the linking field.

Every row from the primary survives. Secondary values are attached where the linking field matches and come through null where it does not. Anything present only in the secondary is not shown.

The join type is fixed and cannot be changed. If you need an inner or full outer relationship, you need an actual join, which means the data has to be reachable from one connection.