Is it possible to calcuale CLV at an early stage for individuals?

by Liri   Last Updated May 22, 2020 22:19 PM - source

I'm working on a project in which we have to predict the Customer Lifetime Value (CLV) for a group of customers. In order to calculate CLV in a non-contractual setting, we use probabilistic BTYD models in combination with the gamma-gamma submodel. With only a transaction log we can estimate the future cash-flow of an individual. However, most often we need a transactionlog with at the very least one year of data.

However, CLV calculation might be even more useful if we can assign a CLV value to a customer before an individual has bought something. A company can then focus on acquiring the most valuable customers. Obviously, we don't have transaction records for these individuals. I was curious to know if someone knows how to calculate the CLV for such individuals? What type of data do we need? And what type of models can we use to predict "early" CLV for these kind of individuals?

Any tips, suggestions or resources are much appreciated!



Related Questions



Attribution Models reset goal after conversion

Updated January 28, 2019 10:19 AM


Telemarketing Data Analysis

Updated November 24, 2018 04:19 AM