Consolidating internal customer data helps to combine data from multiple systems into a 360-degree view of your customers. This consolidated view provides your organization with an excellent starting point toward understanding who your customers are. However, this information doesn't often provide a complete picture of your customers. You also need to have relevant data from external sources to provide context and a deeper understanding, which will help target your customers more appropriately. Additionally, with technologies such as AI and machine learning, you can use historical trends to make predictions that your organization can act on.

The Audience Insights capabilities of Customer Insights provides two features that address these areas:

  • Data enrichment - Merges other data sources with your organization's data.

  • Predictions - Uses AI and machine learning to make predictions.

Data enrichment

Data enrichment allows organizations to merge data from other authoritative data sources with your Microsoft customer data. This feature allows your organization to take more informed and targeted approaches toward nurturing and marketing customers.

For example, an organization that sells recreational vehicles and outdoor equipment can build marketing segments based on internal systems that identify customers who would be interested in or are due for an equipment upgrade. However, what the organization might not know is which customers prefer specific brands or types of equipment. By enriching customer data with information such as brand affinities, you can create more targeted marketing segments of customers who prefer brands of equipment that you currently have large quantities of in stock. Audience Insights provides organizations with the ability to enrich their customer profiles with data such as interests, brand affinities, demographic information, location-centric data, and more.


As your organization conducts business with your customers, you'll collect increasingly more historical data. You'll have records that define who customers are, what types of purchases they make, what subscriptions they have, and more. Organizations can use this available data to start making predictions to help drive activities.

For example, customers who purchase memberships or subscriptions with a company are not likely to renew if they don't see value in continuing with it. By using AI and machine learning on your organization's historical data, you can make predictions on how likely customers are to cancel by comparing their activity to the historical records that you have.

If it's identified that the customer is not likely to continue, you can initiate activities with the customer to try reducing the likelihood of cancellation such as sending coupons or adding them to a nurturing program.

Audience Insights provides multiple ways for organizations to use AI to understand their data and make predictions based on what is happening. These methods could be as simple as predicting the value of a missing field such as a record state, customer gender, or likelihood of a specific outcome. You could also use Audience Insights to represent more advanced scenarios that use custom Microsoft Azure machine learning models from within Customer Insights.

The rest of this module will examine the different options for data enrichment and predictions that are available.