Crowdsourcing as a mechanism of open innovation is a popular way for organizations to solicit ideas from external agents. Our research focuses on the relationship between examples in problem statements provided to a crowd and the subsequent number of ideas submitted by the crowd. We investigate the quantity, as well as the diversity, of these examples. Using generalized regression to analyze data from 122 problem briefs posed to a crowd on an intermediary platform between 2018 and 2021, we find that the relationship between the number of examples given to a crowd and the number of ideas received from the crowd is an inverted U-shaped one. Further, we find that providing a diverse set of good and bad idea examples increases the number of new ideas received from a crowd. Our research extends the nascent crowdsourcing research on using problem statements to guide the crowd’s response to innovation problems.