The traditional analysis of unfilled vacancies is grounded primarily in the interpretationof the Beveridge curve and mainly in the explanation of skills mismatch. However,it is easily verified that this is insufficient. In this text we try to collect, based on theavailable literature, a number of variables that can consolidate the comprehension ofthis phenomenon. We try to build a model that discriminates these explanatory variables(administrative issues, vacancy features, labour market and job seeker characteristics)in order to provide users and decision makers a set of decision-making aids,based on evidence criteria, to set reliable employment and training policies and not onpolitical-ideological fluctuations