Looker

Looker, a business intelligence and big-data analytics platform, enables you to explore, analyze, and share real-time business analytics seamlessly.

The Braze and Looker integration allows Braze users to leverage first-party Looker Blocks and Looker Actions user flagging via the REST API. These flagged users can be added to segments to target future Braze campaigns or Canvases. To use Looker with Braze, we recommend sending your Braze data to a data warehouse using Braze currents, then use Braze’s Looker Blocks to quickly model and visualize your Braze data in Looker.

Prerequisites

Limitations

• This process only works with data that has not been pivoted.
• The API processes a maximum of 100,000 rows at a time.
• The final count of a user’s flag may be lower due to duplicates or non-users.

Integration

Looker Blocks

Our Looker Blocks help Braze customers quickly access a view of granular data we offer via Currents. Our blocks provide pre-made visualizations and modeling for Currents data so Braze customers can easily implement analytic patterns like retention, evaluate message deliverability, take a more detailed look at user behavior, and more.

To implement the Looker Blocks, follow the instructions in the README files of the GitHub code.

• Message engagement analytics block README
• User behavior analytics block README

Both integrations assume that your initial Braze integration, as well as your Braze integration with a Looker-compatible data warehouse, is appropriately configured to capture and send necessary data.

Available Blocks

Looker Actions

Looker Actions allow you to flag users within Braze via the REST API endpoint from a Looker Look. Actions require that a dimension is tagged with braze_id. The Action will append the flagged value to the user’s looker_export custom attribute.

Step 1: Set up a Braze Looker action

Set up a Braze Looker Action with your Braze REST API key and REST endpoint.

Step 2: Set up Looker Develop

Within Looker Develop, select the appropriate views. Add braze_id to the dimensions tag and commit the changes. This braze_id tag is use to determine which field is the unique key.

1
2
3
4
5
6
dimension: external_id {
    type: string
    primary_key: yes
    sql: ${TABLE}.external_id ;;
    tags: ["braze_id"]
}

Make sure to commit the changes. Looker Action will only work with production settings.

Step 3: Set user attributes in tags

Optionally, any attribute can be set using a braze[] tag with the attribute name in the brackets. For example, if you wanted a custom attribute user_segment to be sent, the tag would be braze[user_segment].

Note the following limitations:

• Attributes will only be sent if it’s included as a field within the look.
• Supported types are Strings, Boolean, Numbers, and Dates.
• Attribute names are case-sensitive.
• Standard attributes can also be set as long as they match the standard user profile names exactly.
• The full tag should be formatted within quotes. For example, tags: ["braze[first_name]"]. Other tags can also be assigned but will be ignored.
• Additional information can be found on GitHub.

Step 4: Send the Looker action

1. Within a Look with a braze_id dimension selected, click the settings gear ( ) on the upper right, and select Send….
2. Select the custom Braze Action.
3. Under Unique Key, provide the primary user mapping key for the Braze account (external_id or braze_id).
4. Give the export a name. If none is provided, LOOKER_EXPORT will be used.
5. Under Advanced Options, select Results in Table or All Results and then Send.
   
   
   
   If the export was correctly sent, then LOOKER_EXPORT should appear in the user’s profile as a custom attribute with the value you entered in the action.
   
   

Example outgoing API

The following is an example of an outgoing API call, which will be sent to the /users/track/ endpoint.

Header

1
Authorization: Bearer [API_KEY]

Body

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
{
   "attributes" : [
      {
        "external_id" : "user_01",
        "_update_existing_only" : true,
        "looker_export" : { "add" : ["LOOKER"] }
      },
      {
        "external_id" : "user_02",
        "_update_existing_only" : true,
        "looker_export" : { "add" : ["LOOKER"] }
      },
      {
        "external_id" : "user_03",
        "_update_existing_only" : true,
        "looker_export" : { "add" : ["LOOKER"] }
      },
      .....
   ]
}

Segment users in Braze

In Braze, to create a segment of these flagged users, navigate to Segments under Engagement, name your segment, and select Looker_Export as the filter. Next, use the “includes value” option and provide the custom attribute flag you assigned in Looker.

Once saved, you can reference this segment during Canvas or campaign creation in the targeting users step.

Troubleshooting

If you’re having issues with the Looker Action, add a test user to internal groups and check that the following:

• The API key has the users.track permissions.
• The correct REST endpoint is entered, such as https://rest.iad-01.braze.com.
• A braze_id tag is set in the dimension view.
• Your query includes the Id dimension or attribute as a column.
• Looker results are not pivoted.
• The unique key is correctly selected. Usually the external_id.
• braze_id in the dimension is different then the braze_id in the api. braze_id in the dimension is use to indicated that it’s the id field for the Braze API. For most purposes, upon send external_id is the primary key.
• The external_id user exist in the Braze platform.
• The looker_export field is set as Automatically Detect under Braze Platform > Settings > Manage Settings > Custom Attributes.
• The changes are commit to production. Looker Action works on production settings.