Currently, connectors with predefined schemas still perform a historical sync just to fetch the schema. This is incredibly inefficient, sometimes resulting in the schema fetch taking multiple days, extracting billions of rows, and unnecessarily consuming API quota.

Please load the predefined schema and allow us to select or deselect tables and fields before running a historical sync. This will greatly improve connector performance and efficiency.

For example, see connectors like Google DV360, SA360, CM360, Amazon DSP, and The Trade Desk.