Activity Streams: Content and Flow

The realtime trend continues unabated, with presentations at TechCrunch50, Facebook’s recent updates, and next-generation newspaper designs all extending the impact and value of the stream in social media. Disaggregation begets reaggregation, as demonstrated by the newcomer threadsy this week. As client applications and new services add organization and structure to activity, news, status, and twitter streams, we see hints of what is likely to come in the months ahead.

I think there are two distinct trends at work here. One, the popularity and adoption of the stream as a form of social conversation. And the other, the conversion of realtime information into value that can be consumed outside the stream. Or to put it another way, the value of being in the flow, and of watching it from the river’s edge.

These trends lead to some interesting implications for social media professionals, from designers and developers to brands and businesses. We’ll take them separately, and for simplicity’s sake distinguish them by interaction “in the flow” and interaction “around the flow.” Interaction in the flow is conversation and talk itself — in the form of tweets and updates distributed among friends and across various social spaces. Interaction around the flow involves the value-added actions and activities that use social content, such as rating, diggs, tags, and so on.

In the flow
Many of us are in the stream to communicate. We use the tools available in ways not entirely dissimilar from the ways we have used message boards, IM, chat, and email in the past. We experience it as a kind of online talk. Here, the interaction systems emerging around stream applications focus on ways of improving communication. Twitter’s @reply and retweet are examples of these. So, too, is Facebook’s recent adoption of activity tagging. Feed readers for streams both reaggregate these distributed conversations and provide for interaction within.

As in all forms of talk, the critical design and experience elements and features include addressing (individual, social, and public audiences intended by the user), subject or object, topic (hashtags, tags), time stamp, and other references (could be @names or could be links). Other distinctions not yet supported would include other linguistic types (request, invitation, answer, greeting, etc), urgency, commercial/individual, and more.

A lot of interesting things happen around conversation and designers are only beginning to wrap their heads around the possibilities for surfacing value, extracting meta data, structuring and organizing talk itself, and so on. Because the primary value of the user experience lies in communication itself, the possibilities are virtually endless.

We can easily imagine a wide range of activities that are currently page or site-based being handled instead by the stream. Invitations, meeting requests, buying and selling, questions/answers — these and much more could be transacted by means of messages “off the page” and extracted or sorted out of streams by smart clients or aggregators. Analytics companies will have a gold mine of relationship data to scrape and visualize, for example, for use by those who want to see how influencers reach their audiences, around what topics, how quickly, with what redistribution, and so on.

The conversation space holds many more opportunities than we can currently take advantage of, in part because many applications are still trying to simplify the experience of being in the flow. At present this requires aggregation of messages posted across numerous contexts. Over time, however, it seems inevitable that conversational tools will be able to offer not only the direct messaging experience but also a variety of benefits from use of metadata, analysis, search, and structure/organization.

Around the flow
For those who spend less of their social media time in the flow, there are the interactions with content instead of person. Many of the long tail services create value through interactions with content that are designed to surface and rank by popularity, trend, similarity, rating, and so on. This is the world of taste-making, and it uses indirect social interaction (meaning not person to person communication) to qualify social content items. Recommendations services depend on the contributions of users to qualify and differentiate content: the more ways there are of differentiating content items, the more ways there are of relating it and providing navigation through it.

The primary goals of interaction models used around the flow involve separating content from the conversational stream, extracting meta data where possible, assigning categories and embedding within content structures and navigational systems. Then the social challenge becomes making it accessible (search, browse, and categorization) and making it socially interesting (lists, rankings, votes, etc).

The disadvantage this older page-based method of social experience has with respect to the recent conversational trend is of course that it’s at a remove from the user. The factors that compel us to talk are not available here. Attention is not paid to people directly, but indirectly through means of content. The advantage, however, is of course in the many ways already developed for organizing content and making it available for re-use within other contexts.

[Note: All interactions with social content first involve a selection of something. These indirect kinds of social interaction assign value to the content item (a vote up or down, a rating, a favorite, a tag, etc). Selections in the stream, by contrast, create value by distributing (sharing, replying) communication. There is a critical distinction between the direct communication interaction model and the indirect social action model. Communication uses language; social action uses symbolic tokens and signifying systems like emoticons, icons, ratings, votes, etc.]

In each of these two trends, value is in the relationships, either between people or among content elements. Communication itself creates value, but of a kind known best by those involved and extracted only with difficulty. Social content can more easily represent value assigned to items, but must then find ways to restore what made it interesting in the first place.

One could see a new breed of social networking experiences built around messaging, if conversational features can be codified and structured for ease of participation and consumption. It will be interesting to see whether or not this happens. In either case, the emergence of interaction models appropriate for communication and social participation, to streamline communication and to make social content a more interesting experience, holds a lot of promise.

This entry was posted in Design, Streams, SxD Theory and tagged , , , . Bookmark the permalink.
  • http://twitter.com/wojtek3000 Wojtek Szumowski

    I think that powerful attractor of stream is that it makes you feel like you participate in something that others participate in. Stream gives us a sense that we are having a social experience, even without active participation (like RT, or @reply).
    Stream feels like a pure form of communication, free from pursue of goal, appealing directly to our desire to communicate.

  • http://www.piedcow.com malatmals

    i think the money will be when advertisers start bidding on the placement into the stream… say to sponsorA hey put your ad here where theres 10000 eyeballs a second and an uptrend and oh by the way its $1000 and say to sponsorB the same but its $1100 now for u… ten minutes later the price could be $10 as the advertisers switch streams to follow the eyeballs.

  • http://www.gravity7.com/blog/media/ gravity7

    Bidding makes perfect sense. In fact the marketplace will likely resemble the stock market, and those who design the systems may well hail from Wall St hedge funds. Flow is about rate, acceleration, distribution, reach, and as such the value of attention is measured in realtime, and predicted by trend analysis.

    Feeds offer the possibility of product placement — not keyword, not context, but placement within the activity stream and possibly in and around the words of users themselves. Embedded advertising. I'm not for it but it's where advertising wants to go, if users let them.

  • http://twitter.com/ronchung Ronald Chung

    Adrian, immensely enjoyed your analysis on in-flow conversation and sideline engagement. Activity streams are definitely going to be a big trend going forward. Let's chat about taking real-time streams beyond where they are today.

  • http://www.gravity7.com/blog/media/ gravity7

    Thanks much! I've written on this in the past, you might poke around old posts using the Table of Contents link up i the sidebar. Or let me know what you're thinking — I'm up for discussion!

  • http://sull.tumblr.com sull

    yes, great post!

    i recently shared some thoughts around the idea of “stock and flow” in the context of real-time streams on the net.

    http://vocal.ly/p4l

  • http://twitter.com/mariagrineva Maria Grineva

    Thanks for the good post, Adrian.

    We are working on product that represents news from personal stream in the form of newspaper.
    The Twitter Times : http://www.twittertim.es , interesting to know your opinion about it

  • http://www.gravity7.com/blog/media/ gravity7

    Newspapers are a tried and tested format for news ;-) I could see a Netvibes kind of product for tweets — makes sense. Will check it out!

  • http://www.gravity7.com/blog/media/ gravity7

    Interesting points. A couple things on flow worth considering, and these are somewhat philosophical.

    a) it's system, not structure. We've been used to a structured page experience of content consumption. Communication is flow, and communication tools are better understood and designed as systems. To wit, the minimum UI design requirements of twitter. Working from systems forces you to consider time and timing, instead of pages, containers, boxes.

    We havent even begun to explore the possibilities of time-based navigation systems. Trend and metrics tools do it — timeframes, sliders, compressing the view of time, etc. But communication and traffic are different — making sensible nav for time-based communication systems requires some new thinking. (That's assuming we begin to use it that way — twitter is still a tool for now, and the recent past.)

    B) Many folks who write about flow treat it as a flow viewed as it passes — a passing stream of news. Actually, we're in the flow. This may seem a minor point, and one pertaining just to metaphor. But that's only if flow or stream is about news and content as information. If it's about interaction and communication, the distinction is more important.

    When you're talking to friends you are “observing” the situation, but you're in it, temporally, too. Users experience the stream at different speeds, with varying degrees of attention, and w varying degrees of continuity of presence (in it, in and out of it). These are what count when people talk about information overload — not so much the amount of info (afterall the web is massive), but the user experience (lack of flow control, exposure, attention requirements).

    As an experiment, imagine twitter were an audio stream. How long could you stand it?! Hearing can't select, speed up, or reverse time. Eyes can, and it's that selection of what we pay attention to, plus decision to interact, that takes time and effort. And which might offer possibilities for navigation. If one represented the stream in different ways (there are ways to visually represent time-based experiences).

  • http://bernardmoon.blogspot.com Bernard Moon

    hey, how come there isn't a “like” button here? when there is, put me down for one.

  • Barrett

    Very interesting read. One thing I think will help interaction designers create better filtering systems is the understanding that a user's current role gives him an inherent perspective on the information he is consuming, and thus the actions he takes in that role are relevant to all other users who are acting in that same role.

    Imagine for instance that as a business professor I read an article in Google Reader that makes an impression on me as “interesting” and “educational.” Now imagine I am able to tag the article with my impressions of it and other business professors from across the country are able to see my tags, as well as tags from other colleagues on that article. They could then more quickly and easily filter dozens of articles just by looking at the impressions other business professors have of them after reading them. If they wanted to read articles that were “educational”, they could do that. If they wanted to avoid reading ones that were “shallow”, they could do that too.

    By allowing moms to see which articles other moms thought were inspiring, or doctors to see which articles other doctors thought were helpful, or web designers to see which articles other web designers thought were cool, then filtering systems would go a long way towards helping users find content that is more relevant and valuable to them.

  • aslevin

    I like the way this is starting to tease out the interactions & value that become possible with the elements of the flow & stocks generated by the flow. I think there are going to be several levels of this:
    * activities/messages themselves
    * contextual remixes of activities (around people, tags, time, links, etc)
    * aggregate value of activities
    * analytic insights from aggregate activities

    A few more observations in different directions…

    The flow is pretty broken now. It's possible to reconstitute a conversation on Twitter but it's pretty painful; conversations are direct on friendfeed but harder to discover. Maybe there are principles from wave that will help matters? Or make things worse :-)

    The metaphor of the flow is pretty limited – I think kevinmarks is trying to talk about this as well – if you restrict yourself to the last 5 minutes or 30 minutes you are reducing a lot of potential communication – humans are not that amnesic. We need flow and remix.

    I think the evolution from page to flow is not quite as stark as it seems – we've had comments for a long time; they're becoming more discoverable, we've had events for a long time; they're becoming more aggregatable. Although that observation is from the POV of the user experience – from the POV of the curator/service provider, the pallette is just emerging.

    one tweak, I think that the stream elements are often not going to be the transaction itself, but the event generated by the transaction, streamed and remixable. think of a last.fm event, a representation of a tune played, become abstractable, tagged, remixable by metadata

    one interesting experiment. among a group of people in the wild over time, compare rating and sharing behavior. How is it similar, how is it different. Is what you rate highly the same as what you share? Is it context-dependent?

  • http://www.gravity7.com/blog/media/ gravity7

    Adina,

    I'm with you on what we could do with items published into streams, and using meta data, aggregation, algos and analytics to add a layer of value over top. For discovery, remixing, restreaming, etc.

    My point on stream vs page is more for the designer, since the usual means of containing and laying out content have disappared. More on that in the post I did yesterday on Structure.

    I would also add the relational dimension, which is where we can not only glean from social graph mapping but also from the velocity, reach, redistribution, timing, tagging and more of the user's communication. Combine this with — and this would be hard to do — conversational analysis (what kinds of statements are these? questions, answers, invitations, offers…) and you have new possibilities more accurately aligned to what this is: talk.

    Knowing who talks to whom, who is simply announcing/posting, who responds, and what network effects pertain to talk, could make all manner of advertising and marketing more effective — eg marketing to the social graph through the individual with the most topical credibility for example, or advertising in realtime using the conversational speed and reach of high value users.

  • http://www.gravity7.com/blog/media/ gravity7

    Goog reader is already capitalizing on comments and sharing and with some success. A similar reader for streamed messages would make a lot of sense, and in some ways this is Fbook's recent implementation of update tagging.

    It seems we all agree on the opportunities to be had from meta data about the stream.

    thanks!

  • http://sull.tumblr.com sull

    more on stocks and flows

    http://vocal.ly/pd1

  • http://www.liveairshox.com/nike-shox-deliver-c-87.html Hlovewa

    good!

  • Pingback: Malaletra Blog - Edición y Diseño + Tecnología

  • Pingback: The Theory Behind Social Interaction Design | Johnny Holland