An integrative compositional modeling framework for cross-cultural co-creation of multiespecies dulce de leche in the southern cone
Year: 2026
Type: article
Study design: Observational Study
Abstract: Co-creation has emerged as a promising strategy for integrating consumers and producers into food innovation; however, its application across culturally distinct contexts remains methodologically challenging. Cross-national studies often rely on averaged outcomes that obscure local variability, limiting the transferability of results. This study proposes an analytical framework combining co-creation with Bayesian Dirichlet regression to model compositional preferences while minimizing overgeneralization. Data were collected from 950 consumers through online questionnaires, complemented by focus groups with Brazilian milk producers. Focusing on dulce de leche – a culturally embedded dairy dessert in Brazil, Argentina, and Uruguay - the study examined sociodemographic, perceptual, and sensory determinants guiding the co-creation of a hybrid multispecies formulation. Results revealed significant cross-country differences in expectations and proportional preferences. Familiarity with non-bovine milks, positive sensory associations, and the valuation of creaminess and characteristic dulce de leche flavor emerged as consistent drivers of predicted milk proportions, alongside country effects. The model identified an analytically derived formulation scenario (54.5% cow, 19.4% buffalo, 16.1% goat, and 10.0% sheep milk), representing a probabilistic synthesis of consumer preferences across countries, with projected mean expected sensory intensities. This framework contributes to advancing cross-cultural co-creation research and provides practical guidance for product development in traditional dairy systems. • Significant cross-country differences emerged in sensory and perceptual drivers. • Bayesian Dirichlet regression modeled co-creation data across three countries. • Cross-cultural variability was captured without overgeneralizing sensory data. • Optimal blend combined cow, buffalo, goat, and sheep milk for balanced acceptance. • Framework offers a replicable tool for culturally sensitive food innovation.
Source: Food Quality and Preference
Language: English
Subfield: Food Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: hybrid
Estimated APC paid: $4,420
OpenAlex ID: W7155194401
