Churn analysis studies user growth or shrinkage using behavioral features. This study uses churn analysis to examine pro-Russian and pro-Ukraine Telegram communities’ dynamics during the Russia-Ukraine conflict. The study answers two questions - (1) can churn management strategies help track campaigns/social movements based on user engagement trends? (2) can churn analysis help evaluate the effect of the growth or shrinkage of users on narrative propagation? The pro-Russian and pro-Ukraine narratives were generated from the Telegram channels to predict churners. We observed a significant change in the churn rate of pro-Russian narratives, while there was a notable fluctuation in user engagement with the pro-Ukraine narratives. Consequently, pro-Ukrainian narratives propagated faster than pro-Russian narratives, as demonstrated by the epidemiological model assessing narrative spread. This suggests a dynamic landscape in the dissemination and reception of these competing narratives.