Retention · Random-forest churn prediction followed by a field experiment comparing risk-based and random targeting
Proactive customer retention management in a non-contractual B2B setting based on churn prediction with random forests
Industrial Marketing Management · 2022
Summary
Risk-based proactive contact reduced churn significantly versus random targeting and had positive revenue impact; recency, frequency, monetary value, and last field-representative contact were important features.
For your marketing
Possible founder application
Use as a bounded example of proactive customer retention management in a non-contractual b2b setting based on churn prediction with random forests; compare its context and design with your own market before applying its result.
Reported result
What the study found
Risk-based proactive contact reduced churn significantly versus random targeting and had positive revenue impact; recency, frequency, monetary value, and last field-representative contact were important features.
Research question
What the authors investigated
Proactive customer retention management in a non-contractual B2B setting based on churn prediction with random forests?
Design
Random-forest churn prediction followed by a field experiment comparing risk-based and random targeting
Sample and setting
Non-contractual B2B wholesale customer setting
Scope and transfer
Limitations to keep in view
The evidence comes from one wholesale B2B setting and a non-contractual context; model features and treatment effects may not transfer to contractual, consumer, or other industries. The accessible publisher page was used for this listing; correction history was not independently checked.
Research sources
The article reports a B2B wholesale field experiment in which contacting customers with the highest predicted churn probabilities reduced population churn versus random targeting and improved revenue development. Publisher page was directly retrieved; detailed full-text methods were not independently verified.