Previous studies suggest, contrary to our naive understanding that ideas for innovation are generated by supply-side such as product developers, that the ideas can be actually generated by consumers, especially the users called “early adopters” who keep some distance from product developers in product adoption and knowledge. This research tested, through two experiments for idea generation, which was more important to idea generation for innovation; information, i.e. preceding ideas about new usage of a product, or individual cognitive feature, i.e. innovator or early adopter. One experiment was conducted with general consumers, taking as an example idea generation for new products and services related to information technology, and the other was done with R&D members in a real company. The results suggest that information, i.e. preceding ideas about new usage of a product, was important to generation of creative ideas for innovation and, at the same time, that it was early adopters rather than innovators who could make effective use of the information in the idea generation: This is considered to call for a rethink of “sticky information hypothesis” by von Hippel (1994), which claims the dominant role of information in innovation. This research also suggests, through an additional experiment for information retrieval, that tendencies of innovators to avoid communication with people other than innovators and to think much of function and spec that a product has may degrade their performance in creative idea generation.