PurposeThis study explores the direct relationship between social media user-generated content (UGC), online search traffic and offline light vehicle sales of different models.Design/methodology/approachThe long-run equilibrium relationship and short-run dynamic effects between the valence and volume of UGC, online search traffic and offline car sales are analyzed by applying the autoregressive distribution lag (ARDL) model.FindingsThe study found the following. (1) In the long-run relationship, the valence of online reviews on social media platforms is significantly negatively correlated with the sales of all models. However, in the short-run, the valence of online reviews has a significant positive correlation with all models in different lag periods. (2) The volume of online reviews is significantly positively correlated with the sales of all models in the long run. However, in the short run, the relationship between the volume of online reviews and the sales of lower-sales-volume cars is uncertain. There is a significant positive correlation between the volume of reviews and the sales of higher-sales-volume cars. (3) Online search traffic has a significantly negative correlation with the sales of all models in the long run. However, in the short run, there is no consistent conclusion on the relationship between online search traffic and car sales.Originality/valueThis study provides a reference for managers to use in their efforts to improve offline high-involvement product sales using online information.