Santos, AndréAntónio, NunoRita, Paulo2025-01-132025-01-082050-3318PURE: 105978314PURE UUID: c0d1fd3e-9e6a-4aa1-a57f-2c912131f880Scopus: 85217223758WOS: 001391706400001ORCID: /0000-0001-6050-9958/work/175650758http://hdl.handle.net/10362/177358Santos, A., António, N., & Rita, P. (2025). The Diderot effect: a data-driven validation. Journal of Marketing Analytics. https://doi.org/10.1057/s41270-024-00371-6 --- %ABS1% --- Nuno António and Paulo Rita were supported by national funds through FCT (Fundação para a Ciência e a Tecnologia), under the project - UIDB/04152/2020 - Centro de Investigação em Gestão de Informação (MagIC)/NOVA IMS) (https://doi.org/10.54499/UIDB/04152/2020).Although it is theorized that consumer decisions are commonly irrational and based on systematic biases, there remains a need for more data-driven research to validate and expand upon these assumptions fully. This study is one of the few that, based on analysis and interpretation of complex data instead of qualitative methods, validates one of those biases, the Diderot effect. This study presents a conceptual model and a pioneering research approach combining indirect data and machine learning techniques to validate the manifestation of the Diderot effect on the purchase process of products of a specific category through an online retailer. Results showed that a laptop computer could be one of those products and that consumers might be more predisposed to making unforeseen purchases when they are already in a spending mindset. The findings highlight opportunities for marketing professionals to leverage the Diderot effect, creating value for consumers and organizations.953087engDiderot effectConsumer Buying BehaviorComplementary ProductsGeneralized Sequential Pattern AlgorithmBiasesMachine LearningEconomics, Econometrics and Finance (miscellaneous)Strategy and ManagementStatistics, Probability and UncertaintyMarketingSDG 12 - Responsible Consumption and Productionjournal article10.1057/s41270-024-00371-6https://www.scopus.com/pages/publications/85217223758https://www.webofscience.com/wos/woscc/full-record/WOS:001391706400001