Who cares for happy cows? Exploring views of dairy stakeholders around an imaginary automated animal-based welfare assessment tool.

Introduction: Precision Livestock Farming (PLF) provides additional opportunities beyond focusing solely on production and health-related traits, including the evaluation of animal welfare. This study examines the complexities of adopting an imaginary automated welfare assessment tool in Swedish dairy farming. Materials and methods: Through an iterative qualitative design, we engaged dairy farmers (n = 10), advisors (n = 5), dairy industry experts (n = 2), and PLF managers (n = 3) to co-develop insights and integrate multiple stakeholders' perspectives. Online focus groups (n = 7) served as a platform to explore participants' cultural nuances, discourses, and practical challenges surrounding animal welfare indicators related to feeding, comfort, health, and complementary behaviours. Results: A reflexive thematic analysis exposes strains between farmers' perceptions and other stakeholders' meanings and practices. Despite other stakeholders' assumptions of 'farm blindness,' farmers demonstrate awareness and interest in comfort-related welfare indicators. However, they experience difficulties implementing changes due to limited agency and infrastructural capacities. Other stakeholders often interpreted the lack of farmers' actions as indifference, overlooking farmers' nuanced prioritisation strategies. Equally, conversations underscore farmers' doubts about the commitment and backing of supervising bodies as they hesitate regarding data sharing among interested stakeholders. Crucially, this highlights a lack of shared understanding and motivation for the welfare assessment, and a lack of long-term advisory support that aligns with farmers' capabilities and constraints. Conclusion: Our study underlines the importance of bridging the gap between scientific knowledge and on-farm practices, particularly in defining actionable guidelines for addressing welfare concerns. Regardless of the concretion of our imaginary automated tool, we conclude that stakeholders could readily foster greater engagement with animal welfare issues by recognising farmers' agency capabilities and providing tailored, contextually relevant support that signals the industry's support to farmers, not only a self-need to retain a delicate license to operate thus truly facilitating meaningful gains in animal welfare.