Hello Stranger! A Study of Introductory Communication Structure and Social Match Success Daphne R. Raban University of Haifa, Israel draban@gsb.haifa.ac.il Stephen T. Ricken, Sukeshini A., Grandhi., Nathaniel Laws and Quentin Jones New Jersey Institute of Technology, USA sr82, sg49, ndl4, qjones @NJIT.EDU Abstract Social matching systems recommend people to people. One of the key challenges such systems face is how to build introduction mechanisms that minimize privacy concerns, while maximizing the chance of successful introductions. Another challenge is to design systems that are suitable for mobile devices. We explored the relationship between introduction success and four communication interfaces: 1) a simple chat interface, 2) a restrictive profile exchange interface, 3) a hybrid of the two with temporary profile information, and 4) a hybrid of the two with full profile information. Our findings show that the restrictive profile, which is suitable for mobile devices, supported a comparable level of willingness-to-meet as did the other conditions but it did so in a considerably shorter time. Users preferred the UI of the fourth condition which suggests a future development direction for introduction system user interfaces. We discuss the implications of this finding for design of mobile social-matching applications 1. Introduction Social matching systems are designed to recommend potential social interaction partners to individuals [12]. Such systems can be comprised of several components including: profile management user interfaces (UI); user profiling systems; affinity modeling systems; match management systems; match recommendation UIs; and introduction UIs. Profile management UIs and user profiling systems utilize explicit user profile data (e.g., demographic, personal attribute, and preference information). Affinity modeling systems on the other hand utilize profile and/or implicit user behavior (e.g., data-mined attributes of friends) to provide inputs to match management systems which make social match recommendations that are conveyed to the user via match recommendation UIs. Users can interact with potential matches using an introduction UI, which will ideally ensure safe and anonymous introductions, where appropriate, especially in next generation systems that leverage the ability to locate or reach a person via mobile phone. Designers of social matching systems have thus far emphasized the design and development of simple user profiling and matching systems. For example, friend-of-friend systems provide users access to various parts of their friends profiles, thus helping them visualize their immediate social network [4]. In existing social matching systems such as eHarmony and OKCupid, introductions between matched users are relatively unmediated. Both parties receive a match alert but have no dedicated, direct UI for an initial interaction. Instead, users utilize a combination of synchronous and asynchronous private messages, profile comments, and friend requests to move the introduction forward. Introduction UIs in such systems are generally simplistic [3], [7]. Initial interaction via indirect means is slower than real-time interaction, and carries the risk of participants losing interest, especially if the individuals interacting are complete strangers with no social links between them. Our interest in computer-mediated introductions stems from a larger research initiative to improve the social climate of a mid-sized urban university whose students have expressed discontent with the quality and quantity of social opportunities on campus. To this end, we are actively designing and developing CampusMesh, a real-time social matching and networking application to run on smart mobile devices. CampusMesh will support social introductions and ongoing relationships between members of the university community through social match alerts and reminders (e.g., "there is a user nearby that you should meet" "Fred, whom you met before and put on your friend list, is in your vicinity", "ask Joe to show you his new camera, while he is adjacent to your current location"). To ensure user privacy, CampusMesh will provide a mobile introduction system that gives users control over the revelation of their personal information (see Figure 1). One of the aims of our research is to gauge how simple and open such Proceedings of the 42nd Hawaii International Conference on System Sciences - 2009 1 978-0-7695-3450-3/09 $25.00 © 2009 IEEE