“Good” interface design is often grounded in principles around usability or accessibility (e.g. Nielson 2005, Marcus et al 1999, JISC 2009) but such guidance addresses inclusivity for all potential users of a website. In social media site design however the interface often breaches such principles but, in so doing, subtly defines the audience that a site wishes to attract. In some cases there is, perhaps, a logical basis for this: YouTube autoplays video clips as it assumes all users will be able to play a clip and are wanting to do so, in so doing however it discourages users with slower connections, those without speakers or headphones attached to their computers etc. Facebook has removed most of the buttons for posting updates – a single “Share” button is placed on most areas of the site whilst the home page design requires the user to hit “return” rather than click any button. These are not interface design choices aimed at meeting the needs of all users, they are instead choices to encourage savvy users to add comments and discourages all users from deciding not to post a comment they have started typing. Bryson et al (2006) highlights some of the more subtle factors in attracting selective audiences to online social sites where the design reveals default assumptions about potential users . Interview subjects in this study voiced discomfort with tacitly assumed gender, ethnic or sexual identities in online spaces for queer women and the subjects consequent sense of “un/belonging” to these spaces. Nakamura (2008) discusses the potentially limited ambitions of “racial performance” in digital spaces as well as the invasive issue of identity tourism. If design and algorithmic interventions around tacit assumptions and prejudice (e.g. straight dating ads delivered to non straight participants in a social networking site) can have negative effects even in specialised communities of acceptance how much greater the impact of assumptions in mainstream spaces where performance and embodiment of self is required as part of a learning process? Spaces such the Gorean community in Second Life (described by Bardzell and Odom 2008) provide a degree of tailored independence in mainstream social spaces but there is no panacea: all social spaces will, to an extent, include both positive and negative interventions that will affect different users differently.