Publication Details
Abstract
The growing volume, variety, and velocity of consumer data have reduced the adequacy of marketing research approaches that rely exclusively on descriptive statistics and linear models. This article develops an artificial intelligence (AI)-driven methodology for conducting marketing research in the food products market, with a proposed empirical application to Uzbekistan. The framework integrates theory-guided survey measurement, structural equation modelling, unsupervised consumer segmentation, supervised purchase-intention prediction, and explainable AI. It evaluates the effects of perceived product quality, price perception, food safety, brand trust, product information, online reviews, acceptance of AI-based recommendations, privacy concerns, and environmental considerations on purchase intention. A cross-sectional survey of approximately 500 adult food purchasers is proposed. Conventional explanatory analysis is combined with K-means or hierarchical clustering, random forests and gradient boosting, while SHAP values make model predictions interpretable. The framework addresses a methodological gap between explanatory consumer research and prediction-oriented AI analysis. It offers a reproducible sequence for data collection, validation, segmentation, prediction, interpretation, and managerial translation. Because primary data collection is pending, the article presents a testable methodological protocol rather than fabricated empirical results. The approach can help food producers, retailers, and digital platforms generate more timely, granular, and accountable consumer insights while preserving theoretical validity and responsible data governance.