2.2.2 Customer AI - Create a New Instance (Configure)
Customer AI works by analyzing existing Consumer Experience Event data to predict churn or conversion propensity scores. Creating a new Customer AI instance allows marketeers to define goals and measures.
Setup a new Customer AI instance
In ÃÛ¶¹ÊÓƵ Experience Platform, click Services in the left menu. The Services browser appears and displays all available services at your disposal. In the card for Customer AI, click Open.
Click Create instance.
You’ll then see this.
Enter the required details for the Customer AI instance:
- Name: use
--aepUserLdap-- Product Purchase Propensity - Description: use: Predict the likelihood for customers to purchase a product
- Propensity type: select Conversion
Click Save and continue.
You’ll then see this. Select the dataset you created in the previous exercise which is named --aepUserLdap-- - Demo System - Customer Experience Event Dataset. Click Add.
You’ll then see this. you need to define the Identity field. Click None.
In the popup, select Identity Map (identityMap) and then select the namespace Demo System - CRMID (crmId). Next, click Save.
Click Save and continue.
Select Will occur in your specific dataset, and define the field commerce.purchases.value as the target variable.
Next, set your schedule to run Weekly and set the time as close as possible to your current time. Ensure that the toggle Enable scores for Profile is enabled. Click Save and continue.
After you configure the instance, you can see it in the Customer AI service list and you can also preview the summary of the setup and execution details by clicking on the Customer AI instance row. The summary panel will also display error details in case errors have been found.
Once your model has run, you’ll see this.
Next Step: 2.2.3 Customer AI - Scoring Dashboard and Segmentation (Predict & Take Action)