The paper describes the results of an experimental study carried out by sending bogus CVs in response to online posts. To each post related to a job vacation in a Veneto industry a number of five CVs was created and sent. The CVs were created so to experimentally evaluate the likelihood of discrimination against certain categories of candidates. The variability of call back rates of the addressed industries was analysed to elicit possible relationships between recruiters' taste and gender, nationality, academic curriculum and other social characteristics of applicants. Contrary to expectations, women obtained more call backs than male counterparts and graduates in a social or humanistic discipline were called back more than graduates in a scientific or technical one. Even the possession of a car was negatively evaluated by recruiters. As expected, instead, foreign-born candidates obtained far less call backs than Italian ones. A multivariate analysis drew us to the conclusion that the largest part of online posts was destined to candidates who belonged to intermediate categories, that is, to candidates being the best once the bottom and the top candidates were skimmed. The worst candidates were excluded for intuitive reasons, while the excellent ones - we conjecture - were left aside because recruiters guessed they would demand more than companies were available to give. These outcomes could help graduates in their duty of preparing an ad hoc CV.