« March 2006 | Main | June 2006 »
Search's got Dogpile, Chat's got Trillian, Shopping's got Pricegrabber (and dozens of others), Classifieds have Edgeio, and now social networking's got Stalkerati. What's next? What other applications (not just content) lend themselves to this abstraction vector?
Woohoo! My earlier post on "What Can Web 1.5 Teach Us About Web 2.5?" is now Google's top organic search result for the term "Structured Collaboration".
I'm open to offers for a nice fat banner ad on Octavianworld, but you better hurry, I've got DoubleClick on the other line :-)
This week I participated on an evaluation panel for student projects with live clients in a senior-level computer science course I help teach at MIT. It's incredible to see how much these incredibly talented students accomplish in a single term, juggling other courses. Early exposure to real users made some of these projects especially compelling.
Everyone is out there trying to figure out the next MySpace. One problem MySpace suffers from is that it's not very visually appealing to a younger audience that values 3D games over 2D text. Enter new phenomena such as Cyworld, which has swept Korea and is now about to enter the US. One critique of Cyworld is that the visual metaphors effective in Korea may not translate well here, so the hunt is on for environments that might work better.
One application presented at the class that suggested interesting new directions for where MySpace-style communities might go was "atMyPad". Imagine that on visiting http://www.myspace.com/octavianspace you see not the current "portal style" layout, but rather enter an apartment that looks more like a Second Life environment. Different objects in that apartment represent different things that interest me. For example, if my blog includes posts about fish, a fish tank visually represents posts I have categorized or tagged with the term "fish". This visualization could be animated to reflect activity by me in this category -- perhaps a feeding frenzy if there's lots of comments, perhaps my fish get bigger as my Technorati rank rises, or perhaps the fish float belly up if I haven't posted in a while. Think "dynamic visual tagging".
There's an interesting business opportunity in this that the atMyPad builders are exploring. Just as wireless carriers sell me ringtones, atMyPad could sell me objects to decorate my pad according to the topics of interest represented by my tags and categories. Next, taking a page out of Linden Labs' book, atMyPad's builders could provide an API or their own simple scripting tool and object library to allow me to build -- and resell, with a commission to atMyPad -- objects to others. Going even further, sponsors might pay to have their objects distributed to decorate my pad and stuff its closets, and perhaps these might even represent 3D wish lists of items friends could buy for me in real life. In return, perhaps I, the pad's owner, could stream my iTunes playlists in tinny mono format out of the "stereo" object in my pad for my friends to enjoy while they look around my place. And, maybe there's an Oddcast-style avatar that might take care of my guests by answering questions about objects via text-to-voice rendering from my associated blog.
Eventually, perhaps as presaged by the Skype acquisition, eBay stores could go this way as well?
An entrepreneur approached us recently, puzzled and a little frustrated. She had what was, on the face of it, a great idea for a personal gift business. Even better, one tailor-made for viral marketing: it was the kind of thing that you'd expect people would pass along as an affordable and useful service to family, friends, neighbors, co-workers, fellow church-goers... But despite doing the de rigueur marketing du jour -- paid search and telling folks she new about it -- she wasn't making any appreciable gains toward the 100,000 user milestone she had set for herself and for investors as a gate for further effort and funding.
Viral marketing is conceptually easy to get, but tactically hard to do. One way to understand it better is to actually approach it the way epidemiologists do, which is to understand disease vectors and transmission mechanisms. The former refers to the paths through which an illness spreads, while the latter refers to how sickness gets passed along, and by whom. Listening to the entrepreneur it became clear that while she had a highly nuanced view of her ultimate target customer segments, she had a very uniform view of the folks who would get the message to them.
To unpack the challenge, we suggested the following exercise. Assume for a second that the virus will spread in a simple 1:10 pattern -- that is, each person tells ten, who then each tell ten in turn. In six "hops" you get to 100,000 people (ignore the decay rate for the moment). Question is, how do you sustain signal strength and consistency across each hop? The answer lies in figuring out who in advance your likely carrier for each hop might be, based on formal and informal roles; next, understand his tribe, his mechanisms for engaging it, and his motivations for doing so; then package your "virus" for maximum effect in that context.
Here's how this might work in practice. Let's take a popular example: "most-emailed articles". Today, the tools used by popular services like NYT Online include a form element where you can add your comments on an article you are forwarding to an acquaintance. One essential lesson from "Structured Collaboration" best practices is that blank text fields are some of the least effective elements for enabling people to interact. Most people, confronted with the "comments" field, might type something like "Thought you'd find this interesting" if they type anything at all. I presume they wouldn't have sent the article if they didn't think the recipient wouldn't find it interesting -- making the comment gratuitous, and missing the opportunity to convey something else.
That "something else" could be information that could provide *context* to the recipient. Let's assume the form element was, instead of a comment, a drop-down list or set of check boxes that described the recipient, or recipient group in some way. For example, recipients could be "work", or "school", or "church", or "government", etc. What would this allow us to do? It would allow us to convey the article in the context of other related articles that together might point to a theme.
This synthesis could help to convey a point much more powerfully than sending along the isolated example might. (To sustain usage of a mechanism like this, the email might provide only a short article summary and a link that would bring each successive recipient back to a site from which the enhanced forwarding mechanism could be used again.)
Taking this one step further, add tagging. Other readers subscribing to those tags (via email or RSS, doesn't really matter) will be more "susceptible" to reading those forwarded articles knowing they have been categorized into a bucket they care about (at least until tag spam causes this increased receptivity to decay).
Now, let's add a motivation layer. For example, let's assume that if you forward an article, or for that matter any other viral marketing element, you and the recipient are automatically entered in a sweepstakes drawing, or get a coupon. Providing context to the recipient also provides information to the sponsor/ marketer that allows these offers to be more closely tailored.
On a related front, my friend Bill Ives just signed on to help another friend, Peter Gloor, at Peter's social network analysis software firm iQuest Analytics.
I'd be interested in hearing about any applications people know of that demonstrate these ideas today.
