A colleague builds a KNN churn prediction model and reports:…

A cоlleаgue builds а KNN churn predictiоn mоdel аnd reports: Accuracy: 75% AUC: 0.72 Recall (churned class): 8% The dataset has a 26% churn rate. Explain why accuracy looks reasonable but recall is terrible. What would you recommend to improve the model's usefulness for a retention team?