In this guide you'll learn what our Audience Segmentation module is all about and learn about the main parameters in using it for your audience selection and personalized targeting campaigns.

Overview

The Audience Segmentation API allows you to segment users based on various criteria, enabling you to target specific audience groups for personalized marketing campaigns, content delivery, and user analysis. Segmentation is achieved through facets, which are key attributes or characteristics used to categorize users. This documentation provides an overview of facets and their types, along with examples and a comprehensive list of possible elements within each facet type.

Facets

A facet is a specific attribute or characteristic used for audience segmentation. It provides valuable insights into user behavior, preferences, and demographics. The Audience Segmentation API supports four types of facets:

• Interests
• Demographic
• Companies
• Pre-defined Personas

When building audiences, you'll select facets from one or more of these facet types to segment and group users.

Below, we'll go deeper into each type of facet.

Demographic Facets

These elements within the demographic facet provide insights into various aspects of users' demographics, financial situations, living arrangements, and life stages. By incorporating these elements into your audience segmentation, you can better understand and target specific groups based on their unique characteristics and circumstances.

Interests Facets

Interests facets segment users based on their preferences, buying intent, or areas of interest. These facets enable you to tailor your content and offerings to users who share common interests, resulting in higher engagement and conversion rates. Examples of interests facets include:

Entertainment: Museums, Zoos, Live Concerts, Gambling, ...
Hobbies: Arts & Crafts, Musical Instruments, Reading, Games, ...
Travel: Cruises, Hotels, Plane Tickets, Money Exchange, Car Rentals, ...
Foods & Drinks: Bars, Coffee Shops, Restaurants, Juices, Food Delivery, ...

Pre-defined Persona Facets

Pre-defined persona facets segment users based on predefined archetypes or profiles - which we call personas. These personas are built by combining demographic, interests or company facets to target specific financial products. Examples of pre-defined personas facets include:

Financial Profiles: Saver, Savvy Investor, Life Insurance Candidate, ...
Personal Profiles: Car Owner, Pet Owner, ...

User Attributes

Hyperplane offers the capability to enrich user segmentation by employing additional attributes, when available, to fine-tune the filtering of responses from certain Audience Segmentation endpoints. One key endpoint that benefits from this enhanced filtering is POST /persona/{{persona_id}}/users. Explore an illustrative example of this in our recipe documentation here.

When supplementary user data, beyond transaction details, is provided (such as user identity or KYC data), our platform enables you to filter users on this endpoint using the following attributes:

• Birth Date: Apply a date filter based on a user’s birth date.
• Declared Monthly Income: Implement a numerical filter tied to a user’s declared monthly income.
• Job Title: Utilize a categorical filter related to a user’s job title.
• CBO Code: Employ a categorical filter using a user's occupation code (CBO).
• Branch: Implement a categorical filter associated with a user’s banking agency/branch.

🚧

Should user identity data be absent for users within the provided transaction data, applying these filters will yield zero users, as identification with the requisite attributes will not be possible.

Moving forward, we aim to expand our filtering capabilities. Imminent enhancements include the addition of an in-house inferred income filter and further identity filters, enabling even more nuanced user segmentation.

Company Facets

Company facets categorize users based on their interaction or purchase behavior with specific brands, companies or merchants. These facets are particularly useful for B2B marketing, partnerships or when targeting specific industries or companies.

This facet type is still in beta and will be released soon.

Naming Convention of Facets

Facet names follow a naming convention that helps organize and categorize the facets. The naming convention separates facets and their levels using colons (:). The first part of the facet name represents the facet type, followed by the facet name itself, and optionally, multiple levels forming a hierarchy. It's important to note that facet names are case insensitive.

For example, consider the following facet names:

• interest:bicycle
• interest:shopping_and_fashion
• interest:shopping_and_fashion:clothing
• interest:shopping_and_fashion:accessories
• demographic:financial:affordability:paycheck_to_paycheck
• demographic:financial:affordability:breadwinner
• demographic:financial:affordability:affluent

In these examples, the facet type is indicated at the beginning of each name (e.g., pre_defined_persona, demographic, interest), followed by the facet name and, optionally, multiple levels forming a hierarchy. The hierarchy allows for more granular segmentation within a facet type.

Examples and Explanation

Let's explore a few examples to understand the naming convention and hierarchy in facets:

interest:bicycle

Facet Type: interest
Facet Name: bicycle
This example falls under the interest facet type, targeting users with an interest in bicycles. It represents a specific segment of users who are interested in cycling, whether it's for recreational purposes, competitive sports, or bike-related hobbies.

interest:shopping_and_fashion:clothing

Facet Type: interest
Facet Name: shopping_and_fashion:clothing
This example falls under the interest facet type, specifically targeting shopping and fashion interests, with a more specific level of interest in "clothing."

pre_defined_persona:lifestyle:pet_owner

Facet Type: pre_defined_persona
Facet Name: lifestyle:pet_owner
This example represents a pre-defined persona facet related to lifestyle, specifically targeting individuals who are pet owners. The facet name includes a hierarchy level to indicate the persona category. It signifies a segment of users who have pets as part of their lifestyle and are likely to have unique needs and interests related to pet care, pet products, and pet-friendly activities.

demographic:financial:affordability:paycheck_to_paycheck

Facet Type: demographic
Facet Name: financial:affordability:paycheck_to_paycheck
This example belongs to the demographic facet type, focusing on financial factors related to affordability. The facet name includes multiple levels forming a hierarchy. In this case, it represents the affordability level of being "paycheck to paycheck".

API Conventions

Header

For the header of your request, you'll need a few pieces of information:

• authentication token
• module ID

Below is a request example with what a header should look like:

curl -X 'GET' \
'https://api.<URL>.chico.ai/v1/persona' \
-H 'authorization: <AUTHENTICATION_TOKEN>' \
-H 'module-id: <MODULE_ID>' \
-H 'accept: application/json'

For a comprehensive code guide on this, check out our Audience Segmentation Recipe.

Request

When creating a new segmentation, you'll provide a list of facets, based on their facet types (demographic, interest, etc).

Here is an example:

{
  "persona_name": "life insurance candidates",
  "locations": [],
  "interest_facets": [
    { "name": "child_care", "weight": 0.2},
    { "name": "kindergarden", "weight": 0.2},
    { "name": "savings_and_investment", "weight": 0.2},
  ],
  "demographic_facets": [
    { "name": "financial:affordability:affluent", "weight": 0.7},
    { "name": "financial:affordability:breadwinner", "weight": 0.8},
  ],
  "company_facets": [],
  "pre_defined_personas": [
    { "name": "investor:savvy_investor", "weight": 0.75},
  ]
}

Notice how we used child_care instead of interests:child_care under the interest_facets request. Both options are ok to use.

Also, as a reminder, these names are case insensitive, so it would have been equally acceptable to have used CHILD_CARE or INTERESTS:CHILD_CARE as an input.

List of All Possible Facets

Below is the list of all the possible facets you can use in the audience segmentation module.

Demographic Facets

Interests Facets

Pre-defined Personas Facets

Company Facets

Still in beta. Will be released soon.