April 9, 2020 — OWOX BI Pipeline. New pipeline Google BigQuery → Yandex.Audience available in open beta

Create automatically updated audience segments based on any type of user data you store in Google BigQuery and upload them to Yandex.Audience. You can use these segments in various Yandex services.

The new pipeline is in open beta now and is free for all OWOX BI subscription plans, including the "Start" plan.

How does it work

A Google BigQuery → Yandex.Audience pipeline automatically uploads audience segment data from your Google BigQuery project. The segment data must be united with an SQL query.

The data is uploaded directly to Yandex.Audience and is updated automatically once per day. You don't need to gather the data to a .csv table and upload it manually via the Yandex.Audiences interface.

What you need to set up the pipeline

Before setting up the pipeline, make sure you have this:

1. The data you want to unite into an audience segment. It must be stored in a single Google BigQuery project.

2. An SQL query to unite these data. You'll need to paste this query in the pipeline settings after you've created the pipeline. The query result must provide a list of Client IDs based on which the segment will be formed.

Before setting up the pipeline, test the query running it in BigQuery. If it works there, it will work for the pipeline.

Note:A Yandex.Audience segment must contain at least 1,000 rows. If the query run result contains fewer than that, OWOX BI won't be able to upload the segment to Yandex.Audiences.

3. If you want to build a segment based on Yandex.Metrica Client IDs, you'll need to provide your Yandex.Metrica tracking ID. Read in Yandex help center where to find your tracking ID.

Note:If your OWOX BI project is subscribed to "Reports&Attribution" or "Enterprise" plan, the OWOX BI team can help you build segments to upload. Just write to us at bi@owox.com or in the support chat.

How to set up the pipeline

1. On the OWOX BI main page, click Create pipeline:Create_pipeline_blank_en.png

2. Select the data source — Google BigQuery:Create_pipeline_BQ-YA_1_en.png

3. Select where you need to upload your data — Yandex.Audience:Create_pipeline_BQ-YA_2_en.png

4. Provide access to Google BigQuery. You need to use the account that has access to a Google BigQuery project that contains the data you need:Create_pipeline_BQ-YA_3_en.png

5. Select the Google BigQuerу (Google Cloud Platform) project that contains data you want to create a segment based on:Create_pipeline_BQ-YA_4_en.png

6. Provide access to the Yandex.Audience account you want to upload the segment to:Create_pipeline_BQ-YA_5_en.png

7. Enter the name that will be used for your segment in Yandex.Audience and other Yandex services:Create_pipeline_BQ-YA_6_en.png

Select the Data type for your audience identifiers to build the segment based on. If you collect Yandex.Metrica Client IDs into your user behavior tables, we recommend selecting the Yandex.Metrica Client ID data type here:Create_pipeline_BQ-YA_7_en.png

If you select Yandex.Metrica Client ID, you also need to enter your Yandex.Metrica counter ID:Create_pipeline_BQ-YA_8_en.png

It's your tracker ID from Yandex.Metrica. Here's the guide on how to find it.

8. Click Continue. You'll be directed to the created pipeline's page. Note that the pipeline has the Blocked status now.

9. To activate the pipeline, click on the Query tab. There, paste past your SQL query that will form the segment from your data stored in BigQuery:Create_pipeline_BQ-YA_9_en.png

The result of this query should provide you with a list of identifiers of the type you've chosen as Data type during the pipeline creation (it may be Client IDs, phone numbers, emails etc.).

Before you paste the query in the Query tab, run it in Google BigQuery to check if it works.

Note:A Yandex.Audience segment must contain at least 1,000 rows. If the query run result contains fewer than that, OWOX BI won't be able to upload the segment to Yandex.Audiences.

10. Once you've pasted the query, click the Activate button next to the pipeline status. In the opened dialog, click Query is ready:Create_pipeline_BQ-YA_11_en.png
If the query is written correctly, the pipeline will change the status to Running. This means that OWOX BI has begun building the segment. Once the segment is built, it will be sent to Yandex.Audience. Yandex.Audience will process the segment. This usually takes up to 2 hours.

Once the segment is processed, the details of the uploaded segment will appear in the History tab on the pipeline page:Create_pipeline_BQ-YA_10_en.png

In this table:

  • The Processing Yandex.Audience segment status indicates that the segment is still being processed. While it's processed, it's unavailable in Yandex.Audience.
  • The Upload complete status means that the segment is successfully uploaded to and processed by Yandex.Audience. You can check it in the Yandex.Audience interface.
  • Last modified is the time when the segment was uploaded.
  • Data processed is the segment's data size.
  • Rows received is the number of records in the segment.
  • Unique rows validated is the number of records Yandex.Audience validated as unique.
  • Rows match is the number of records in the segment that were successfully processed by Yandex.Audience. In the screenshot above, one of the uploaded segments contains 27,562 unique Client IDs, 27,101 of them are processed successfully. This means that your ads can be shown to 27,101 of 27,562 Client IDs in the segment.

The pipeline uploads new segment data on a daily basis. Note that each time the segment is updated, the data is being reuploaded from scratch. As a result, the segment may be unavailable for up to 2 hours while its data Is being updated.

If you delete the Yandex.Audiences segment, OWOX BI will automatically create a new segment within 24 hours and upload relevant data from the BigQuery source into it. The new segment will have the same name as the old one but will have another ID. To continue using the segment’s data, you’ll need to replace the old segment with the new one in your ad management service.

Note that this pipeline is currently in free open beta. Its functionalities are subject to change in the release version. We'll be glad to hear from you any feedback on the pipeline's work and stability. If you have anything to share, please write to us at bi@owox.com.

Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request


Please sign in to leave a comment.