Findings from a multi-site user needs and requirements survey to inform development of the WiseFood application

Individuals experience persistent barriers to adopting sustainable dietary practices, including cost, availability, and low food literacy (2,3) . While digital tools and Artificial Intelligence (AI) offer promising avenues for supporting sustainable dietary behaviour change, few interventions target the household as a unit of change (4) . WiseFood (5) aims to address this gap by using AI to empower households in making informed, sustainable food choices. It seeks to achieve this by co-designing AI-powered tools through multi-site Living Labs (LL) across Europe. This study explores household decision-making to identify user needs and requirements (UNR) for an inclusive, evidence-based innovative digital platform designed to help citizens make healthier and more sustainable food choices. A 37-item household survey was developed across LL sites, with input from nutrition experts and pilot-tested for clarity and relevance. The survey covered six sections: demographics, behaviours and attitudes, views on AI in food decisions, desired features, interaction preferences and barriers, and expectations for WiseFood; it was disseminated online via REDCap and JotForm. Standardised recruitment strategies were established to ensure consistency across LL sites. Purposive recruitment targeted diverse household types including single-person, single-parent, cohabiting couples (with/without children), older adults and varied socioeconomic backgrounds. Quantitative data were analysed using descriptive statistics in Excel while open-ended responses were analysed using content analysis. Ethical approval was obtained from committees in Ireland (RCSI Research Ethics Committee (REC202504022)), Slovenia (Research Ethics Committee of ITC Murska Sobota (EC-ITC-2025-01)) and Hungary (Research and Innovation Committee of the University of Veterinary Medicine Budapest (2025/08/27/1)). Between June-July 2025, 144 participants completed the survey: cohabiting couples with children (n = 38, 26.4%), cohabiting couples (n = 26, 18.1%), single-person (n = 26, 18.1%), single parents (n = 22, 15.3%), older adults (n = 21, 14.6%), and other (n = 6, 4.2%). Thirty percent of participants (n = 43) identified as being very health-conscious, while taste and sensory appeal were the most influential factors in food choice, rated as important by 81.9% (n = 118) of participants. Only 36.8% (n = 53) considered sustainability important when making food choices. Meal planning (n = 98, 68.1%) and budgeting support (n = 65, 45.1%) features were identified as most important for a digital tool like WiseFood. While only 15.3% (n = 22) had previously used AI for food choice, 72.7% (n = 104) were open to using a tool like WiseFood. Barriers included digital fatigue (n = 14, 36%), complexity and usability concerns (n = 16, 41%), and lack of trust in AI (n = 15, 38%). Findings highlight that digital tools such as WiseFood must be practical, cost-sensitive, inclusive, and equipped with actionable, multi-modal features to support diverse households. These insights will inform the co-design phase of the WiseFood Project.