Shared mobility services are announced as a game changer in transportation and a promising solution to reduce congestion and improve performance of urban mobility. However, only few studies aim at identifying if it really exists a significant demand for such services. This paper propose a first attempt toward this objective by finding twin travelers into ridesourcing trips data. Using the Didi Chuxing's dataset, a general methodology is defined to assess the similarity between trips and to cluster analogous travelers. The study reveals, among others, that at least 18% of the trips can be paired with introducing tolerable delays.