As an analysis tool for systems' development, participatory design (PD) is based on the claim that users should be regarded as experts when they enter the design process. However, experience proved that this "expertise" happens in different degrees and that, in a large-scale system, it is hard to find the real experts among a crowd of thousands of users. PD demands a constant interaction between user and designer, and this could be the source of quality feedback to improve the interface. Nevertheless, how can users send good feedback when dealing with a transnational system designed for thousands of clients in different countries? This scenario will be more common in a computerized world in which stand-alone programs are being replaced by Web applications that can be accessed worldwide through a browser. This represents a severe threat to PD's intentions, and in order to survive, this method will have to merge with larger scope research techniques and then-through random sampling that represents the diverse, international body of users-obtain user feedback. Just like in mass communication models, when it is not possible to listen to every spectator and the "letters to the editor" get lost in a big box, quality will yield to quantity in user feedback.