Create an Audience

Creating an audience begins with defining your use case, segmentation approach, and engagement strategy. To do this, consider the following questions:

• What is the business goal you are trying to accomplish?
• Which customer segments are important to achieving this business goal?
• Once you identify the relevant users, what is your strategy for engaging them?

Example Use Cases

Increase User Retention for a Mobile App

Retaining users is crucial for the sustained success of a mobile application. By identifying and re-engaging inactive users, you can boost overall engagement and reduce churn.

Goal: Improve user retention by targeting users who have not opened the app in the last 7 days.
Segmentation Strategy: Identify users who have been inactive for a week but used the app actively within the previous month.
Engagement Strategy: Send personalized push notifications offering a discount or exclusive content to re-engage users.

Predictive Audience for Likely Purchasers

Leveraging predictive analytics allows businesses to anticipate user behavior and proactively engage those most likely to convert, thereby optimizing marketing efforts.

Goal: Increase conversion rates by targeting users predicted to make a purchase in the next 7 days.
Segmentation Strategy: Utilize Predictive Audiences to identify users with a high likelihood of purchasing based on machine learning models analyzing past behaviors and interactions.
Engagement Strategy: Deliver personalized email campaigns featuring product recommendations or special offers to these high-likelihood purchasers.

Optimize Engagement in a Subscription Service

Identifying subscribers at risk of churning enables proactive engagement strategies to maintain a stable subscriber base.

Goal: Minimize churn by identifying at-risk subscribers.
Segmentation Strategy: Identify users whose subscription renewal is within the next 30 days and who have decreased engagement (e.g., fewer logins or interactions).
Engagement Strategy: Offer these users tailored incentives such as a loyalty bonus or a discounted renewal rate to encourage continued subscription.

Create a New Audience

Once you have clearly defined your use case, it’s time to begin building your audience.

Navigate to the Audience Creation Page

1. From the mParticle Overview Map, select Segmentation.
   
2. On the Audiences landing page, click Create New.
   

Configure Your Audience Strategy

Individual audiences are contained within folders called Audience Strategies. The first step in creating a new audience is to set the configurations for this folder:

1. Enter a name for your folder.
2. Select the platforms and feeds that will supply data to define the audiences within your folder.
3. Click Create.

Define Your Audience

After saving your audience strategy configuration settings, you’ll enter the Editor. Here is where you can add, view, edit, and connect audiences.

Follow the steps below to create your first audience within this folder:

1. Click the Add criteria box to open the Audience Builder Modal.
2. In the audience builder modal, choose the environment (Production, Development, or both) from which the audience will receive data. (Note the audience environment considerations below.)
3. Click Add Criteria to begin adding criteria for your audience.
4. Select the type of data you would like to use for your first criteria (more on criteria types below).
5. Referring to the data types and matching rules below, use the criteria editor to further target the users who fit your use case. .
6. Once you have added your first criteria, click Done.

After you have added your first criteria, a number displays that represents the estimated audience size: . This estimate is based on a sample of data. As you continue to add criteria, you will see an estimated size for both individual criteria as well as for the whole audience.

4. Continue adding more criteria using And, Or, or Exclude logic to further refine your audience.

Precise vs. Estimated Audience Sizes

Throughout the Audience creation process, you will see both preliminary and precise audience sizes.

Preliminary Estimates

Preliminary estimates are displayed before the audience has been fully calculated, and are denoted with a ~ character throughout the Audiences feature. When an Audience is created or edited, for example, preliminary estimates are shown to give an approximate idea of audience size.

Precise Estimates

Once an audience has finished calculating, the precise estimate will be displayed for that audience. This can take up to 20 minutes for the precise estimate to finish calculating.

Activate an Audience in a Campaign

Once you have created an Audience that you want to forward to an external tool for use in a campaign, click the Connect Output button in the Audience tile, then follow the steps to connect that audience to any of your connected outputs.

Main Criteria Types

You can build criteria based on two main sources of data:

Events

Event criteria check for specific events and their properties, and their availability is subject to the data retention policy of your account. Within the new criteria option in the audience builder, the following options create event-based criteria:

• Events
• Ecommerce
• Crashes
• Installs
• Uninstalls
• Sessions
• Upgrades
• Screen views

User Profiles

These criteria check your active user profiles, and their availability is subject to the user profile retention policies of your account. Within the new criteria option in the audience builder, the following options create profile-based criteria:

• Users: Access user profile information such as user attributes, calculated attributes, current audience memberships, consent state, location, etc.
• User Predictions: Predictive Attributes that you have defined and generated for your users.
• Attribution: Access user install and uninstall information to build criteria based on the attributed campaign and publisher.

Data Types and Matching Rules

Audience criteria can be created with several different data types, each with its own matching rules.

Strings

When building audiences based on string attributes, several case-insensitive matching rules can be applied:

• Contains / Does Not Contain: Matches substrings. For example, "blue" matches both "blue" and "blue shirt".
• Exact Match / Does Not Match: Requires the entire string to match. For example, "blue" matches "blue", but not "blue shirt".
• Pattern: Allows wildcard matching. * represents any number of characters, and ? represents any single character. For example, "bl?e" or "b*e" would both match "blue".
• Includes / Does Not Include: Matches an exact value within a list. For example, specifying "Chicago" in a list of movies returns "Chicago (2002)" and "Chicago (1927)", but not "Chicago Cubs".
• Partial Match: The inverse of Includes. For example, specifying "Chicago" would return all movies with "Chicago" in the title.

Date-Based Criteria

Filters based on fixed calendar dates. For example, events occurring after 09/12/2018. Date-based criteria are defined in UTC and are not relative to when the audience is calculated:

• Before: Excludes the specified date. For example, “Before September 9, 2018” includes events up to 11:59 PM UTC on September 8.
• After: Includes the specified date. For example, “After September 9, 2018” includes events from 12:00 AM UTC on September 9.
• On Date: Covers events from 12:00 AM to 11:59 PM UTC on the specified date.
• Between Dates: Includes events from the start of the first date to the end of the second date.

Recency-Based Criteria

Defines a period relative to the current time. For example, users active within the last 7 days. Recency-based criteria select events occurring within a timeframe relative to ‘now’.

• For example, if calculated at 1:00 PM on September 9, 2018, a recency filter for the last 7 days includes events between 1:00 PM on September 3 and 1:00 PM on September 9.

Attribution

Attribution criteria segment users based on campaign interactions, such as app installs or re-engagements.

• Profile Criteria: Select Attribution, then choose Install or Uninstall to filter based on campaign and publisher fields from attribution events.
• Event Criteria: Select Events > Attribution to use install or engagement events, filtering by event attributes. This allows you to select any information included with the event as custom_attributes.

Note: All event criteria are subject to audience event retention limits.

Identity

Identity criteria segment users based on their stored identities. You can test for the existence of a specific identity or apply string-based logic. These criteria are scoped to the workspace in which the audience is created. For example, if your account has three workspaces, an audience in one workspace only includes users active in that workspace.

Location

Location criteria allow segmentation based on geographic information.

• Equals: Filter users in a specific city, state, zip code, or DMA (Designated Market Area), based on IP geolocation.
• Within: Filter users within a set distance of a global city using latitude and longitude coordinates.

Cart

For ecommerce events, you can target users who added items to their cart but did not complete a purchase.

• Cart Abandonment: Select New Criteria > Ecommerce > Shopping - Cart Level > Cart Abandonment to define this segment.
• Specify the time period to wait without a purchase before including users in this audience.

Attribute Key

Use Exists or Not Exists to check for the presence of an attribute.

• For example, User Attribute: Gender EXISTS evaluates as true for both Gender = "Female" and Gender = undefined.

Additional Audience Features

Auto-complete

As you define your audience criteria, a list of suggested matching values will appear based on what you’ve entered.

This feature works both when building new audiences and fine-tuning existing ones, helping you save time, reduce manual effort, and improves accuracy. To use it, you must have one of the following standard Roles: User, Admin, Audiences-only, Support, or Admin+Compliance. Alternatively, you can create a Custom Role with any of the following tasks: audiences:draft, audiences:edit, catalog, or audiences.

Boolean operators

Once you have added criteria, you can use the Boolean operators And, Or, and Exclude to create logical relationships with subsequent criteria.

Audience Environment Considerations

Be mindful of your selected audience environment:

• Use Production for real customer data to prevent test data from overwriting profiles in partner systems.
• Ensure parent and child audiences share the same environment. If environments differ, child audiences will not inherit users from their parent.

Audiences in Development display a badge in the Audience Strategy Editor, while Production audiences do not.

By following these steps and best practices, you can build and activate audiences that align with your business objectives, enabling more targeted user engagement and monetization strategies.