Sensory profiling and consumer acceptance of mozzarella cheese in a pizza application across 5 Chinese market test sites.

Mozzarella cheese (MC) is increasingly used in pizza, bakery, and convenience foods in China, but its sensory performance and consumer acceptance may vary across regional market contexts. Because MC is usually consumed after heating rather than as a standalone cheese, evaluating consumer responses in a realistic pizza application is important for product optimization. This study characterized 22 commercial low-moisture shredded MC samples after baking on pizza using quantitative descriptive analysis (QDA), and 4 representative samples were selected for consumer evaluation. A total of 320 young consumers from 5 Chinese regional market test sites, including Beijing, Chengdu, Guangzhou, Shanghai, and Wuhan, evaluated the samples using a 9-point hedonic scale and a 5-point just-about-right (JAR) scale. The 22 samples differed significantly in all 16 QDA attributes, showing broad variation in appearance, dairy aroma, saltiness, sourness, melting quality, oil separation, and chewiness. Mixed-model analysis showed significant effects of sample, test city, and their interaction on overall liking, with small-to-moderate effect sizes. Dairy-related flavor and yellowish appearance were associated with higher liking for some samples and consumer clusters, whereas excessive saltiness was the clearest penalty for MC-3, with total mean drop values of 2.10 in Beijing and 1.41 in Guangzhou. Excessive oil separation and insufficient dairy flavor also reduced liking in several sample-city combinations. K-means clustering identified 3 consumer segments, indicating that preference patterns were not fully explained by test city. Exploratory PLSR-based external preference mapping suggested associations between liking and dairy aroma, color, smoothness, melting quality, mouthcoating, saltiness, sourness, and chewiness. These findings indicate that MC products intended for pizza applications in China should be optimized by considering both common sensory penalties and city-associated consumer response patterns.