Innovation contests often result in several hundred ideas generated. Raters have to process this huge amount of ideas that consist of attributes like idea descriptions and various types of feedback infor-mation with limited cognitive resources in order to separate good from bad ideas. It is not clear to what extent raters attend the available information during idea selection. In order to improve our un-derstanding of how to best support raters in idea selection, this study investigated the influence of vari-ations of the presentation mode (two versus four ideas per screen) on the attention paid to information on idea attributes using eye-tracking. We investigated attributes that refer to idea descriptions, feed-back about the content of ideas (creativity score, tags) and about the community comprising the idea-tors and the crowd (historical success of the ideator, likes). The results of our study show that with fewer alternatives per screen, feedback attributes received more attendance, while we found no signifi-cant difference for the processing of idea descriptions. These findings provide first insights into the information-processing behaviour of raters and can inform the design of selection platforms and theory building on the effects of feedback in idea selection.