Skip to main content

Creating a Transactional churn

Overview

You can select a Transactional churn to run the churn model for identifying the customers with one-time purchase of products and services. This document can help to predict whether a customer is at risk of no longer performing transactional activities of your products or services.

To predict with a Transactional churn

Follow the below steps to create a customer prediction using transactional churn model:

  1. Go to Predict > Built-in.
  2. On the Predictions page, click the Create tab.

Alt text


  1. Select Master Data from the dropdown list. When you select the master data, the system validates the selected master data type and filters the predictions based on the selection.
  2. In the Customer churn model, click Use this model.

The Customer churn model form appears.


Alt text


  1. Click Get Started.

The Transactional churn model page appears.


Alt text


Add Model name

  1. In the Transactional churn model page, type a unique name in the Name text area.
  2. Type a unique name in the Output table name.
  3. Click Save & Proceed.

The Model preferences page appears.


Alt text


Add Model preferences

  1. In the Model preferences page, enter a value in the Prediction Window text area.
  2. Enter a value in the Churn Definition text area.

imageNote

The value that you enter should be minimum 60.

ItemDescription
Prediction WindowNumber of days under which the customers who may churn.
Churn DefinitionNumber of days where a customer has churned if they have made no purchases.
  1. Click Save & Proceed.

The Add required data page opens.


Alt text


Add Purchase Logs

Follow the below steps to customize the purchase history of your customers by assigning fields from your data that corresponds to your selected attributes:

  1. In the Add required data page, click Add New.

The Add Customer Data form appears.


Alt text


  1. Select Purchase history from the drop-down list.
  2. Select Transaction ID from the drop-down list.
  3. Select Transaction date from the drop-down list.
  4. Select Value of transaction from the drop-down list.
  5. Select Unique product ID from the drop-down list.
  6. Specify the information if this transaction was a return from the drop-down list.
  7. Click Save.

The created purchase log gets created under the Table column.


Alt text


  1. Click Save & Proceed.

The Additional data page appears.


Alt text


Add Additional data

Follow the below steps to add additional data for the model to predict the customers having higher risk of churn:

  1. On the Additional data page, click Add New.

The Add Customer Data form appears.


Alt text


  1. Select a Customer activity table from the drop-down list.
  2. Select a Primary key from the drop-down list.
  3. Select a Timestamp from the drop-down list.
  4. Select an Event from the drop-down list.
  5. Select Details from the drop-down list.
  6. In Activity type, perform one of the below actions:
ToDo
Select an Activity Type from the drop-down listClick Select from existing.
Type a new activity type in the Activity Type text areaClick Create new.
  1. Click Save.

imageNote

In case, you do not have any additional data to add, you can skip this step.

The Additional data gets created.


Alt text


  1. Click Save & Proceed.
  2. The Data update schedule page appears.

Alt text


Update Schedule

Follow the below steps to update schedule to run the transactional churn model:

  1. In the Data update schedule page, perform one of the below actions:
ToDo
Import new data to your model on a weekly intervalClick Weekly.
Import new data to your model on a monthly intervalClick Monthly.
  1. Click Save & Proceed.

The Review your model details page appears.


Alt text


To review and run your model

Follow the below steps to save your churn model or run it:

  1. In the Review your model details page, click the Edit icon to edit the details.
  2. Once you complete your review or editing, perform one of the below actions:
ToDo
Save and run your churn modelClick Save & Run.
Save your churn modelClick Save & Close.