As technology advances, experts around the world have been able to collaborate through crowdsourcing platforms. The organization then began to present users with projects of more serious difficulty for crowdsourcing software development. Crowdsourcing projects that require higher quality results depend not only on the products of users but also on the user's participation, ability, and user initiatives. This study explores the factors that enable crowdsourcing software development around a crowdsourcing platform that deals with projects that implement algorithms or code. Furthermore, the success of crowdsourcing competition is measured by analyzing user engagement and initiatives. For this study, 415 competition data from the Kaggle platform were collected and analyzed as a regression model. As a result, the increased competition duration and the influx of high-rated users encouraged crowdsourcing projects by increasing user participation and initiatives. It also revealed that the reward strengthened user initiatives but reduced user participation and that the parallelism of the competition reduced user code-sharing but increased user discussion.