Most online services and applications on the Internet rely on digital traces resulting from choices made by their users, in particular, by means of rating-based recommendation systems. Therefore, it is crucial to understand how such traces affect individual and collective decision-making. We have conducted experiments to measure how groups of individuals interact with digital traces and determine how they could use these traces to cooperate and find the cells with the highest values in a table of hidden numbers. Our experiments and data-driven model show that digital traces spontaneously induce cooperation between individuals. However, the way individuals use the traces to deliver information to others and the reliability of that information largely depends on the degree of competition between individuals.