Semi-reciprocal Transparency in Social Networks

I'm put off by the offers I see on various social networking sites for seeing "who's viewed my profile". While this feature may be attractive to some profile owners, it has a dampening effect on my use of these networks – I am less likely to look at someone's profile if I know that they might know that I've been looking. It reminds me of trade shows I've attended where my name badge has a bar code, and vendors in the booth ask me if it's OK if they scan my badge to record my contact info in their database (I always decline). If I'm interested in making further contact, I am perfectly capable of initiating that on my own. I don't want to be the target of an unsolicited followup along the lines of "thanks for stopping by our booth, let me tell you more about our product/service" … or, in the case of an online social networking site, "thanks for viewing my profile, let me tell you more about…". And some social networking services – or applications available on them – offer far more intrusive revelations of activities beyond simply viewing profiles.

The first site on which I encountered this offer of looking at who's been looking at my profile was LinkedIn, where this extra one-way transparency is one of the benefits of upgrading to a "professional" account (with price levels ranging from $25 to $500 per month). Even without the upgrade, I can find out some general features about some of the people who have viewed my profile, ranging from a "Recruiter at Nokia-Siemens Network" – a set of 32 possible LinkedIn users – to "Someone in the Information Technology and Services industry" – a set of 167,741 possible LinkedIn users (associated with "Accenture OR Andersen+Consulting"). In nearly every case, I have a pretty good idea about who the specific profile viewer is, regardless of the size of the set of potential users that is thinly masked by the more general characteristics provided at the non-professional (i.e., upgrade-free) level. I don't use LinkedIn much – and I'm not a recruiter or a salesperson – and my profile isn't viewed by many people; I can imagine that if I was a more active LinkedIn user, with something to sell (a job, company, product or service), I'd be interested in – and willing to pay for – a list of specific users who have viewed my profile.

As it is, though, the idea that any professional-level user of LinkedIn will know if and when I visit his or her profile makes me less willing to do so. When I'm notified that someone has invited me to "join my network on LinkedIn", or one of the people in my existing networks has asked a question, created a new group, updated their status, or engaged in some other notification-worthy activity, do I want them to know that I viewed their profile but then didn't take any further action (e.g., accept their invitation or answer their question)? Not necessarily. In many cases, I'd rather just ignore the invitation or notification … and, thereby, the user [profile].

I'm reminded of an early experimental media space application, the Bellcore Cruiser, designed as a virtual or online collaboration tool for distributed work teams that would replicate – or approximate – practices typical in shared physical workplaces. While many of the awareness and communication capabilities are rather commonplace today – via tools such as instant messaging, Skype and a variety of special purpose videoconferencing systems – Cruiser was extremely innovative when it was first deployed in the early 1990s [more information can be found in a seminal CHI 1992 paper, "Evaluating video as a technology for informal communication", by Robert Fish, et al.].

Cruiser enabled remote co-workers to connect with each other via computer workstations augmented with microphones, camers and special monitors in three different modalities:

• cruise: a real-time audio/video call initiated by a caller that required explicit consent from the callee before establishing the link
• autocruise: a series of 5 random audio/video calls initiated by the system – also requiring explicit acceptance – that was designed to simulate "corridor browsing" in a physical workplace
• glance: a 1-second video-only view into another person's remote office (if the person had not set their status to "private"), which would need to be followed up by a cruise invitation in order to initiate a full two-way, audio/video connection

All connections in Cruiser were reciprocal, so even the recipient (or target) of a glance was notified that someone was glancing at them. Not surprisingly – to me – there were 6 times more cruises than glances recorded during a 10-week study of the system, even though I suspect that in shared physical workplaces, brief glances far outnumber more engaged connections (e.g., casual conversations or more formal meetings) and most such glances go unnoticed. When I first learned about the system, I remember thinking that if I were a Cruiser user, I would not be inclined to "glance" at a colleague – if he / she would be notified – unless I had a definite communication goal in mind … in which case I would simply initiate a cruise.

Many related research prototypes and studies have subsequently been done (the ACM Digital Library lists 57 citations to the CHI 1992 paper), and there are many commercial systems in widespread use today that offer audio/video connections and capabilities for setting or assessing the status of other users. The point, for me (here), is that explicitly notifying a person that someone else is simply glancing at him or her – whether in real or virtual space – raises the social costs of glancing … and reduces the likelihood that such glancing will occur. When the nature of glancing is a real-time video feed of a person, it make sense for the privacy of the glancee to take precedence over the privacy of the glancer, but when the nature of glancing is viewing an explicitly constructed online profile of a person, it seems to me that the privacy of the glancer ought to have equal or greater precedence.

In effect, I'm suggesting that not all social media spaces ought to provide fully reciprocal transparency: not every action, e.g., glancing, is – or should be – noteworthy (or notification-worthy). There is something more disturbing about tying the level of reciprocity to the amount you're willing to pay … although I suppose this is consistent with other disturbing "facts" of life, such as the level of political contributions corresponding to the level of "access" lobbyists and special interest groups have to legislators (a recent example of which can be seen in the Sunlight Foundation's mapping of the Health Care Lobbyist Complex).

Returning to social media, it should be noted that LinkedIn does allow a user to specify three levels of what is revealed about them when they visit another person's profile: name and headline, anonymous profile characteristics (such as industry and title), and nothing. The default setting is the so-called "anonymous" characteristics, which as I mentioned above, are not as anonymous as they may seem. I don't know how many people have changed their default settings, but I suspect most users of most systems don't change most default settings, and thus many users – of LinkedIn and other social networking systems – are revealing more about their profile viewing activities than they may realize. I'm not sure if I would have even known about the revelation – or my ability to control it – if not for LinkedIn's aggressive marketing of their professional levels of membership (and the ubiquitously highlighted feature of finding out "who's viewed your profile").

LinkedIn revealed the "who's viewed your profile" feature in a blog post about two years ago, and based on comments, it is very popular among many people (mostly those who are the viewers of profile viewers). The capability for viewing who has viewed your profile on MySpace has also been available for at least two years. There are many services available for tracking profile viewers on MySpace – my favorite [name] is ProfileSnitch.com (a sample of which is shown below).

I'm not a MySpace user, so I don't know whether or how MySpace users can adjust profile settings to affect the activities that are tracked by such services. For all I know, this kind of tracking (or "snitching") may be seen as a positive feature by the snitchees on MySpace.

A more troubling level of revelation recently came to my attention through a friend who told me about the Compare People application on Facebook, that offers to help a Facebook user who has installed the application

Find out who stands where in various categories: cutest, sexiest, smartest and many more. Most importantly, find out where you fit in!

Apparently, 1,867,765 Facebook users have added this application, resulting in 65,738,628 people being involved in 3,478,630,212 comparisons. In a review by Rae Hoffman, "Compare People Facebook App Pulls a Bait and Switch?", she notes that the application originally seemed innocuous enough, allowing users to opt out of being compared, and to restrict disclosure of users' comparisons of (or votes for) their friends. However, the application introduced a "premium service" in which information about comparisons could be disclosed, e.g.,

Who should you date?
We crunched the numbers, checked the stars, and came up with your matches

Who are your true friends?
See who has the highest opinion of you

Who are your best references?
See who has the highest professional and academic opinion of you

Your wins and losses
A question by question recounting of exactly who you won and lost to

In a followup post on the Compare People Premium Service, Rae noted that even after uninstalling the application and blocking [further] comparisons, she was still being listed in rankings.

For what it's worth, this application is not particularly appealing to me – at either it's free or premium levels – as I have no interest in comparing my friends or being compared by my friends, and I would rather not know about about the results of any comparisons initiated by or involving my friends. Apparently, however, 34 of my friends have added the application, so for all I know I may be listed in the rankings … but I'm not willing to add the application in order to find out.

The reason I mention it, here, is that this seems like a particularly egregious example of unintended transparency within a social networking system. It's one thing to let someone know that I've viewed their profile, it's another thing to let them know that I've compared them – and how they've fared in the comparison(s). This may all be "acceptable" within the context of Facebook's terms of service, but I suspect that many of the Compare People users may not realize the level of transparency of their actions; despite over 1 million users, only 4,000 have reviewed it … though those that have offered feedback appear not to like it very much (its average popularity is only 1.7 out of 5).

And, although revealing LinkedIn profile viewers may be less intrusive than revealing Compare People ratings (and raters), I wonder how many other LinkedIn users are aware of the level of transparency … and/or are modifying their use of the system in view of its revelations … potentially reducing the value of the system to the very people who are paying for it.