Adjusting Propensity Model Scores During Economic Shifts: A Framework for Short-Term and Long-Term Adaptation

Abstract

The automotive industry experienced significant disruptions during the COVID-19 pandemic, with consumer behavior undergoing dramatic shifts. Vehicle purchases declined due to economic uncertainties, while service intervals were extended as customers postponed maintenance visits. These changes rendered pre-pandemic propensity models less effective in predicting customer behavior, as they relied on outdated patterns. Rebuilding models with updated data was not feasible in the short term due to limited post-pandemic data. This paper presents a framework to adjust propensity model scores using post-recession priors, ensuring relevance during dynamic periods. The methodology leverages observed response rates to recalibrate predictions, maintaining business utility. Applications in vehicle purchase and service propensity modeling are explored, with short-term strategies for immediate adjustments and long-term strategies to enhance model resilience.