This study investigates how consumer satisfaction in the restaurant sector varies across seasonal periods and geographic contexts, using over one million Yelp reviews from Philadelphia and Tampa—two cities with distinct climatic and economic profiles. By integrating structured data (star ratings) with unstructured data (text-based sentiment scores), the analysis applies sentiment analysis, Ordinary Least Squares regression, ordered logistic models, and Random Forests to detect spatiotemporal patterns in consumer evaluations. Results show that satisfaction is generally higher in warmer seasons, but the strength of seasonal effects differs across cities, with Tampa exhibiting greater seasonal variation than Philadelphia. Sentiment scores are more sensitive to seasonal shifts than star ratings, and systematic divergence between the two emerges across cities and seasons. These findings suggest that economic structures, such as tourism reliance, may shape satisfaction more than climate alone, and highlight how textual analysis captures emotional nuance often masked by numeric ratings. The study also demonstrates how combining structured and unstructured data can reveal nonlinear and context-dependent satisfaction dynamics that conventional models may overlook. This contributes to consumer research by offering an integrated framework for modeling seasonal and geographic variation in satisfaction and by validating the utility of sentiment analysis in large-scale behavioral studies. Practical implications for service management and design are discussed, and future research directions are proposed to extend this work across diverse service contexts and environmental conditions.