Application of RFM and PLS-SEM to Analyze University Students’ Purchase Behavior Toward Co Cây Hoa Lá Ginger–Lemongrass Herbal Shampoo on TikTok Shop in Bac Tu Liem

This study examines university students’ purchase behavior toward C? Cây Hoa Lá ginger–lemongrass herbal shampoo on TikTok Shop in Bac Tu Liem by integrating recency–frequency–monetary (RFM) analysis with partial least squares structural equation modeling (PLS-SEM). An online survey generated 458 responses, of which 387 valid questionnaires were retained for analysis. The structural model tested the effects of seven factors—perceived product quality, perceived value, brand trust, social influence, TikTok Shop shopping experience, attitude toward natural products, and price sensitivity—on purchase intention. The measurement scales met the required reliability and validity criteria. The KMO value for the independent-variable group was 0.688, total variance explained reached 72.376%, and both composite reliability and average variance extracted satisfied the recommended thresholds. Brand trust exerted the strongest effect on purchase intention (β = 0.365), followed by perceived product quality (β = 0.297), perceived value (β = 0.214), attitude toward natural products (β = 0.174), social influence (β = 0.151), TikTok Shop shopping experience (β = 0.143), and price sensitivity (β = 0.117). The model explained 37.9% of the variance in purchase intention. RFM analysis showed that Lost Customers constituted the largest segment (23.51%), followed by Potential Loyalists (19.90%) and Hibernating customers (18.35%), whereas Champions accounted for 10.34%. These findings provide an empirical basis for strategies aimed at strengthening brand trust, improving perceived quality and platform experience, and managing customers according to their respective RFM segments.