# Behavior Activation Overview

This feature is a public Beta release.

There are three broad categories of customer data: Identity, Attributes, and Behavior.

* User Identity: first name, last name, and email.
* User Attributes: age, income, interests, and so on.
* User Behavior: purchase logs, website usage logs, customer support interactions, and so on.


For more details on attributes and behaviors, see [Attributes and Behaviors](/products/customer-data-platform/data-workbench/parent-segments/overview-of-master-segments).

With Behavior Activation, you can create activations that export behavior data along with attribute data. Behavior Activation removes the previous limitation that activation was primarily limited to attributes or behavior data that has been converted into attributes with complex workarounds for activation purposes.

This topic contains:

* [Benefits of Behavior Activation](#benefits-of-behavior-activation)
* [How Behavior Activation Works](#how-behavior-activation-works)
* [Limitations](#limitations)


# Benefits of Behavior Activation

Behavior data of customers can be highly valuable for marketing, sales, engagement, or any customer experience (CX) campaigns. For example: personalization of email content, product or service recommendations based on a customer’s past behavior or creating reports based on customer behavior data. General benefits of using Behavior Activation are as follows:

* Harness power of behavior data in campaigns such as personalization.
* Reduce workarounds that convert behavior data into attributes, for activation.
* Decide during campaign which behavior data to activate.
* Removes or reduces dependency of business teams such as marketers on data engineers or IT teams.


Use case scenarios:

* Personalization : An e-commerce company wants to send personalized emails to its customers that have added items in cart and abandoned them without completing an order.
  * With behavior activation, the company is able to identify specific items added to the cart and exports those to downstream systems.
  * The items in abandoned cart can then be shown in an personalized email and remind customers to complete the purchase.
* Upsell : A pet food company seeks to recommend additional products a customer may be interested in, based on a recent purchase.
  * The company uses a recommendation model to generate a next best product recommendation and stores it in a behavior table.
  * With behavior activation, it activates the next best product for a customer and sends it over a personalized communication.
* Churn Prevention : A media company identifies customers that have not logged in to check out their products in the past two months, which indicates a possible churn risk.
  * Using information such as the time since last login, and customer's favorite product, the media company sends a personalized communication to the customer and provides a nudge to check it out again.


# How Behavior Activation Works

* Behavior table selection:
  * You can select a table only from the behavior tables that have been configured in segment rules.
  * A behavior table that resides inside a referenced segment, cannot be selected for Behavior Activation. To enable selecting this table for Behavior Activation, add it to the current segment with the activation.
* Activated behavior data:
  * To determine the activated behavior data, Behavior Activations filters in two steps: 1. Segment rules determine what profiles will qualify for that segment. 2. From the behaviors of the qualifying profiles in step 1, the activated behavior data is determined by the segment rules from the behavior table selected in the activation.


For steps on how to configure Behavior Activation, see [Configuring Behavior Activation](/products/customer-data-platform/audience-studio/activation/configuring-behavior-activation).

## Limitations

Behavior Activation has certain limitations currently. These will be addressed in upcoming releases.

* Only one table data can be selected for behavior activation.​
* A behavior table should be used in only one rule set in segment rules.
  * If the same column from a behavior table is used in two or more rule sets, then no behavior data gets activated.
  * If different columns from the same behavior table are used in two or more rule sets, then the activated behavior data might be incorrect.
* Behavior data can be correctly exported only when the segment rules meet the following conditions:
  * Aggregation operators usage: Sum, Count, Average only
  * Match logic usage: All (not 'Any' nor 'Advanced')
* If you want to order the activation data by a certain column data, ensure that the source column name has not been mapped to a different output column name but has the exact same name.