In industrial settings, brainstorming is seen as an effective technique for creativity in innovation processes. However, bulk of research on brainstorming is based on an oversimplified view of the creativity process. Participants are seen as idea generators and the process aims at maximizing the quantity of ideas produced, and the evaluation occurs post-process based on some originality and feasibility criteria. Design theories can help enrich this simplistic process model. The present study reports an experimental investigation of creativity process within the context of real-life design ideation task. Results lead to the rejection of the classical 'quantity breeds quality' hypothesis. Rather, we observe that successful groups are the ones who produce a few original propositions that hold great value for users while looking for ways to make those propositions feasible.