Abstract : Our research work is focused on innovation and more specifically on the role of information in the product's conception process. Web 2.0 is quickly growing because contents are mostly generated by users and it represents now a wide and borderless source of information that can be used for innovation purposes. If Web improved information search techniques in technological and competitive surveys, we argue that Web 2.0 allow a better knowledge of customers needs and expectations. We propose in this thesis a qualitative and quantitative analysis of information generated or relayed by Web 2.0 users regarding their relevance into the innovation process. Our experiments deal with the study of Social Bookmarking and especially the Delicious.com Web application that allows users to memorize and share their bookmarks' lists. We would show that regarding to the increasing volume of information indexed by search engines, exploring this co-constructed directory offers a significant improvement in Web information retrieval. We also work on the social aspect of social bookmarking by proposing a way to sort participants in order to detect expert or influent users on a particular subject. Our study related to users comments on Amazon.com shows that customers are more and more involved into the innovation process of companies by providing them interesting and relevant comments about the product they bought. We therefore recommend meanings to extract then to analysis those customers feedbacks to improve the way that products are conceived. Finally, we define an innovation method that consists in setting up a permanent survey of Web 2.0 activities then to produce a document named " expectations notes ". that could be used for directing the further innovation processes.