RESUMO
The declining birth rate, population ageing and the outbound migration of young people in recent years have created obstacles for local companies in recruiting skilled personnel. Similar factors have also affected recruitment in the outlying island of Penghu, Taiwan. Multiple-attribute decision-making (MADM) techniques for evaluation model development were incorporated in the present research to investigate the key determinants of talent recruitment by the coffee house operators in the outlying island of Penghu. The decision-making trial and evaluation laboratory (DEMATEL) approach was first implemented to (i) examine the extent of mutual impact between evaluation predictors and (ii) develop an influential network relation map (NRM) for indicator and sub-indicator assessments. Subsequently, the DEMATEL-based analytic network process (DANP) technique was utilised to compute the weights of every indicator and sub-indicator. Additionally, the modified VlseKriterijumska Optimizcija I Kaompromisno Resenje or VIKOR (in Serbian) approach evaluated every indicator and sub-indicator performance gap for empirical data analysis with suggested conclusions and recommendations. The results showed that the first dimension that should be improved is past experience, followed by personality traits and professional competence, according to the causal diagram of the NRM. The weights acquired through the DANP indicated that professional competence was the most influential dimension, followed by personality traits, whereas past experience was the least influential. Past experience revealed that the highest gap value is the most important for development if coffee house operators want to attain the appropriate levels. Finally, the theoretical, managerial, and practical implications were discussed. Notably, the research outcomes could optimise talent recruitment, establish appropriate talent recruitment methods for coffee house operators, and enhance the human resource quality in the outlying island of Penghu.