Video game companies often have to choose among thousands of ideas to decide which ones to turn into video games. Despite the huge amount of money at stake, this process known as green-lighting in the game industry is largely a guesswork based on experts’ experience and intuitions. In this paper, through a case study with a prominent Swedish gaming company we identify the key factors influencing green-lighting decisions and develops a typology for evaluating game concepts more effectively. These factors can be used to build a data-driven model that integrates domain knowledge with AI/ML techniques, including natural language processing (NLP), to predict a game’s potential return on investment (ROI) during the funding request process. Further, the typology provides a structured framework for decision-makers, helping to reduce risk and ensure investments are aligned with market demand.