CARRE: Counterfactual Action Retrieval and Reason Evaluation for Explainable Churn Prescription

Abstract

Churn models typically identify high-risk customers but do not specify which feasible retention action should be considered or why that action is appropriate. We present CARRE (Counterfactual Action Retrieval and Reason Evaluation), a three-stage framework that combines retrieval-augmented candidate generation, cost-aware counterfactual scoring, and large language model (LLM) reasoning. CARRE retrieves a predefined catalog of retention actions, estimates model-predicted churn-risk changes under explicit feature transformations, and generates a structured churn reason and a profile-grounded explanation for the selected action. On the IBM Telco Customer Churn dataset, CARRE achieves 79.8% greater mean model-predicted risk reduction than the plain SHAP baseline and 80.4% greater reduction than the cost-controlled SHAP+Cost baseline across 313 high-risk test cases; its cost-normalized efficiency is 10.5% higher than that of plain SHAP. On a 136-case reason-stratified evaluation sample, diagnosis-driven prompt refinement increases weak-label agreement from 79.4% to 90.4%, with no auxiliary-plan constraint violations; because the same sample was used for error diagnosis and re-evaluation, the post-refinement result is not an independent estimate of generalization. For 135 explanations generated using the pre-refinement v2 reason outputs, two cross-vendor LLM judges assign mean scores ranging from 4.02 to 5.00 out of 5, although one judge saturates on actionability, and a deterministic audit finds no contradictions among 66 verifiable profile claims. Retrieval ablations show that k=5 provides the best evaluated compromise between high candidate coverage and downstream reasoning agreement in this dataset. These results illustrate how retrieval, model-based counterfactual scoring, and language generation can be separated and jointly evaluated in a prototype churn-prescription pipeline.

Authors 3

Minjoo Kim, Sangjin Park, Seung Hwan Cho

Specification

Official page

arxiv.org/abs/2609.09766

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

Arxiv announce type

replace

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

arXiv id

2609.09766

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

Categories

cs.CL

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

PDF

arxiv.org/pdf/2609.09766

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

Primary category

cs.CL

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

Published

11 Sept 2026

Source:arXiv (Atom API + RSS)·T1·observed 7 h ago·high

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance