Google Analytics → Google BigQuery: Create a user behavior data streaming pipeline

The Google Analytics→Google BigQuery pipeline collects raw non-sampled data about user behavior on your website, both at hit and session level.

The collected hit data is being uploaded to BigQuery tables in real time. Session data gets to BigQuery within the following day.

Note:When you create a hit and session data pipeline, OWOX BI automatically creates a cost data pipeline from Google Analytics to Google BigQuery.

What you need before creating a pipeline

1. To use Google BigQuery and create a streaming pipeline in OWOX BI, you need a Google Cloud Platform project with activated billing. Read more about how to set up Google BigQuery in this article.
2. To create a user behavior data pipeline, you need a subscription to the plan Marketing Data, Reports & Attribution, or Enterprise in OWOX BI. Session cost attribution for this pipeline is available only in the subscription plans Reports & Attribution and Enterprise.

How to create a pipeline

To create a Google Analytics → Google BigQuery user behavior data pipeline, go to the Create a pipeline page and take the following steps:

1. Select data source: Google Analytics:

2. Select the data type to be uploaded to BigQuery: User behavior data:

3. Provide access to Google Analytics:

4. Select the Google Analytics property and view as the data source.

Here, specify the Google Analytics view you want to gather data to. You need to have the Edit permission for this Google Analytics property.

Important:To avoid data discrepancies between data collected to Google Analytics and data uploaded to Google BigQuery via OWOX BI, select a Google Analytics view without any filters applied.

Note that you can set up only one pipeline per property.

5. Provide access to the Google BigQuery account you want to upload Google Analytics data to:

6. Select a Google BigQuery dataset.

First, select the Google BigQuery project that contains the dataset you want to gather data to. Then, select the dataset.

If needed, you can create a new dataset right here.

You can set up only one OWOX BI pipeline for each BigQuery dataset. Creating several pipelines for one dataset may cause the hit data to be overwritten.

7. Integrate the tracking code on your website. 

The method of implementing the tracking code depends on your project. Read about the tracking with Google Tag Manager, analytics.js, Measurement Protocol, Tealium IQ, and AMP pages, Android and iOS apps tracking in the list of the tracking code implementation methods.

Find it on the created pipeline's page: Settings > JavaScript code for hit and session data collection:

Done. The Google Analytics Real-Time Data → Google BigQuery streaming pipeline is all set.

A dataset with hit data tables will appear in your BigQuery project in the first minute of the next hour. Session data will get to BigQuery in 12 hours after the beginning of the next day.  

Please note: In OWOX BI pipeline, the maximum hit payload size is 16 KB. The hits exceeding this limit won't be processed and sent to BigQuery.
In Google Analytics, the maximum hit payload size is 8 KB.