Ideation is an increasingly demanded skill in the market. In view of this reality, crowdsourcing is a process that has gained use and notoriety in the last decade as a good way to encourage collaborative innovation. Its practice is based on the opening of themes and problems to large groups, so that, in this way, ideas and efforts can be shared openly through an online platform. However, the high volume of contributions in the process requires assertiveness in the selection of the ideas to be implemented. The purpose of this article is to propose criteria, methods and factors to evaluate ideas generated from a crowdsourcing approach. To this end, a systematic literature review was carried out by reading 90 (ninety) articles on the themes of crowdsourcing and evaluation of ideas. In addition, a case study was carried out at the Office of Ideas, an institution created within the Federal University of Rio Grande do Norte. This Institution aims to carry out projects to improve and innovate products and services for the university community, and acts essentially in the processes of capturing, evaluating and implementing projects in the organization. In this way, the internal processes of an innovation ideas office were observed and this made it possible to validate the elements that should compose an evaluation system, proposed through a focus group with representatives from various sectors of the organization. From these steps, the following elements for evaluating ideas were selected: Popular and technical filter, which evaluates ideas from a Popularity Index (IP); Detailing of ideas, which aims to minimize the factors that influence the evaluation process and get to know the ideas more deeply, solving doubts about their specifications and; Final evaluation by multicriteria and consensual analysis, which evaluates the ones using the multicriteria of originality, acceptability, applicability and effectiveness and, the influencing factors, which should be studied and considered by organizations.