Text-to-Price-Sensitivity: From User-Generated Content to Price Responsiveness in FMCG Markets
(A2026-134187)

AUTHORS

Anett Erdmann, Universidad Pontificia Comillas/Universidad Pontificia Comillas; Madeleen van Veen, University of Münster; Manfred Krafft, University of M�nster/University of Auckland

ABSTRACT

Price responsiveness is commonly quantified through price elasticity yet fails to capture the psychological evaluations underlying consumers’ reactions to price changes. This study introduces Text-to-Price-Sensitivity (T2PS), a theory-driven approach that infers consumers’ price sensitivity from user-generated reviews. We develop two measures: (1) a price-sentiment lexicon and (2) supervised text-embeddings to determine semantic price-sensitivity similarity. Building on a dataset spanning four years of consumer reviews, prices and sales across three FMCG categories, we show that T2PS can be used complementarily to traditional price response measures. In addition, we demonstrate systematic heterogeneity in text-based price sensitivity, and three distinct product clusters. Product-level demand models confirm that T2PS predicts sales beyond economic price elasticity, offering a theoretically grounded, scalable way to integrate digital consumer voice into behavioral pricing analytics.