The paper analyses product knowledge as a key factor for classifying and selecting the more innovative users within virtual brand communities. Based on a multi-method approach, the work is applied to the Ducati Hypermotard blog, with the aim to study the recurrent contents and subjects observed in users’ messages and their distribution in relation to the members’ ability to support product innovation. The existence of a link between user’s product knowledge, text contents, and user innovativeness is confirmed. Finally, a model for the automatic text-based classification of messages and users is proposed.
Keywords: Product innovation, virtual brand community, lead user, product knowledge, content analysis; text-mining