Customer Tables In BigQuery

Connecting to a customer table in BigQuery allows automation of customers and their attributes into the audience platform providing for easy data integration with most CMS systems or other data practises.. 

To Get Started

Before setting up a connection in Thirty-One Circles you first need to organise your data into a table in your own BigQuery instance. The table will need the following columns and data types:

  • email_hash > “String”
  • phone_hash_f1 > “String”
  • phone_hash_f2 > “String”
  • consent > “Boolean”
  • attributes >”String” & a repeated field
 

If you would like help getting your customer data into BigQuery or in this format please reach out to our team who will be happy to assist. Note that email_hash, phone_hash_f1 and phone_hash_f2 all follow the format as described in our hashed data guide

Connecting To Thirty-One Circles

Once your table is set up, you can connect to it by adding a new “BigQuery Automated Attributes” data source and then grant access by adding the user and permissions highlighted in the connector and provide the table id.

The table ID should be in the form of “project_id.dataset_id.table_id” and once entered the access and the format of the table are validated by the system 

Connecting To Thirty-One Circles

Once the table is connected it may take up to 24 hours for the data to be available in the audience platform, you can then form audiences based on all the attributes uploaded using an “Attribute audience” 

Key Notes:

  • You may need to wait up to 24 hours for processing new data. 
  • Only rows with valid hashed data are used. Attributes should container letters, spaces and numbers only. 
  • If the users advertising consent is provided this will update the consent used in the audience platform. If you would prefer consent is updated from the on site tag, set consent in the table to be null and the latest value from other sources will be used (or if consent is not yet set will default to denied). 
  • Customer attributes defined in a BigQuery table should not be defined in any other file upload as these will be overwritten. 
  • To force remove an attribute from all users, make sure all rows no longer have that attribute and add a row to your table with the hashed_email as null and the attribute to remove which will trigger the removal process. This processes is only needed if no rows in the latest update contain that attribute.