Nice to Know You: Familiarity and Influence in Social Networks. Maurits Kaptein Tilburg University Department of Statistics maurits@mauritskaptein.com Clifford Nass Stanford University Communication Department nass@stanford.edu Petri Parvinen Aalto school of Economics Marketing Department petri.parvinen@aalto.fi Panos Markopoulos Eindhoven University of Technology Industrial design p.markopoulos@tue.nl Abstract Advertisers on Social Network Sites often use recommendations by others in a user’s networks to endorse products. While these familiar others are hypothesized to be more effective in influencing users than unfamiliar others, there is a catch: familiarity does not necessarily ensure similarity to the familiar person, a potential problem because the combination of familiarity and dissimilarity has been hypothesized to lead to lowered compliance. In an experiment (N = 44), we test people’s compliance to similar and dissimilar familiar others in an online environment: we show that in both cases, familiarity leads to increased compliance. The work highlights the importance of familiarity on influence and suggests that gaining familiarity even in situations of dissimilarity is effective. 1. Introduction Most users of Social Networking Sites (SNS) frequently encounter new connections: new friends are suggested by the SNS itself and new people are encountered in online spaces such as blogs or forums. Many of these new connections were previously unfamiliar to users. SNS services use these new connections and older ones to advertise products and services that are ostensibly recommended by or used by others in one’s social network. The general reasoning behind this compliance-gaining strategy is that people are inclined to listen to people they “like” [11], a property implicitly assumed to be true of one’s connections on a SNS. Cialdini [11] refers to “liking” as one of the “six weapons of influence” and reviews the overwhelming evidence that people are indeed inclined to listen to others that they like. This, and other similar work [24], has led to the enormous interest of marketers in social media campaigns and social advertising in which existing links in one’s social network are used to promote products or services. Thus, (product) recommendations or calls to action coming from a familiar other selected from one’s own social network are expected to be more effective then general appeals or appeals made by unfamiliar others. These so-called social recommendations are currently fairly commonplace [29,[30], and often direct measures of interpersonal similarity are at the core of these recommendations [3]. Liking—as promoted by Cialdini [11]—is not easily established, however, and being “linked” in a social network might be too loose a criterion to predict actual interpersonal liking up to a level that it indeed increases the persuasiveness of an appeal. New links, which are relatively unfamiliar to users, are likely to be less influential then older, familiar links. In this article, we determine whether increasing familiarity indeed leads to increased influence and thus a potential for more effective social recommendations. Closely connected to the literature on the effects of familiarity on people’s perception of each other and the influence people exert over each other is the literature on similarity. The Similarity Attraction Effect [10,18] states that people like people that are similar to them in background, occupation, personality, or a host of other traits and states. This would lead one to believe that both similarity as well as familiarity both increase liking. However, evidence to the contrary also exists: (see [19]) shows that liking might decrease when increasing familiarity with dissimilar others. Hence familiarity and similarity do not operate independently. In the remainder of this article, we first review the literature on both familiarity and similarity and their effects on compliance. This literature leads us to hypothesize that familiarity will have a positive effect on the influence people exert over each other-and thus compliance to requests made by “friends” in SNS’s only in cases that the connection is similar to oneself. Next, we describe an experiment to test whether compliance to a familiar other is indeed higher than to an unfamiliar other, and we separate both the similar and dissimilar case. The results show that gaining familiarity, even in situations of dissimilarity, is better than gaining nothing at all. However, in line with work by Norton et al. [19] we do find the effects of familiarity to be lower in cases of dissimilarity. We discuss the implications of this finding for the design of advertisements in SNS’s. Surprisingly we also observe another effect: similarity judgments towards