Google Ads Boilerplate
Only in Essential Plan and above
The Google Ads boilerplate provides pre-configured models and semantics for analyzing your Google Ads performance.
Prerequisites
- Importing Google Ads Data into GCP BigQuery. Follow the Google documentation: Google Ads transfer.
Steps to Use the Google Ads Boilerplate
Before continuing, please make sure you've successfully imported your data into GCP.
Click the Google Ads option in the Start with a Data Boilerplate section to add a Google Ads Boilerplate.
Ensure you have imported your Google Ads data into Bigquery for this step. It is recommended that you do this step in a separate browser tab.
You will be asked to provide Bigquery Data Viewer permissions to the Wren AI service account. To do this, you must go into your GCP project, select the three dots on the resource containing your imported Google Ads data, and select manage permissions.
Select the Add Principal button to grant new access. Paste the Wren AI Service Account Principal Address generated on your Wren AI Cloud Project page. In the assign roles section, please select BigQuery Data Viewer and save.
The Viewer Role permission is read-only access to Bigquery data and metadata. This allows Wren AI Cloud to see your tables and explore schema to generate the queries in Wren AI Cloud Console. We can will not modify or delete any of your GCP resources.
Once the Wren AI Service account has been added to your GCP project, please input your project ID. This can be found on the GCP Developer Console; click on the GCP project, and at the top, it will tell you the project ID.
Next, add your dataset ID. Access the BigQuery dataset ID in the GCP Developer Console and inside the desired project. In the resources section, find the dataset, click on the 3 dots next to the dataset name, and select copy ID.
When done, click continue.
Review Pre-filled Fields
The project modeling page displays all fields and their sub-values from the Google Ads Boilerplate. We've added a visual marker to indicate the boilerplate's origin, making it easier to differentiate between views and models.
Learn more about views