Session: It's a Big Web! CHI 2012, May 5–10, 2012, Austin, Texas, USA Talking in Circles: Selective Sharing in Google+ Sanjay Kairam1, Michael J. Brzozowski2, David Huffaker2, Ed H. Chi2 1Computer Science Department 2Google Inc. Stanford University Mountain View, CA [email protected] {zozo, huffaker, edchi}@google.com ABSTRACT If users are aware of the consequences of ‘over-sharing’, Online social networks have become indispensable tools for then why continue to share so much? This question information sharing, but existing ‘all-or-nothing’ models constitutes what Barnes has called the ‘privacy paradox’ for sharing have made it difficult for users to target [4]. In this paper, we study individuals who are actively information to specific parts of their networks. In this sharing while successfully handling privacy concerns. We paper, we study Google+, which enables users to selectively observe their behavior to identify strategies for improving share content with specific ‘Circles’ of people. Through a sharing precision through selective sharing. combination of log analysis with surveys and interviews, we investigate how active users organize and select We base our analyses on usage data from the field trial of audiences for shared content. We find that these users Google+, a social networking service introduced in 2011. frequently engaged in selective sharing, creating circles to Through the incorporation of ‘Circles’, this service is manage content across particular life facets, ties of varying designed around selective sharing and consumption of strength, and interest-based groups. Motivations to share information. In the following section, we provide some spanned personal and informational reasons, and users brief theoretical background on needs for selective sharing frequently weighed ‘limiting’ factors (e.g. privacy, and prior research on challenges for users sharing using relevance, and social norms) against the desire to reach a existing social technologies. We then provide a brief large audience. Our work identifies implications for the introduction to how these needs are addressed in Google+. design of selective sharing mechanisms in social networks. BACKGROUND Author Keywords In applying Altman’s theory of privacy boundary regulation Social media; group-based access controls; social networks [3] to the online world, Palen and Dourish argue that the disclosure of information is a necessary consequence of ACM Classification Keywords participating in a social world [22]. In the offline world, H.5.3 Group and Organization Interfaces some information (such as presence) may be freely INTRODUCTION disclosed to the general public, while other information Social media, from blogs to forums to online social (such as personal secrets) may only be revealed to a networks (OSNs), have become a major venue for social privileged few. Using Goffman’s dramaturgical metaphor interaction and communication. In the United States, over for interpersonal communication [9], social interaction 65% of adult Internet users are currently active on social provides a stage and an audience for whom we tailor self- networking tools [17]. This phenomenon is global; at the representations, disclosing what we see fit. start of 2012, Facebook, the largest OSN, claimed over 800 Prior work has revealed the existence of ‘facets’ [8] or million active users, 75% of whom were outside the US [7]. ‘modes’ [20], representing the various audiences for whom A key affordance of these tools is interaction with we must fluidly adapt our performances over the course of audiences to which users may not have had contact offline; our day. Generally, in offline settings, information often these interactions accompany disclosures of personal disclosed is ephemeral and the audience evident, enabling a information. Research on social media over the past decade tight loop between what we share and with whom we share has shown that users are revealing a startling amount of it. As Palen and Dourish identify, moving interactions personal information through blogs [14, 25], profiles [10], online means that content shared may persist beyond the and status updates [19]. In situations where readership may scope of a conversation and reach unintended audiences. not be clearly delineated, users often underestimate the size Many existing tools for online communication are designed of their potential audience [1], in many cases leading to to enable ‘public’ sharing to wide groups. Rather than tension [8, 23], regret [26], or tangible negative focusing on the natural ‘facets’ of a user’s life, many social consequences in one’s personal or professional life [25]. networks, for instance, grant default access to content to a user’s ‘followers’ or ‘friends’, restricting those who fall Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are outside this boundary. While this model may be simple for not made or distributed for profit or commercial advantage and that copies users to understand, it ignores the ‘natural’ groupings which bear this notice and the full citation on the first page. To copy otherwise, may exist within these networks [16], making selective or republish, to post on servers or to redistribute to lists, requires prior control over content difficult and prone to error [24]. specific permission and/or a fee. CHI’12, May 5–10, 2012, Austin, Texas, USA. Copyright 2012 ACM 978-1-4503-1015-4/12/05...$10.00. 1065 Session: It's a Big Web! CHI 2012, May 5–10, 2012, Austin, Texas, USA Difficulty in disclosing information selectively to various users separate their networks for the purposes of selective life facets can lead to what Marwick and boyd have called sharing and consumption. ‘context collapse’ [18]. Studies of OSN use in the work- In the subsequent section, we report findings from a place reveal that this leads to very real tensions for users of qualitative study of users sharing comfortably on Google+. such systems [2, 5, 23]. As a result, users may engage in We analyze the behavior of these ‘expert’ users through drastic behaviors to avoid future privacy violations, such as data collected via surveys and interviews in order to self-censoring or ceasing sharing altogether [24, 26]. understand how they use the product to successfully tailor Tools such as email fall on the other side of this spectrum, self-presentations for various parts of their network through easily enabling ‘targeted’ sharing of specific items to selective information sharing. We focus specifically on how individuals. By essentially requiring access controls to be they organize their contacts into Circles, what factors set for each item shared, such systems provide a high motivate them to share, and contextual considerations for degree of control over access to content. While prior work choosing an audience for a particular piece of content. has shown that particular sub-groups of email contacts may QUANTITATIVE LOG ANALYSIS be common across many users [20], the creation and In this section, we conduct analysis using usage logs from maintenance of persistent groups for access in these the field trial of Google+ to examine overall user sharing systems can often be cumbersome or difficult. patterns and characterize how users organize their networks Introducing Google+ based on names assigned to Circles they have created. Google+, a social networking service launched into field High-Level Sharing Patterns trial by Google in 2011, was designed to address some of To understand the extent to which users were engaging in these problems. To aid users in selectively disclosing selective sharing with Circles, as opposed to public or information to common sub-groups of their network, targeted sharing, we first examined logs of per-user sharing Google+ introduced ‘Circles’, an intuitive mechanism for choices aggregated over the week of July 20-26, 2011. organizing contacts. Similar to ‘lists’ or ‘groups’ available in networks such as Facebook or Twitter, Circles are user- Data created groupings of contacts which may be overlapping or User IDs were hashed and sampled randomly from the hierarchical, allowing users a great deal of flexibility in population of users who had shared at least once during the organizing their networks. Each time a user shares a piece data collection period, providing a representative sample of of content, he or she makes a contextual decision about the 100K active users. To preserve user privacy, we did not audience with whom to share it; content can be shared conduct any analysis concerning specific Circles, but rather publicly or shared selectively to one or more circles. Users restricted our analysis to public (posts shared using the can also share to ‘friends of friends’ by selecting an ‘Public’ setting), selective (posts share to one or more ‘Extended Circles’ option, though this feature is not Circles), and targeted (posts shared using a ‘+Mention’) analyzed in detail in the present work. sharing. In order to compare our results to those found in the later qualitative study, we focused on users sharing via In addition, specific individuals can be referenced through the desktop web application, resulting in a data set of inclusion of a ‘+Mention’. Mentioning users individually in 64,005 users who shared content during this week. this way can be used to target content directly at those users. Use of a ‘+Mention’ when sharing to a wider As shown below in Figure 1, the number of items shared audience can be a means of notifying the mentioned user weekly per user roughly follows a power-law distribution. and directing their attention to the post, similar to a Using Clauset et al.’s approach [6], we estimate the ‘@mention’ on Twitter. In addition to providing a simple parameter α as 2.578, meaning that the distribution matches mechanism for selective sharing, Circles also support those of a variety of man-made and naturally occurring selective consumption. Users can filter their incoming phenomena, giving us confidence that out data was stream in order to see content being shared only by representative of the wider population members of a specific Circle.