Import segments

Firebase provides tools to tailor a user's experience by targeting groups of users through Firebase services such as Remote Config, Cloud Messaging, and In-App Messaging. Using a linked BigQuery account, you can import segments you may have identified outside Firebase to create targeted experiences with Firebase services.

Set up imported segments

You can import data for your segments into Firebase using Google Cloud BigQuery. BigQuery provides several ways to load data, so you are free to choose what works best for your configuration.

Once the integration is enabled:

• Firebase creates a dataset in BigQuery that you own, but Firebase has read access to.
• Firebase periodically ingests the data, making your updated segments available in the Firebase console for targeting.
• Firebase only has read access to this data. Firebase keeps a copy of this data in its internal storage.
• Any data that is deleted from the BigQuery data set is also deleted from Firebase data storage.

Enable BigQuery import

1. Go to the BigQuery integration page in the Firebase console.
2. If you haven't previously set up BigQuery integration, follow the on-screen instructions to enable BigQuery. Firebase console">
3. Enable the Imported Segments toggle.

When you enable importing segments from BigQuery:

• Firebase automatically creates a new BigQuery data set named firebase_imported_segments. This dataset contains empty tables named SegmentMemberships and SegmentMetadata.
• The dataset ‘firebase_imported_segments’ is also be shared with a Firebase service account with the domain @gcp-sa-firebasesegmentation.iam.gserviceaccount.com.
• Firebase runs a job at least every 12 hours to read from this dataset, and may import more frequently than 12 hours.

Import data into BigQuery

You can use any supported mechanism to load your data into BigQuery to populate the SegmentMemberships and SegmentMetadata tables. The data must follow the schema described below:

[
  {
    "name": "instance_id",
    "type": "STRING"
  },
  {
    "name": "segment_labels",
    "type": "STRING",
    "mode": "REPEATED"
  },
  {
    "name": "update_time",
    "type": "TIMESTAMP"
   }
]

[
   {
      "name": "segment_label",
      "type": "STRING"
   },
   {
      "name": "display_name",
      "type": "STRING"
   }
]

Set up billing for BigQuery

If you are trying out the new feature for an app with very few installations, you may only need to set up the BigQuery sandbox.

However, if you are using this for a production app with many users, you must set up billing for BigQuery usage to pay for storage as well as the mechanism used to load data into BigQuery. You will not be charged for any reads initiated by Firebase.

Deactivate the integration

To deactivate this integration, go to the BigQuery integration page in the Firebase console and deactivate the Custom segments toggle.

Use imported segments

Once the data is ingested, it will be available in the Firebase console for targeting with services such as Remote Config or In-App Messaging. This works just like targeting with properties or Google Analytics audiences.

You can use "Imported segment(s)" as one of the targetable attributes and segments you imported will be available for selection. They also include an estimate of the number of app instances that belong to each segment.

An estimate of the number of instances that match the entire targeting criteria is also available. This is updated as you make any changes to the targeting criteria.

Use cases

There are a number of ways you might use imported segments to create targeted user experiences. This section outlines some common scenarios where you might wish to use this feature.

Send notifications to a group of users

Imagine you have an app that allows in-app purchases with a shopping cart. You might also use custom-built or third-party analytics solutions (ones not powered by Google Analytics) to collect various metrics associated with user behavior in your app. Using these metrics, you could identify a group of users who have added items to the cart, but not completed checkout.

Now imagine you want to use Firebase Cloud Messaging to send a notification to these users to remind them they have items in their cart. You can create a segment called "incomplete-checkout" and include these users, identified by their Firebase installation ID, and upload it to BigQuery to share with Firebase.

Once Firebase ingests this data, it is available in the Notifications composer where you can create a new notification campaign targeting "incomplete-checkout" to send a message nudging the users to complete checkout.

Configure an app for a subset of users

Suppose you use an in-house analytics solution that indicates some users are having trouble navigating the app. To help those users, you want to configure the app behavior for these users to include a short tutorial video.

You can incorporate Remote Config in your app and use a parameter, named something like "needs_help", in your app to conditionally show the tutorial video.

Using your analytics data, create a segment named "troubled-users" and include appropriate users, identified by Firebase installation ID. Then upload this segment and its members to BigQuery to share with Firebase.

Once Firebase ingests this data, it is made available in the Remote Config console as a targetable segment. You can then create a condition targeting "troubled-users" and set the "needs_help" parameter to true for this condition and false by default. Once this config is published, the app shows the tutorial video only to users in the "troubled-users" segment.

Follow user journeys across devices

Imagine you built a restaurant-review app using Firebase and Google Analytics. Using metrics collected, you find users often access the app from both a mobile device and a tablet. You also discover that your users prefer to write reviews on the tablet, while they may read reviews from any device.

Some users start writing a review on their phone and give up, possibly due to the smaller form factor. You decide to send a notification to such users on their tablets prompting them to finish their reviews.

To do this, you could set an internally-generated reviewerId as the UserId using Google Analytics for signed-in users and trigger an event to identify cancelled reviews. You can then export your app’s Google Analytics data to BigQuery.

By analyzing this data in BigQuery, you can identify the Firebase installation ID of tablets for users who didn’t finish writing a review on their phone. You can name this group "tablets-of-users-who-cancelled-on-phone" and upload the segment to BigQuery to share the list of members with Firebase.

Once Firebase ingests this data, it is available in the Notifications composer as a targetable segment. You can then create a new notifications campaign targeting "tablets-of-users-who-cancelled-on-phone" to send a message nudging these users to complete their review on their tablets.