Skip to content

mParticle

mParticle’s customer data platform empowers you to do more with your data. Sophisticated marketers use mParticle to orchestrate data across their entire growth stack, enabling them to win in key customer journey moments.

The Braze and mParticle integration allows you to seamlessly control the flow of information between the two systems:

  • Sync mParticle audiences to Braze for Braze campaign and Canvas segmentation.
  • Share data across the two platforms. This can be done through the mParticle kit integration and the server-to-server integration.
  • Send Braze user interaction to mParticle through Currents, making it actionable across the entire growth stack.

Prerequisites

Integration

Audiences

Use Braze and mParticle’s partnership to configure your integration and import mParticle audiences directly into Braze for retargeting, creating a full loop of data from one system to another. Any integration you set up will count toward your account’s data point volume.

Forwarding Audiences

mParticle offers three ways to set cohort membership attributes, controlled by the “Send Segments As” configuration setting. Refer to the following sections for the processing of each option:

Single string attribute

mParticle will create a single custom attribute called SegmentMembership. The value of this attribute is a string of comma-separated mParticle audience IDs that match the user. These audience IDs can be found in the mParticle dashboard under Audiences.

For example, if an mParticle audience “Ibiza dreamers” has an audience ID of “11036”, you can segment these users with the filter SegmentMembership — matches regex — 11036.

While this is the default option in mParticle, most Braze users opt to use single array attributes for the filtering experience when creating segments in Braze.

mParticle segment membership

Single array attribute

mParticle creates a single custom array attribute in Braze for each user, called SegmentMembershipArray. The value of this attribute is an array of mParticle audience IDs that match the user.

For example, if a user is a member of three mParticle audiences with the audience IDs of “13053”, “13052”, and “13051”, you can segment for users who match one of those audience with the filter SegmentMembershipArray — includes value — 13051.

One attribute per segment

mParticle will create a boolean custom attribute for each audience a user belongs to. For example, if an mParticle audience is called “Possible Parisians”, you can segment these users with the filter In Possible Parisians - equals - true.

mParticle custom attribute

Both single array attribute and single string attribute

mParticle will send attributes as described by both single array attribute and single string attribute.

Both single array attribute and one attribute per segment

mParticle will send attributes as described by both single array attribute and one attribute per segment.

Both single string attribute and one attribute per segment

mParticle will send attributes as described by both single string attribute and one attribute per segment.

Single array attribute, single string attribute, and one attribute per segment

mParticle will send attributes as described by single array attribute, single string attribute, and one attribute per segment.

Step 1: Create an audience in mParticle

To create an audience in mParticle:

  1. Navigate to Audiences > Single Workspace > + New Audience.
  2. To connect Braze as an output for your audience, you must provide the following fields:
  1. Lastly Save your audience.

You should begin seeing audiences syncing to Braze within a few minutes. Audience membership will only update for users with external_ids (that is, not anonymous users). For more information on creating Braze mParticle audience, see the mParticle documentation on Configuration settings.

Step 2: Segment users in Braze

In Braze, to create a segment of these users, navigate to Segments under Engagement and name your segment. The following are two examples of segments depending on the option you selected for Send segments as. For more details on each option, see Forwarding audiences

  • Single array attribute: Select SegmentMembershipArray as your filter. Next, use the “includes value” option and input your desired audience ID. mParticle segment filter "SegmentMembershipArray" set as "includes value" and audience ID.

  • One attribute per segment: Select your custom attribute as the filter. Next, use the “equals” option and choose the appropriate logic. mParticle segment filter "in possible parisians" set as "equals" and "true".

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

Deactivating and deleting connections

Since mParticle does not directly maintain segments in Braze, it will not delete segments when the corresponding mParticle audience connection is deleted or deactivated. When this happens, mParticle will not update the audience user attributes in Braze to remove the audience from each user.

To remove the audience from a Braze user before deletion, adjust the audience filters to force the audience size to 0 before deleting an audience. After the audience calculation has completed and returns 0 users, delete the audience. Then, the audience membership will update in Braze to false for the single attribute option or removes the audience ID from the array format.

Data mapping

Data can be mapped to Braze using the embedded kit integration if you want to connect your mobile and web apps to Braze through mParticle. You can also use the server-to-server API integration to forward server-side data to Braze.

Regardless of which approach you choose, you must set up Braze as an output:

Configure your Braze output settings

In mParticle, navigate to Setup > Outputs > Add Outputs and selected Braze to open the Braze kit configuration. Save when completed.

Embedded kit integration

The mParticle and Braze SDK will be present on your application through the embedded kit integration. However, unlike a direct Braze integration, mParticle takes care of calling the majority of Braze SDK methods for you. The mParticle methods you use to track user data will automatically be mapped to Braze’s SDK methods.

These mappings of mParticle’s SDK for Android, iOS and Web are open source and can be found on mParticle’s GitHub page.

The embedded kit SDK integration allows you to take advantage of our full suite of features (push, in-app messages, and all relevant message analytics tracking).

Step 1: Integrate the mParticle SDKs

Integrate the appropriate mParticle SDKs into your app based on your platform needs:

Step 2: Complete mParticle’s Braze event kit integration

While the Braze SDK does not need to be directly included within your website or app for this mParticle integration, the following mParticle Appboy Kit must be installed to forward data from your app to Braze.

mParticle’s Braze event kit integration guide will walk you through custom mParticle and Braze alignment instructions based on your messaging needs (Push, Location Tracking, etc.).

Step 3: Connections settings for your Braze output

In mParticle, navigate to Connections > Connect > [Your desired platform] > Connect Output to add Braze as an output. Save when completed.

Not all connection settings will apply to all platforms and integration types. For a breakdown of connection settings and the platforms they apply to, see mParticle’s documentation.

Server API integration

This is an add-on to route your backend data to Braze if you’re using mParticle’s server-side SDKs (for example, Ruby, Python, etc.). To set up this server-to-server integration with Braze, follow mParticle’s documentation.

Connections settings for your Braze output

In mParticle, navigate to Connections > Connect > [Your desired platform] > Connect Output to add Braze as an output. Save when completed.

Not all connection settings will apply to all platforms and integration types. For a breakdown of connection settings and the platforms they apply to, see mParticle’s documentation.

Before enabling “Enriched User Attributes” or “Enriched User Identities” we recommend reviewing Data point overages to ensure you are aware of how these settings will impact data point usage.

Data mapping details

Data types

Not all data types are supported between both platforms.

  • Custom event properties support string, numeric, boolean, or date objects. It does not support arrays or nested objects.
  • Custom attributes support string, numerical, boolean, date objects, and arrays but does not support objects or nested objects.

Data mapping

User identity mapping

For each mParticle output, you can select the external identity type to send to Braze as the external_id. While the default value is customer ID, you can choose to map another ID, such as MPID, to send to Braze as the external_id. Be aware that choosing an identifier other than customer ID may influence how data is sent in Braze.

For example, mapping MPID to your Braze external_id will have the following effects:

  • Due to the nature of when MPID is assigned, all users will be assigned an external_id on session start.
  • Currents setup may require additional mapping due to differing data types between MPID and external_id.

Forwarding erasure requests (data subject requests)

Forward erasure requests to Braze by configuring a data subject request output to Braze. To forward erasure requests to Braze, follow mParticle’s documentation.

Potential data point overages

Enriched user attributes

Enabling enrich user attributes/identities (server-to-server only)

In the mParticle connection settings, Braze recommends turning off Include Enriched User Attributes. If enabled, mParticle will forward all available user attributes (such as standard attributes, custom attributes, and calculated attributes) from the existing profile to Braze on each logged event. This will result in high consumption of data points since mParticle will send Braze the same unchanged attributes on each call.

For example, if a user adds first name, last name, and phone number during their first session and later signs up for a newsletter adding the same information, in addition to email, triggering a newsletter sign-up event:

  • If turned on (default), five data points will be incurred. (sign-up event, email address, first name, last name, and phone number)
  • If turned off, two data points will be incurred (sign-up event and email address)

Considerations of turning off enriched user attributes

There are a few considerations to be aware of when turning off Include Enriched User Attributes:

  1. The server-to-server integration uses the mParticle events API to send events to Braze. Each request is triggered by an event. When a user attribute is changed, such as updating an email address, but is not associated with a specific event (for example, a profile update custom event), the new value is only passed to an output like Braze as an “enriched attribute” in the payload of the next event triggered by the user. When Include Enriched User Attributes is turned off, this new attribute value unassociated with a specific event will not be passed to Braze.
    • To solve this, we recommend creating a separate “user attribute updated” event that only sends the specific user attribute(s) that have been updated to Braze. Note that with this approach, you are still logging an additional data point for the “user attribute updated” event, but data point consumption will be far less than sending all user attributes on every call with the feature enabled.
  2. Calculated Attributes are passed to Braze as an enriched user attribute, so when “Enriched User Attributes” is turned off these will no longer be passed to Braze. To forward calculated attributes to Braze when “Enriched User Attributes” are turned off, a calculated attribute feed could help without pushing all the attributes. The feed will fire an update downstream to Braze when a calculated attribute changes.

Sending unnecessary or duplicate data to Braze

Braze counts a data point each time an attribute is passed to Braze, even if the value is unchanged. For this reason, Braze recommends only forwarding data needed to action on within Braze and ensuring that only deltas of attributes are being passed.

Github   Edit this page on GitHub
HOW HELPFUL WAS THIS PAGE?
New Stuff!