Adrian Chan Archive

I am aware of the irony of posting about the the buzz on Google Buzz this week. But there’s no other way to contribute than to heap yet more on the pile.

I’ll skip over the many good points that have been raised this week within buzz and alongside it. If you are reading this, you have probably read them.

I want simply to make a few observations about the Buzz user experience, some of which are simply unavoidable, and many of which belong to the “conversation” space in general.

Talk is a difficult thing to facilitate using social tools and services. This is because in addition to the content itself, there are contributors, readers, relationships, audiences, social scenes, and public. And talk is a form of social action. A statement alone is communication. A response is action that communicates. Many kinds of actions involving talk exist (greetings to wedding vows), involving varying degrees of sincerity, expectation, commitment, trust, and so on.

Continue reading “Breaking down the Gbuzz” »

The recent skittles twitter campaign used a feature in limited testing at twitter. It’s called integrated search, or real-time search. You could see it at work Sunday & Monday on the skittles.com homepage, or in the picture here taken from a deck by Fred Wilson and covered recently on Cnet. New search results are posted to the top of a search results page in real-time, effectively transforming search into conversation.

It has the effect of aggregating conversation within twitter, by threading posts around the search phrase or keyword. This strikes me as a potential game-changer for twitter, for several reasons.

We currently hold "conversations" on twitter with followers. We have to search to find non-followers around topics. But there are barriers to bringing them into the conversation. Results are past results, and we have to follow/be followed back before conversation becomes possible.

So conversations tend to happen between people who follow each other. If they are topical, they tend not to mention the topic. And this makes them less easy to find in search. Twitter addressed this recently. If there has been conversation between users (using @replies), it is now visible with the "show conversation" link.

But there are limitations to the usefulness of the "show conversation" implementation:

• to engage in that conversation would require that we follow and are followed back
• "conversations" are often off topic, or get off topic quickly
• the focus is on the people @replying to each other, not on keywords

"Show conversations" doesn’t really capture conversations, but captures an exchange between users who have @replied each other. Only the first tweet in the exchange has to contain the search keyword.

Twitter certainly realizes that it needs to searchable. But it also realizes that search results are limited to our use of search words and phrases. And limited by the fact that we have only 140 characters at our disposal. If twitter went after conversationality, it could do so only by aggregating the conversation around an exchange between users who follow one another — not around topics.

The following-follower model that has made twitter so incredibly viral has been a constraint on conversations. Each of us has only a small window through with to see what a small number of people are talking about. And only a limited means of capturing and sustaining conversation with people around a topic.

The theoretical description of this problem is this: tweets are only loosely coupled. They are loosely coupled between users, and loosely coupled by topic:

• Tweets are not coupled to each other unless they include an @reply or D message. The latter doesn’t count for public conversations. @replies only count if our account settings are to set generously (there are three settings).
• Tweets tend not to sustain topics because they must be so short, because we tend to initiate and then drop and change what we tweet about, and because the twitterverse serves the purpose of talking about and creating news. In news, we are more likely to pass something along than to engage in discussion.

Twitter was designed in such a way to prohibit conversations. Not intentionally, of course, but symptomatically. Conversations require a kind of coupling between statements and responses, and people in conversation, that twitter makes incredibly hard to achieve.

Alfred Hitchcock used to say that he never made a "Whodunnit" movie. His movies were "For whom was it done?" In fact a lot of his movies begin with the crime. In some, the victim of the crime turns out to be the criminal himself.

In all of Hitchcock’s films, we the audience witness some aspect of the crime. And because Hitchcock was a master of camerawork, and used his camera to let the audience in as a witness, we’re usually in on something that one or more characters don’t know. Jimmy Stewart’s neighbor leaving his apartment in Rear Window, as Jimmy reaches for something he has dropped. The killer’s shadow on the shower curtain in Psycho. A vertiginous zoom in on Kim Novack’s curled hair — an audience reveal that winds up the plot’s second, and formal spiral in the mystery Vertigo.

Hitchcock’s films were as riveting as they were not only for his splendid choices in casting his lead actress, but for his singular talent at subordinating characters to formal puzzles and logics. He is credited as being the first to involve the audience in solving, or "creating," the film. He was notorious, too, for glossing over his actors’ needs and for attending instead to the visual narration of the particular puzzle at hand. It mattered more to him the direction in which his actors were looking than capturing their motivation.

Hitchcock knew that a mystery thriller could become endlessly suspenseful if actions were not simply as they appeared, but were instead motivated by another, for another, or on behalf of another. This allowed him to continuously shift the "guilt" and "suspicion" from character to character. We in the audience had the job of figuring out who was who, and who was who to whom.

The solution to the puzzle, and to the crime, always came out when relationships among the characters could be resolved.

Action is more interesting when it is a matter of interpersonal motive and relationship, rather than the accomplishment of the task itself completed by the action. It’s a pity there are few good imitators of Hitchcock. (Although there are some; and social films like Crash, Amor es Perros, Red, White, Blue, Babel, and others in which relationships form out of coincidence and chance in a way capture the state of social fragmentation endemic to contemporary society.)

We in social media can learn from Hitchcock. We can learn to ask not "What did the user do" but "For whom was it done." Was it done for his/her own self-image and repute? Was it done for the attention of another? To solicit reciprocal interest of another? To gain notice by a group, club, or circle of peers? To obtain status in front of an audience, or to receive the validation of peers?

I wonder what kinds of social media Hitchcock would design, if he were in our industry. How might he use his "camera" to show the audience something that was off screen to the actors involved in a situation or social interaction. What kinds of relationships he might put people in if he were designing social games. And how he might reveal clues and thread his plot points. Whether the audience might be involved in passing that thread through the warp and woof of a networked social fabric. And how interesting and engaging some of his creations would be, designed not around Who said something but For whom was it said?

Adrian Chan is a social media experience expert and analyst. You can follow him on twitter at gravity7.

In client work recently I have come up against the the importance — and difficulty — of satisfying multiple user positions and experiences. Social media work because the author and the reader are satisfied. Sure, social MEdia need to be satisfying to me, but if they are to scale and succeed, the social system needs to reward readers and recipients, too.

This has brought up a few interesting principles of late. I want to share them just because I find them interesting.

The first is that there’s an asymmetry between the interest which motivates the user who "acts" (creates, posts, etc) and the user who responds. Not all systems are built around a coupled statement-response model of communication, of course.

But there’s an intrinsic interest in response for a lot of users and use cases in twitter and other conversational tools. If I ask a question on twitter, I am motivated by my question, which is something I want an answer to now. The person who is asked (who sees) the question has no interest at all. His or her interest can be piqued and aroused — but is not the same as mine.

The act of answering may more likely be the motivation, and not the content of the question. Question and answer systems are difficult because they involve satisfying two users, the asker and the answerer. If these user experiences are satisfied in real-time, then the interaction itself handles the experience. If they are satisfied out of synch, then each user has to produce his/her own interest: one in the act of asking, one in the act of answering.

This asymmetry extends to other aspects of communication in social media. Take, for example the case of sharing.

Because users are different, and have different personalities:

• Some who share do so because they want to share with (someone).
• Others share because they want to show to (others).
• Some share to exchange for (something).

These are different experiences and are met with different technologies or have different technical solutions.

For example, the user who shares with someone probably posts the photo in order to send it along. Sharing is the act; The photo is the symbol. This user wouldn’t post this particular photo if it weren’t for the person or people s/he was thinking of.

But take a person who finds something online, and book marks it because it is interesting, but has no person in mind to share it with. Later, this person decides either that the thing is interesting enough that others would find it interesting, and shares it. Or s/he thinks of a person who might enjoy it also, and shares it. Sharing in this case has come afterwards. It is a second act, it adds value to this person’s user experience, but wasn’t the original motive or interest.

The user in our first case, on the other hand, wanted to communicate from the beginning. Communicating was the primary act, and was the motive and interest.

These are just two simple examples of how the activity involved in sharing stuff online can be broken down into two acts: one of saving the thing; one of sharing the thing. And that these are different, depending on whether they are governed by the act of saving or the act of communicating.

In the first case the action carries the content. Communication leads.in the second the content precedes the act. Communication follows.

These distinctions may seem trivial but they’re not. They have significant implications for:

• how the system scales
• who finds it useful (and who finds it a waste of time)
• what content is produced as a leave behind
• how personal or public it is in tone
• how easily it can be organized and structured, and so on

Social problems can have only social solutions.

Adrian Chan is a social media experience expert and analyst. You can follow him on twitter at gravity7.

This post is a follow up to the first law. There are two more coming.

The second law of social interaction design is that the functionality of social media is contingent on social practices that use them. Notice that there’s a double contingency there. Social media functionality is contingent on use by users; use by users is contingent on the technologies required for use to be possible. This double contingency means that social tools are inseparable from the users, and the use by users inseparable from the tools. It is not a matter of which comes first, or of how technologies structure interactions, for interactions shape technologies. And as all of us in the "beta" community know, social media design is iterative because it constantly observes its own use.

What this means is that social practices absorb and assume the burden of structuring interaction, and of organizing and coordinating activity where tools make that a possibility. The less that is "designed," the more that is handled by people. Chat, and in some ways Twitter, are among the least designed of social media tools. It is conversation, talk, and communication that organize them.

Asking of a social media tool not what it does, but what it is capable of, is an empowering shift of attitude. It gets closer to the grail of social media design, which is, in the context of mediated interactions, What are people capable of? In the real world, audiences are assembled for all manner of reasons and purposes, and their behaviors are as diverse as those of mobs, gangs, queues, gatherings, marches, protests, and so on.

Audience sizes do not scale linearly, groups becoming audiences becoming crowds becoming masses. Purposes may organize the disposition, expectations, the duration, activity, and even spatial orientation of real-world assemblies. In all cases, one can ask What are the people capable of, and get a different answer. They are capable of waiting, of rushing, of dispersing, coalescing, rampaging, blocking, in silence or with sound and fury. By asking What are the people capable of, we recognize that there is some kind of order and organization, in space and time, and often without the barest of architecture or other bounding.

Because the communities built around social media are for the most part asynchronous, they are framed in time, over time, as much as they are by tools themselves. Audiences exist because users return. For every established online community or user base there are many individual users making a regular practice or habit of use.

This gives us our first corollary, which is that social practices emerge out of aggregate individual user practices. There is no "one" social practice, but several, and experienced by each user uniquely. On tools like Twitter, for example, these practices include various kinds of talk and messaging. They include announcing one’s location, one’s feelings, one’s activities, and one’s plans. Practices that require two or more users include brief rounds (or conversations), replies, retweets, and a variety of types of commenting.

Commenting, in contrast with conversation, is not reciprocal. On Twitter we find people referencing a user who has followed them; referencing a user who has mentioned him/her; referencing individuals, auto-replies, and topics by referencing individuals. These comments may solicit responses, but do not directly respond to those in question. Other social practices include fights, marketing offers, event invitations, application invites, and more. Combined, they represent the growing set of practices to which Twitter is suited, from a core set on out to those at the margins.

We have seen already that the use value of social media is not simply one of utility. Given that each user brings his or her own set of values, and ways of valuing experiences on social media, we have the corollary that social practices need not have any utility. The users engaged in an observable practice may each have their own reasons, and thus reasons for use, hence making it impossible to ascribe one use value to the practice.

This may contradict those who believe that there are in fact collective uses of social media. But if those exist, they exist only in theory — they don’t correspond to user experiences and thus can only be argued on the basis of some other system of measure. Are social media democratizing, or are they subject to crowd psychology? Are they informing, or are they given to gossip and opinion? These are perspectives, beliefs, debates, and controversies, and while of course interesting, lay no claim to user centric design principles or insights. The social practices seen on social media are just that, and any "business" or application of social practices to market opportunities is epiphenomenal to the forms of talk it is based on.

This leads us to a third corollary, that social practices are emergent. In chaos theory terms, emergent phenomena are auto-poetic. That is, they "write" themselves. They are the product of forces that, in observing themselves, reproduce themselves without any particular genetic design, external guidance, or intrinsic goal. To the social media designer, this creates a problem. For the best a designer can do is supply architecture that, when populated, is most likely to result in anticipated or desired social practices.

There are no direct steering mechanisms available in social interaction design; only the educated and informed use of features, design elements, and other design choices intended to enable and extend commonly occurring practices. Here, again, the designer is best equipped who also grasps the psychology of users and the phenomena that occur when they are introduced to one another. This explains the widespread use of best practices in web design and architecture, as well as the sudden and surprising successes of the small few who innovate well.

A fourth corollary of the law of social practices creates a real challenge for social media design: users engage in a practice when they feel like it. This could belong to the law of user centricity, but it has implications for social practices. Face to face social encounters bind participants in time and space, and again through interaction and communication handling and negotiation. There is no binding of users in time and space on social media. It must be handled entirely by actions and communication. Actions are what the user does; communication is what the user says. (Note that in social media, communication can be performed by means of writing, recording and uploading video, audio, getting on webcam, and more.)

It’s for the lack of binding in time and space that we call social media discontinuous, de-coupled, disaggregated, dis-embedded, and so on. The binding that does happen is not an event, as it is in face-to-face situations. It has no situation, and no duration in time. Rather, it is a sustained commitment taken unilaterally by individual users which can produce the effect, and thus the experience, of bound and mutually-framed experience. Being next to one another is possible, in a way, but being with is not. This thwarts the possibility of individuals sharing in each other’s "stream of consciousness" — which is to say the emotional and empathic coupling of activity and presence that makes us seek out social interaction to begin with.

We’re now at our last corollary of design by social practice, and it embraces the previous one: the more open and simple the social tool, the more uses it has for more users. And you may have guessed it already, but this one also presents a design challenge. For while open-ness in structure and design may engender a greater number of uses, too many uses may render the tool useless. Like the blog, Twitter exhibits design simplicity in the extreme. Any user can write whatever s/he wants, whenever s/he wants, and from wherever s/he wants. But unlike the blog, audience members have no control at all over what they read.

The asymmetry of experience returns — and now the creative flexibility of the application results in undifferentiated talk and messaging for the consumer. The user interest satisfied in reading and following Twitter is undirected and unstructured. The experience of the reading user depends entirely on whatever happens to have been posted by whomever happened to have posted it, that the user is following. This kind of arbitrariness and randomness of news and messages would be death in any other medium. The 140 character limit survives as a necessary constraint on the noise level — even though it can contribute nothing to raising the signal level.

We have covered just two laws of social interaction design. Our next, communication, will get at the interaction type that drives status culture and talk-based applications. Can talk be designed?

Adrian Chan is a social media experience expert and analyst. You can follow him on twitter at gravity7.

This is a short post I want to put out there to get discussion going on structure in social media. As I’m still thinking about talk tools, and short-form messaging ("status culture") in particular, I’m having to contend with some tricky conceptual stuff around structured user experiences. Facebook and other social networks are much more structured than twitter, status updates, and short-form messaging. From a Ui and user experience perspective, these tools bring a lot of order and organization to user actions and interactions. That has the benefit of limiting noise and of creating a lot of different sub-system of user actions. Games, gifts, leaderboards, rankings, ratings, post vs comment types, tags, social navigation, what have you. Stuff I and others have written about in terms of pattern languages and design approaches.

Twitter and its kin are unstructured. I’ve come up with the proposition that when structure is under-determined in site/system architecture, social practices handle the organization of experience. The burden of structure is shifted from architecture to interaction handling.

Different types of talk are well documented. Erving Goffman’s symbolic interaction has made huge contributions to our understanding of forms of talk as "framed" encounters. Framing happens in time, in positioning of actors, in turn-taking, "keying" and "footing" changes related to statements and what they mean. But facework is critical to his analyses. Not to mention use of body language, eye contact, tone of voice and so on.

What are the possibilities of open systems of talk? I’ve begun thinking about this from the perspective of multiple personality types and frankly it’s getting ugly. How does a socializer relate to a pundit? What kind of twitter activity attracts a harmnonizer? Does an inviter look for retweets? It’s simple with a single user model, but more realistic if we can account for the different kinds of user personalities and what they are competent and interesting at doing online.

Since social practices emerge on social media without any directed guidance and only through the undirected participation of users who each have their own reasons for doing what they do, the challenge of designing for emergent practices is a tough nut indeed. Where is the threshold for the emergence of a particular practice? And what’s the upper limit for an open tool’s population — the limit point beyond which it drowns in its own unstructured noise?

I was thinking last night about some cool things to do on twitter, for example. But which I haven’t seen. There are four ways to contextualize a tweet: the accountname, pic, @name, and hashtag (#). The rest is the tweet statement itself. So I know a tweet from /bbcnews is news. Or it could be indicated #worldnews. Or the pic could be the bbc logo.

Given these limited means of contextualizing a tweet — that is, providing cues to the reader as how to read it — there is still a lot that one could do.

Eg.

–Use an #clickmypic as a clue. Create a user pic that is legible only in orig size (viewed on profile page). Embed a message or clue in the pic. Tweets could then be created that were:

• trivia pursuit questions: the pic is the category
• save the planet: the pic is a question, e.g. what’s your contribution this wk? Response is whatever small thing you’re doing this wk to save the planet
• coupons/discounts — viewable only if you expand the pic size
• movie character — the reply should be the movie the tweeted movie quote comes from:

–A #tagyourit game. Self explanatory

–#onethingyoudontknowabout me. ditto

–#soundtrack (what i’m listening to)

–#flixsterquiz (never-ending flixster movie quiz question)

There could be tons of these small twitter games, with @naming for pass along. I’d like to see a brand try something like this out. It seems to me that the creative possibilities for open or unstructured talk tools are huge — all that’s needed is the creative, and a simple-enough or familiar enough game structure to make it fairly obvious how to play. (The game rules supply structure, tweets become the game’s "moves".)

To return briefly my problem of personality types, and whether we can find personality in tweets, and twitter (and status update) use practices that correlate with personality types, I think the answer is yes. But it’s neither foolproof nor straightforward. We update and tweet on whim and fancy, mood, and conversationally. Those are practices that fall outside of personality type-casting. I’ve managed to find strong consistencies in how a lot of people update:

• people who tend to describe feelings, moods, or activities (Self-oriented)
• people who solicit a response, address someone else, frequently @name (Other-oriented)
• people who multiply @name, who tweet events they’re at, who they’re with (Relationship/activity-oriented)

I’ve found some fairly consistent example of updates and messages that include:

• identifying with something a person is into (Self is attributed a pastime, goal)
• identifying with a value, cause, political theme (Self is associated with a value) 
• identifying with a group, practice, or status sign (Self is attributed desired status)
• positioning and location (indirectly soliciting contact and making Self available)
• third person comment (Self is reflected upon, "judged" or joked about)
• event-specific (what Self is doing)
• mood or feeling (how Self is feeling)
• etc

These and other kinds of status updates and messages seem consistent with the user’s personality type. Now, theoretically, a functioning social system would reveal that personality types that go well together can actually be seen forming networks. Those who like activity should be found with those who are active. Those who identify with attributes of others should be found with those others. Those who em-cee should be seen mentioning those people they find interesting (em-cees can spot the rockstars, tend to talk about them more than their own Selves). And so on. Many many natural couplings and sets of users whose personalities should produce emergent social practices.

I’m very interested in doing this in collaboration with psychologists and have started doing so. Interestingly, social media tends to be a field for social psychologists — and this is more a matter of personality (even clinical) psychologists. Social psych takes on status, social hierarchy, roles and positions — the kinds of things that are common to community. My approach here is to find personality-based combinations and their practices, which is a different tack (is also more user-experience based).

That’s what’s on my mind. Designing and building successful social media tools, applications, and uses around open systems and especially talk-based systems is creating more challenges for design methodology than did the web-based social networks. I think it can be done, but it’s going to be a lot more sociological and psychological than most design approaches are used to.

Adrian Chan is a social media experience expert and analyst. You can follow him on twitter at gravity7.

I was thinking last night about an essay Michel Foucault once wrote about two competing concepts of the Self in major world religions. It’s been so long that I don’t now recall which essay it was. Foucault is known for theoretical "archaeology" of western thought. And for his work on the the birth of the "Subject" (read: individual). As in, when did the subject, the sovereign person, emerge in thought and culture? And more specifically, when did the Subject become the locus of truth? (He read this through the inquisition, the practice of confessions, and so on).

Anyways, in this essay he compared two views of the Self: the Self that is discovered and known through some kind of religious quest and search. And the Self that is created, invented, through free will, action, choice (and so on).

It occurred to me that a similar bifurcation exists in social media. We have a lot of discovery engines and techniques. Techniques once used to find related documents and data, but now often used to find compatible or similar people. This is an approach that ascribes attributes and qualities to the identity (person, user). They might be interests, demographic data, age, gender, location, even social graph/friend relations. It’s an approach used ultimately to help us find people we might like. Based on the idea that when two things are alike, their shared likeness might lead to further relationships.

But there’s an interesting flaw in the logic. That two things are alike might be liked by one person is fine. But that the two people who like those things might like each other, makes a leap of faith. It rests on the idea that the relationship between two things can be extended to the two people who relate to those things in like ways. We don’t know that this is an extensible logic or idea. Do similar people automatically like each other? Really? If so, aren’t the similarities that would make us compatible, make us friends and friendly, just as likely to be something other than what interests us — our style or personality?

I’m reminded of the logic of dating sites — that a match is a basis for meeting. Anyone who’s tried online dating knows that the first meeting is where chemistry either seals the affair, or dissolves the whole run up into an awkward and disappointing mess.

The logic of long tail can work on objects and things because they are stable. Attributes used to describe them are values that can be shared. They belong to each thing (a movie is documentary) because the two things each share that attribute. The more attributes in common, the more alike they are (these movies are documentaries about penguins).

But is the approach extensible? Do we like each other because we share attributes?

There’s another approach taken in social media — the social graph. This version uses Granovetter’s weak link theory and suggests that the friend of a friend is the most important relationship — because it can introduce us to people who are not one, but two or thee degrees away. We get access to people who aren’t our friends but are closely linked. It’s assumed that trust is extensible from the first degree (I trust you) to the second (I trust someone you know). Not the most convincing idea, but good enough to make friend recommendations.

But in each case, we have only a system of things and attributes.

Human relationships aren’t build on similarity or identity of attributes. They’re a result of interaction, of understanding, of the things we do that move us and by which we move one another.

Our industry needs a richer understanding of the creative acts and the productive aspects of social media use. Of what is required, and what happens, when a connection becomes meaningful to the people connected through what they do, not have in common, with each other. We need to think more about drama. about stories, about conversations and pastimes. About the things and people we anticipate, expect, and wait for. About what time is like, and times are like, online — short and long times, ongoing times, choppy and interrupted times, rhythmic times and times that are over. About how all the dynamics of interaction are transformed but somehow retained and adapted to the way things work online.

Yes, discovery can be produced by searching among common attributes. But the really productive stuff comes out of social practices. Social media may be a means of production. But we are still the production of means.

Adrian Chan is a social media experience expert and analyst. You can follow him on twitter at gravity7.