Predictive transactional churn

Important

This content is archived and is not being updated. For the latest documentation, see Microsoft Dynamics 365 product documentation. For the latest release plans, see Dynamics 365 and Microsoft Power Platform release plans.

Enabled for Public preview General availability
Users by admins, makers, or analysts Nov 20, 2020 -

Feature details

The transaction churn prediction feature enables you, without the help of a data scientist, to predict the likelihood that a customer will stop purchasing products or services. Using the prediction score, you can combine other information about your customers, like customer value, to create segments of high-churn risk or high-value customers. Use this segment to directly target customers through marketing activities, customer support, and other scenarios to reduce churn risk.

Configure the definition of churn as a time-based window specific to your business and define when customers are considered churned. For example, a grocery store might want to consider a customer churned if they have not purchased anything in the past 30 days.

As you continue creating the prediction, we'll guide you on what data is needed, and enable you to map data about your business to fields required to predict churn for your customers. You can also set a schedule to retrain the model based on new information in your system to adapt to changing business circumstances.

The first step in the wizard Model Preferences for transactional churn

The second step in the wizard Add customer data for transactional churn

Data schedule options for transactional churn

See also

Transactional churn prediction (docs)