Hi @Rick x Kulin,

Thanks for sharing this with us.

Using average date between orders is a fine strategy to help tweak the days used in the segment definition. For reference, the days we outline in the article are designed as a general framework that may benefit from a slight change depending on the industry.

I’m interested to hear about the experiences of other community users on this as well.

I’m also going to share another community post which speaks to churn risk prediction that may be helpful for people reading along

I hope that’s helpful.

​@Dov 

Has the “Churn Risk Prediction” property been removed from the available properties available to use in segment definitions and CSV property exports? 

The article “How to segment by customer lifetime value (CLV)” states something that I don’t find to be true anymore which is:

“Exporting CLV data can allow you to further analyze and predict the behavior of different groups of customers. In addition to your CLV and predictive analytics values, you'll be able to export Churn Risk Prediction. Churn risk will be exported into your CSV as a number between 0 and 1.”

I see each profile shows me predictive analytics, including “Churn Risk prediction” which tells me that my client has enough data to determine predictive analytics. However, When I try to export a segment via CSV, there are no property options to select for export that has anything to do with “Churn Risk”. The only predictive properties I have options to select for export or in segments definitions are the following:

• Historic Customer Lifetime Value
• Historic Number Of Orders
• Average Order Value
• Average Days Between Orders
• Predicted Number Of Orders!-->
• Predicted Customer Lifetime Value!-->
• Expected Date Of Next Order
• Total Customer Lifetime Value
• Predicted Gender