Question

BigQuery Data Transfer Service for Google Ads Partitioned Table Joins

  • 6 October 2021
  • 0 replies
  • 9 views

Hello,

I’m utilizing BigQuery Data Transfer Service for Google Ads to pull Google Ads Data then bringing the tables into Views in Looker.

When using the transfer service you are given both views and partitioned tables (shown here) for each report in Google ads. My logic is to use the partitioned tables in Looker to cut down on compute power and costs. Problem is when doing joins I’m getting a lot of fanning out in the data (large totals for stats such as costs, clicks, etc.).

 

Joins are necessary as example you need to use the table Campaign to bring in the Campaign name for the table Campaign Basic Stats. Google provides sample queries, but they are using the views and just point to Querying Partition Tables docs that I’m scratching my head to translate into Looker.

 

I’ve come across these resources but still roadblocks

A Looker Block for Google Ads Transfer, which has a section About Partitioning, which points to this somewhat dated thread. (I’m about to just use the views like shown here).

 

Smart Partition Key and Partition Keys

 

Anyone have experience with these partitioned tables or have any advice?

 


0 replies

Be the first to reply!

Reply