Multivariate typology is a valuable tool in exploring farming systems, facilitating effective communication for the target farmers with their varied needs. The use of practices and dairy herd productivity may vary according to their socio-personal and farm-specific attribute. In heterogeneity, developing unique recommendations and interventions for milk-producing households will be very difficult. The present ex-post facto study was conducted to identify homogenous clusters and assess animal husbandry practices use and dairy herd productivity across different identified clusters. The data was collected from 240 milk-producing households, equally divided across three selected districts, through personal contact using a pre-tested interview schedule. The respondent was - a person belonging to such milk-producing households- who had control over decision-related dairy farms and operations. Two multivariate techniques i.e. Principal component analysis (PCA) and cluster analysis (CA) were used. Four distinct clusters were identified with significant heterogeneity with their socio-personal and farm-specific attributes. More land-owning young and educated farmers in Cluster 1 used more animal husbandry practices than in other clusters and obtained significantly higher milk productivity than households in Cluster 2. Most milk-producing households with large herds and trained farmers (C3) and resource-poor milk-producing households in Cluster 2 and Cluster 4 received a medium level of milk productivity.

Keyword: Typology; Milk-producing households; Dairy herd productivity.