- October
Posted By : Adrian Chan
A Social Interaction Design Primer

Current business conditions are unforgiving, and seem to be taking their toll on the social media industry, whether it’s in the mood on Sand Hill, in the decline of online advertising, or even in the prognosis for Web 2.0 at large. I penned a reflective piece yesterday in which I suggested that companies in this space do more to enhance the user experience, or suffer through the coming Darwinian fight for survival. While execs work to define their business models, I’d like to attempt to capture the basics of social media user experience. For the business all of us are in, at the end of the day, hangs on the participation of users.

Years ago Alan Cooper published a book that has become a classic among designers. Titled Inmates Are Running The Asylum, it set a framework for interaction design based on respecting the user and her needs, goals, and objectives. User-centric design is now the de facto approach taken by software, product, interaction, and all manner of designers, including social media designers. In fact our industry, perhaps more than any other, relies on delivering compelling user experiences for its success. Users are our business model, and failure to engage users not only kills participation, but all other aspects of a social media company’s business also.

As a self-proclaimed social media expert, I’m often asked by companies how to “get users to do” more of this or less of that. Clients understandably identify with their product, and with what they designed it for. But in social media, users do what they think a site or service is for, and not necessarily what it’s designers intended. So I begin an engagement by asking clients to view their product from the user’s perspective. Many small companies do not have a user experience designer on staff, and rely heavily on best practices to steer feature and interface design.

But user experience matters in social media are more complicated than in non-social software. For example, the conventional user-centric view starts with user needs and goals. In social media these are not necessarily rational and objective. They can be much more psychological, and social, for example. Furthermore, the interactions that users have are not just with the software application — they are with other users (through the software). The UI is not an interface to discrete actions and transactions (such as your online banking site); it is a social interface, and through it users feel like they are interacting with friends and audiences.

This complicates matters somewhat for the standard interaction design approach. If the task of conventional software is to provide successful interactions, to inform the user that his actions worked, then what of social media? Communication is by definition an open-ended transaction, not a discrete one. Take the example of a dating site: one user pings another, by messaging or gesture, and hopes to hear back. Does the software designer want to provide a status message about the recipient’s interest? “Your message was received but she’s thinking about it. Please be patient.”? Likely not. In fact, the dating site wants to keep its users on the hook for as long as possible. Ambiguity is in its interest — not clarity and transparency.

We can go one further in distinguishing social interaction design (as I call it, or SxD) from standard user experience and interaction design. For in social software, failure works. Take twitter for example, which is not used for SMS-Web messaging as originally intended, and which hooks many users because in social communication and interaction terms, it is kind of upside-down and in reverse. Users don’t choose who they are talking to. When they post, their tweet appears in “thread” that is a false representation because their post appears next to the tweets of those they follow, not those who follow them. And there is an asymmetry between posting and reading such that users are required to declare their presence. Where a chat room or IM application is designed to capture users’ presence, twitter does the reverse. Users have to declare their presence and attention by using (@ or direct messages): “@username, Nice post!”

Social interaction design works by respecting the psychological and social, the ambiguity not the clarity, the unintended not the intended. The best a designer can do is set up a social architecture that structures and organizes participation well enough that users know what’s going on, and therefore what to do. Social interaction designers start not from user needs but from user interests.

The bottom line for any social media company is know your users. Here again, social interaction design differs from non-social design. There is not just one user. There are not even several “personas.” Instead, users differ by their communication and interaction styles, their ways of being social, their understanding of what they are doing and of what others are doing. For simplicity’s sake, I segment users according to three types of interest: Self Interest, Other Interest, and Relational Interest. This comes from contemporary sociology and psychology, and goes roughly like this:

Self-interested users act from a position of Self
Other-interested users react to an Other (user)
Relationally-interested users interact through social activity

To provide a few examples, there are Facebook users whose activity centers on their own profile, which is a representation of their Self and an extension of it (into the mediated social world that is Facebook). These users may not even visit their friends’ profiles. They interact around their own status updates, wall posts, profile page elements, and so on. Then there are Facebook users who spend more time browsing their friends’ profiles, posting to their walls, reading their friends’ updates. They do not begin communication on their own pages, talking about themselves, but begin by responding to a friend’s post or update. There are then those in the third group, the socializers if you will, who play the numerous Facebook social apps. Drawn to social activity, they go where the action is.

We can see this on twitter, also. Some users post to their audiences about themelves. Some, finding this weird, read first and are inclined to respond. And others get into rounds of conversation, often including their friends by @naming them in their posts.

These are rough distinctions, and I will be the first to admit that there is no research as of yet into their viability. But they correspond to similar distinctions made by psychologists and sociologists, and after a year of thinking about core social interaction design principles, I have yet to come up with anything better.

And as it turns out, a similar three-part approach works for the user interface, which I call the social interface. Standard UI theory comes up short here also, because the UI is not only a representation of software features and functions, but is a medium through which users engage with other users. Imagine that you are sitting across from a friend. There is a screen between the two of you. Now in social media, that screen has three modes. It may be a mirror, and you see yourself reflected. Or it is a surface, and you see what your friend has posted on it. Or it is a window, and you can talk through it with your friend.

Design requirements are different for each mode. In the mirror mode, the interface should present an engaging and compelling reflection. In the surface mode, it should organize and structure content and navigation. And in the window mode, it should become transparent and unobtrusive.

Now these modes of the interface correspond nicely to our three user types. Self-interested users may engage in their projection and expression of themselves (mirror); Other-interested users may respond and talk to others (window). And Relationally-interested users may go where the action is (surface).

Marshall McLuhan, the patron saint of media theory, claimed that every new medium uses an old medium as its content. Social media use mass media for much of their content, and moreso for how to organize and lay out that content. We use print, web 1.0, software application UIs, as well as television, cameras, and radio. But we cannot understand how social media are used if we do not first understand what interests our users, in being on social media, in using them to connect and relate, and in reading what users have left behind.

This is still a very young medium, and there is much yet to come by way of innovating time-based conversation and self-presentation tools (see my brief on Swurl and a piece on designing for lifestreaming). I expect more innovation of the presentation layer, by means of Flash, for example. And of course there is mobile, which we are only beginning to mine for new and compelling experienced.

But if you are in the social media space, and feel that your product captures a good 80% of current uses and best practices (in terms of features and design), you might now invest the next 20% in cementing the user experience. I will post again soon with specific ideas for identifying your core user types and improving user experiences to raise participation levels.

For those of you who are interested in more, see:
Slideshare presenations on What Is Social Interaction Design, User Psychology, and User Competencies (or go here to download originals)
White papers


  • personacreation.com

    October 31, 2008 at 12:10 pm

    Hi Adrian,

    I’ve been following your “user needs” vs. “user interests” distinction in various posts and I agree that it is an excellent, clarifying distinction about social media users. Social media designers will benefit from the social and psychological perspective you so clearly lay out here and elsewhere.

    I disagree, however, with your assertion that personas cannot capture users’ social media interests. My experience is that personas and scenarios can help communicate interests. Why do you not see them as appropriate tools? What other tools would you recommend to convey users’ social and psychological interests to designers and programmers?

    Great post– Cheers!

  • Angela,
    Wait a second, don’t I know you? 😉 (how’s it going?!)

    This may be a matter of semantics, or of my not being clear. Personas are, as far as I get it, heuristics — so they are incredibly useful in the design process for communicating *about* the personified “user.” I’ve got nothing against that.

    What i’m after tho is the inner user experience, which is a matter not of “observing” the user, or “describing” the user — those serve the purpose of communicating *about* the user and are helpful, as personas are, in reports about social media users. Forrester, etc do this very well — there are probably 4 or 5 models of user types, social media market segmentation, and so on that help communicate about social media.

    The inner experience requires not a descriptive but an explanatory theory, if not predictive theory. And as we both know, that’s a hermeneutical challenge. I’m using psychology, against better instincts, because it is a familiar language. I want to know what psychological hooks work online. What people think they are doing (which can be very different from what others think they are seeing them do.) I want a language for phenomena like “I see myself being seen” — stuff that I think motivates users *even if it is subconscious.* I need a framework for how to understand how some social media really engage and compel users because they offer such great presentations of the user and the social. And the list goes on and on.

    I’m using modernism a la Giddens, and Goffman for interaction, Berne et al for transactional analysis, Habermas in part, and so on…

    Writ large, I think users have a private encounter with the social (online). That encounter is where the gold is (literally and figuratively). Personalities are thus, for me, better than personas.


  • Yeah it very difficult to make perfect UI designs

Leave a Reply