The socially-mediated NYTimes: What would interest the reader?

The New York Times has just published an interesting piece on why readers email articles. The research (Social Transmission and Viral Culture) was conducted by Jonah Berger and Katherine A. Milkman of the University of Pennsylvania's Wharton School. In the Times article, Will You Be E-Mailing This Column? It's Awesome, author John Tierney summarizes some of the study's surprising conclusions.

According to Tierney, the study examined reasons behind which articles are emailed most by readers. It addressed questions like "Do people prefer to spread good news or bad news? Would we rather scandalize or enlighten? Which stories do social creatures want to share, and why?" And by Tierney's account, the study discovered that "Perhaps most of all, readers wanted to share articles that inspired awe, an emotion that the researchers investigated after noticing how many science articles made the list."

He goes on to write: "Building on prior research, the Penn researchers defined the quality as an 'emotion of self-transcendence, a feeling of admiration and elevation in the face of something greater than the self.'" Why awe? Because "Emotion in general leads to transmission, and awe is quite a strong emotion." (Berger) And awe is evoked by the Times in stories with inspiring and substantial tales to tell: "They used two criteria for an awe-inspiring story: Its scale is large, and it requires "mental accommodation" by forcing the reader to view the world in a different way."

I haven't read the original report, but what I find interesting in this piece is that the researchers sought answers primarily in the content of the articles selected. Granted, this kind of research can be conducted at scale more easily. Articles were analyzed for their headlines, topicality, their affectivity, and more that can be mapped into algorithmic text analysis. But from what I know of recommender systems and of social media, other reasons explain why people make recommendations also. These are personal and social reasons, related to the selection's reflection on the user and in social cases, to the attention, status, and other aspects of sociality a user may find engaging over time.

Which begs a couple questions: What reasons might readers have for recommending articles that were not covered by the researcher's methodology, and What might the Times do otherwise to increase and diversify reader engagement?

Recommendations in social systems are an interesting practice. And I suspect that the deeper the sociality of the system, the more diverse the motives users may have for participating. Yelp is one such example. As a recommendation and review site, it allows and even encourages users to disclose themselves in their reviews and recommendations. Users are encouraged to build a following, to engage communicatively, and to use reviews as a proxy, if you will, for the self-disclosure that animates online dating and profiling sites.

In review and recommendation systems, users do more than just review a particular business. They talk about and reveal themselves in how they like, what they like, why they like, and how often they frequent the business. The business reviewed is really a springboard from which users can disclose personal tastes, preferences, interests, expertise, habits, and other attributes of personality, style, and character. The review format allows users to talk about themselves without directly talking about themselves (a reason that online dating sites provide numerous versions of the "about me" profile entry).

Recommendations are not yet reviews, of course. Times article recommendations are made by an individual user to his or her followers, or to friends and colleagues. In the latter case, the recommendation will likely be particular to its recipient, as the recommendation is a direct message. But in the newer case of Times followers, the recommendation incorporates the strange/familiar audience makeup we are now so used to in public social media systems. And for this reason, recommendations surely involve self-interest, attention-getting, status, reputation, credibility, and other aspects of social status and status updating common in the age of twitter.

Here, the Times could take advantage of the social incentive structures that motivate deeper and sustained reader engagement. Improving on the user-audience engagement currently satisfied by Times followers, it might build out reader communities. My own recommended articles, for example, are for the most part arbitrary reflections on what I found interesting and worth reading. I shy away from the front page for the obvious reason that front page news needs no surfacing. There are times when I select content that might reflect on my interests, but also times when I select content for its news value, its substance, and sometimes its relevance to the audience I assume (I haven't looked deeply) is following me.

It's in the latter that the Times could expand reader involvement. If I (and I think some of my colleagues, based on what I see them recommending) and others believe we have social media/technology types in our following, then I will sometimes recommend content of that kind. I'm looking for pickup by my audience. That tells me that I care about audience reception — which is a matter of social status. And I don't think I'm alone in this. If I and others care not only about the article's intrinsic interest, but about the interest others take in the article, then sociality is latent, present, and indeed actionable and extensible (by the Times).

A Times community might extend this interest to develop topical expertise and surface topical experts. It might leverage the social incentives that help experts achieve visibility and credibility, reputation (please, not by points), and status. The Times is surely interested in audience participation as a means not only of validating its content but of further distributing it. This is just an example of the mass media incorporating and assimilating social media strategies. The move demonstrates the shift of reader interest and engagement from content per se to Self and its attendant interests in attention and status. To wit, the Times might do well to ask not just What reader engagement makes the Times look good? but What content makes the reader look good?

A Times community of readers might then take advantage of the intrinsic conversationality of active reader engagement. Experts might go up against other experts. Critics might question other critics. I've looked at a number of the personality types I think can be found in social media. For simplicity's sake, I divide them roughly into Self-oriented, Other-oriented, and Relational-oriented types. Their social competencies vary in how they talk about themselves, converse with others, and participate in social games and activities.

The Times could satisfy this diversity of modes of engagement by means of some pretty straightforward social organization. At which point I am certain it would find that user interests in content reach far beyond the article, to the much richer social modes of interacting and communicating with social scenes. Scenes that support Self image, that provide for conversation and discussion, and which enrich and diversify socially topical fields of interest.

The depth of user engagement becomes much more interesting when the modes and means of relationality and relating become interpersonal and social, when a brand's sociability is more than an extension of brand image but is a complex field of scenes and publics that vary in degrees of engagement, attention, speed, and rhythm. When content is validated not just for its substance but for its modes of referentiality — which is to say, the ways in which its substance, its claims, its direct and indirect references (denotation, connotation) take on the social significations that come with social re-use, appropriation, and re-contextualization. Yelp did it. I would guess that the news media can, too.

Swurl: lifestreaming and timelining

Like many of you, I simply can't keep up with the river of lifestreaming applications hitting public beta this year. Many seem to simply do the same thing, more or less, with a bit more of this or a bit more of that to differentiate each from its competitors. But social apps are bound, perhaps more even than "conventional" software, to conform to best practices. Why? Because they are social applications. Social applications succeed only if they can extend the individual user experience out into new and interesting social experiences.

And they do have to be interesting -- for social applications, again more than conventional software, must be interesting. More often than not they are interesting because they are used as tools for talk. Talking with, to, at, amongst, in front of, behind, and to the side of. Talking with friends tends to be interesting to those involved simply because it is among friends. But where the face to face dimensions of social interaction are also rewarding for the obvious reasons, social applications must deliver a working substitute. There is no real "spending time together" online.

Even chatrooms, which are as much a precursor of lifestreaming as anything else online, can only approximate this sense of togetherness. I recall early days in IRC chatrooms where that sense of being there or of being in it was as much due to the suspense and waiting (for somebody to type out their response) as it was due to the "room" itself. One might even argue that this pressure of time grows in the user the slower the technology is to record and transmit time. The longer the latency, the greater the waiting, and thus the greater the anticipation, suspense, and urgency! (Is it not said that suspense in film is simply the time that it takes for something to happen?).

Swurl.com is interesting because it has a visual timeline of the lifestream (pictured above). In calendar format, and well-designed, the timeline looks good and is an attractive visual representation. It's low on conversational content and talk, but it captures the past of a user's activity in a compelling presentation. Plurk.com also has a timeline, but one that is used to steer interaction (and which looks more like a horizontal river display). Not only does Swurl's calendar provide thumbnails of pictures and shortcuts to posts, it expands to accommodate periods of heavy activity. All days do not look alike. I like that.

This variation is important in lifestreaming apps. In contrast to the profile-based site or service, the stream stands in for the profile. The person's talk stands in for profile elements. These choices make sense, because the call to action in a lifestreaming service is talk. It's not browsing, searching, or navigating. At least not quite yet (I believe we're ready for more order and structure). Really, each message/post/tweet in a lifestreaming app is its own call to (inter)action, which is also why most users are in it "now" or never.

Which makes Swurl's representation of past user activity interesting to me. Most lifestreaming have stayed away from the archive of past activity (what's the pleasure in paging backwards through a user's posts?). But there's a lot of value in past activity, and visual coverage of the past can take many forms (think Edward Tufte). We've seen none of them yet (Chirpscreen's slideshows come to mind, though it would be nice to see them become actionable) but I'm certain that we will.

If twitter is the power curve of lifestreaming, then apps like swurl might show us some of the value in the long tail -- the long tail being the past. To picture this, take the standard long tail graph and turn it sideways. The Present is the curve, the Past is the tail.

Mining the tail of time is mining in depth rather than mining across connections. Mining the connections of past time, for lifestreaming apps, might mean drawing connections across the past times (pastimes, experiences, too) of a site's users. Currently, Swurl engages conversations around a user and his or her posts. But we could imagine indexing user streams for the purpose of making connections and extracting content. After all, a user's post posts, talk, uploads, etc are used by many applications to predict or anticipate choices and preferences.

I'm excited to see what Swurl, with minimal complexity, has done to wrap a bit more around lifestreaming than we get out of tools like twitter. Twitter will remain for me my primary talk tool, as it has and will continue to have the best audience awareness. But if you wanted to imagine social networking, and profile-based social networking around lifestreaming instead of profile pages, Swurl would be a good place to start.

Join me on Swurl!


Related reading:
Readwriteweb on Swurl
Techcrunch on Swurl

Of crowds, power, grass, social theory and social media

Crowds and Power
Table of contents grab from amazon.com


I'm reminded now and again of this book, which describes crowds, audiences, tribes, mobs, assemblies, riots, gangs, and all manner of aggregations of people in the kind of prose that's now a rarity. For those of us in social media interested in the many ways in which our technologies assemble audiences, this Canetti offers serious insights.

He details differences between mobs awaiting, crowds gathering, audiences listening, mobs erupting, riots exploding, queues queueing, and so on... To Canetti, each social assembly captures not only a different force, but embodies (literally!) different affects. They are oriented to an other, or huddled in self-defense. They anticipate in patience, in frustration, or in awe. They worship or hound, condense or flee. They surround a leader, pursue him, or depose him in masses gathered in the public square.

Our media don't do quite the same things, of course, but there are similarities and inspirations aplenty here.



And to go with it, perhaps, the film Grass, a 1925 documentary (intertitled but no voice over -- just music) about Bakhtiari tribes gathering their herds for an annual migration to grasslands. Have you ever seen livestock herded downriver by swimmers?

Grass review, from imdb.com : Fantastic documentary of 1924. This early 20th century geography of today's Iraq was powerful. Watch this and tell me if Cecil B. DeMille didn't take notes before making his The Ten Commandments. Merian C. Cooper, the photographer, later created Cinerama, an idea that probably hatched while filming the remarkable landscapes in this film. Fans of Werner Herzog will find this film to be a treasure, with heartbreaking tales of struggle, complimented by the land around them, never has the human capacity to endure been so evident. The fact that this was made when it was shows not only the will of the subjects, but of the filmmakers themselves.

New SxD White Paper: The Social Engine that Drives Review Sites

Social Interaction Design Guide: The Social Engine that Drives Review Sites 2007, pdf, 16 pages. NEW! A Social Interaction Design guide to the social engine and engineering of user motivation and participation on review sites. This lighter-than-usual white paper looks at the social practices engaged in web sites built around user reviews. In particular, the paper examines the way in which reviews can become a kind of personal profiling system for reviewers. It also looks at how reviews create and add value, and poses the question of how business might participate in social marketing of this kind.

The User Experience of Review Writing

Another excerpt from the forthcoming white paper on Review Sites...

Review Writing
The inner experience of writing a review involves a large number of things, and without going into any in depth, we need to acknowledge at least several of them. There is the thing reviewed. There is whom it is written for—this might be "yelpers" or "anyone" or "mommydaddy" or "friend," "stalker," "business owner," "the Almighty," or "the cute Yelper who just requested my friendship." There is the style of writing, which might hew close to the utility of reviewing or stray off into personal ramblings, flashes of wit, hooks and lines designed to get attention, and so on. There is the use of qualifying observations by which a reader can glean, for him or herself and not because the author has said so, the salient selling points of the thing reviewed. There is then, as just mentioned, the recommendation or advice given within the recommendation, which itself can vary among all shades of "should," "perhaps" "tentatively" "confidently" "ought" "must" and "not." There is the revealing of the depth and scope of one's authority on the matter, or not, or lack of it (which is not the same as not revealing, it's a matter of not admitting!). There is the difference between being the first to review, in which case the review may inform subsequent reviewers, because a review can easily be a response to a review, to reviews in a series, or to reviews overall, depending on where the author puts him or herself in his/her emotional/mental relation to the whole proceeding. There is the review as comment to, or commentary on; and in commenting to, one might address reviewers, commentors, their reviews or their comments, though it may be hard for the reader to tell which is which.

The experience of writing a review is in fact complex indeed, and that's not including the potential for misspellings, errors in fact, misinformed or inaccurately attributed perspectives and observations. Nor is it including the post window, tags, and now the addition of icons that can be used to represent a gestural remark, which again may indicate to some a reflection on the review, or the reviewer, and it can be hard to tell which is which since we can't ask the person who selected them. And none of this includes the context of the review, which is to say some reviewers choose a time of day, or a category, an oft-reviewed Thing, trend, or bit of news as a means of attracting more attention (to themselves, their review, the view of themselves as manifest in the review, or perhaps to others. Or the Thing, even!). And again, none of this addresses the site in which the review is posted, its "branding" and community, and the sense that each user may have of what those are, how it serves them, or whom is served, and so on and so forth. The production of a review, as we see, is not so simple as the posting window would have it. From the perspective of social interaction design, at least.

Technorati tags: blogging, interaction design, social interaction design, social media, social networks, social software, sociology, theory, reviews, user generated, web 2.0, web,
Yelp

Understanding the Yelp factor and social reviews

The rise in the popularity of user reviews on social media sites has a lot of people talking. Here is a mode of social interaction online that doesn't require joining MySpace and putting one's Self on the line. At least, not in the manner that many of the community-oriented social software sites would have us do it. In contrast to their more fully-functioned brethren, review sites present a relatively simple value proposition: associate yourself with something, preferable something you like (product, place, experience, travel, it makes little difference for now), and describe it for us in your words. In other words, disclose some of your interests, your style, personality, habits, and preferences, by reviewing something that we can all relate to.

To support this user disposition, the codes of interaction on social media sites tend to be informal, and the proceedings are largely unstructured. There are a few categorization and publishing requirements, of course, but just a few. The system handles the reviews, attaching them to things reviewed, making them search-able, find-able, and organizing reviews collected according to modes of distinction (relevance) by-and-large inherited from search engines and common social software practices.

To the reviewer (user), then, the frame of interaction and value proposition seem fairly straightforward. Where it gets interesting is in what happens next, for review sites involve much more than just reviews. Reviews can be written for all kinds of reasons, some of them having little to do with the Things reviewed. They might also be written to any number of users, for reasons that vary from the highly personal to clichéd. And interactions among reviewers and their readers, too, run from personal and enthusiastic agreement to cold-shouldered neglect. These variations exceed the value proposition of user-generated reviews and give us a compelling case study in social media.

So as social media designers, we need to address two different user experiences, the reader's and the writer's. Our need to motivate and engage the reviewer's participation requires that we design a system to support the writer's subjective experience of writing review. We need to supply an audience, topics, stylistic differences, a participatory genre, if you will. Reviewing Things has to be interesting and compelling and must have purpose, if the writer is it hand over his or her attention to it. But for similar reasons, we to provide the reader with value also. In theory, at least, reviews should display as much objectivity as possible—enough to warrant their utility as reviews (and not just as opinion pieces). Do these two user experience propositions stand in a fundamental conflict?

Finally, we need to examine whether the design of social media can structure the axes of use on either side (reviewer/reader) such that the value produced is the value consumed. This is the nature of the challenge that often faces social media designers: creating an efficient marketplace, without use of real money or real incentives, by enabling the production and consumption of knowledge such that benefits are captured on both the production and consumption side of the equation.

Take the popular review site Yelp.com, for example. Now this site is fascinating, truly excellent in many ways, for it has succeeded in surviving without merchant participation.

The fact that Yelp.com comprises user reviews written without any merchant presence preserves the site's integrity. Reviewers are under no obligation to do anyone a favor; nor do their reviews benefit them in any fungible way. So the system provides a forum in which reviewers may write from whichever position motivates them. And because nobody's going to spend time writing about stuff unless they believe somebody might read it (this, at least, is my hope), Yelp's members tend write for each other.

Out pops the Social, and reviews become a means by which members get attention; describe and reveal themselves through things they know something about; show wit, style, pictures and collections of compliments (which span a range of review-oriented to the unquestionably-no-use-for-this-icon-but-to-flirt); make friends; find popular things and review them because they've been reviewed so many times; become domain experts; wander widely off topic; and so on. And please don't get me wrong—it's hellafun. Indeed Yelp has become an interesting case study in the importance of anticipating the social forces that emerge when a system is launched into the world, and interactions begin to pile up one on the other. For users don't read manuals, or the fine print in the terms of agreement, to learn how the system works or how to use it. Users, and I should simply say "we," look at what others are doing. This tells us what's going on, and with that, how to proceed. If it's empty, well then no point in trying to become popular. If it's full of people, then whatsup and whatsgoingon?!

Social media sites are built on the contributions of users who themselves orient their contributions to the site's organization, theme, and audience. On Yelp.com, for example, some write many; some write deeply; some write to write to others; some write their secret discoveries; others can't believe it when a member holds out that local nugget for all to see (I committed this neighborly faux pas when I revealed that a local grocer squeezes its own orange juice). ;-). Some, having written a few, find those who have written about the same; while others find members they like, and comment on their writings for the association. Some — and don't get me wrong, I do this, and there are no write and wrongs here! — write to the author, some to their scene, some write about themselves with utter sincerity, and some write to cover, dodge, and cloak with seductive mystery all around.

Indeed, the social practices emerging around social media become particularly pregnant in the case of review sites. It is now a bonified genre, though some sites participate thinly, others richly. Contributions are codified across categories, horizontally and vertically. This serves the needs of site and content navigation. In each vertical we have "best of's, lists, recent, trends, price and other qualifications. Vertical organization is simple, as it's vertically organized: books, music, blogs, dvds, consumer electronics, nannies, and (yes, we're that far along) review sites themselves. And of course reviewers, too, are presented by similar qualifying criteria.

Now it would seem that the disinterested review is the most useful. But we're in social media-land now, and a) there's virtually no such thing as taking a disinterested relation to a Thing that is liked; and b) because a writer's reviews are hung out for others to read and know them by, even if it were possible in our write up a disinterested review of the Thing At Hand, social interest (getting noticed, being accepted, liked, and other warm fuzzies) rather blows that all away. Let's be honest here: nobody's going to spend time writing about stuff unless they believe somebody might read it.

We just split the practice along two axes and user experiences: reader and writer. So now let's integrate the social back into the two axes then. No user review on any user-generated content site is published without containing within it some negotiation of the possibility that a post/review/comment may be taken up in communication. Reviews are now a form of talk. Users are interested in each other (even when this takes shape only within their minds only), and this interest can overrun the objectivity that would most benefit the stated purpose of review sites, to wit, qualifying the Thing At Hand with unmotivated user evaluation. Interest in people; interest in content. Does the split pose a problem to the genre?

Social participation is essential if anybody's going to be bothered to write. But social participation may also transform expertise and utility into a popularity contest for compliments and friends and sheer volume of reviews. My personal preference is movies, because I've seen a boat load of them. For many it's food, dining out, bars and restaurants and all that fun stuff that people do when they're not cracking their craniums on the inner workings of social software. Everyday Things belong to familiar turf and territory: good for self-disclosure and personal opinionating. But how will the marketers and markets integrate it all? Because ultimately, business wants a piece of this, and the word is out: there's gold in them thar hills.


Technorati tags: blogging, interaction design, social interaction design, social media, social networks, social software, sociology, theory, reviews, user generated, web 2.0, web,
Yelp