In an email thread about [Ajax] widgets, one super RedMonk friend brought up an excellent point:
I think you’re right that [cross-widget integration] should really be handled by a data service – e.g. the “Anne Zelenka presence awareness service”. We could imagine that in Theory-land [yuh!] where everyone’s identity and data is federated and queryable in a secure fashion, each widget could query the “where’s Anne?” remote service, whose data is provided by some GPS-enabled device you carry on your person. But I don’t think it would be the job for one widget to ask another widget for this information – it belongs in a universally accessible data service, i.e., a URL that returns a wad of meaningful XML (or JSON )
Needless to say, the idea of changing an identity store — or a “directory” — into an Internet hosted, URL-faced service that returns those wads got my heart racing. (Or was it that 3rd cup of coffee?)
More generally, it made me realize that I’m constantly looking for such a service online: somewhere that I can store all of my profile information for access by other services (web based and otherwise) to pull from. Of course, finding that service is only half the battle: other services need to start consuming it until whatever the user-centric, Identity 2.0 service achieve The Biggest Community wins tipping point.
What the Customer Wants
Ultimately, my end-user desires are simple:
- I want to stop entering the same profile information over-and-over again. I probably create a user account 2-3 times a week (this morning it was Dishola.com), adding in the same email, username, picture, and “about me” info.
- I’m sick of entering in my “friends” into every damn site I create a profile on. For social networking sites, this second item is one of those counter-intutive differentiators that most (all?) sites are missing out on.
Twitter is the simplified version of Dodgeball. Whereas dodgeball lets you update your location status, Twitter lets you update your over-all status. It’s a great example of how simplifying a concept can make it more fun and broadly useful. You can update your status through three different means: the web page, an IM client, or a text message on your phone.
Now, niceness of a generalized presence server aside, the annoying part of Twitter is that I have to once again invite all my “friends” to use it or find them in the system. It’d be much nicer if I could just suck all that info from dodgeball, LinkedIn, MySpace, facebook, upcoming, flickr, and whatever other system I’d linked up with other people in.
I don’t fancy sending out a spam message to all these people: instead what I’d like to do is find existing people in the system that I already know and link up with them. Some sites do this by allowing you to import an address book, but that’s always somewhat imprecise esp. if the matching is on email addresses (I used a different email address for every site for spam tracking). Not to mention you usually end up spamming a bunch of people.
Ze Frank’s the ORG is another social-networking-enabled site that’d be much better if I could suck in and find existing users. The great thing about the ORG (so far) is that it’s ignored the taboo of displaying your location to the world. Of course, you can opt-in and out of that.
What this taboo-breaking means is that once I tell it where I am, I get a list of people close to me. This is fascinating because there’s enough people in the ORG that there’s actually a few people in my neighborhood. Will I actually go meet them? Who knows? But it makes the site a lot more connected to the real world, instantly, than other social networking sites.
Consequently, I keep going back to the ORG to see what’s happening and who’s close to me. Of course, this is another instance where I’d like to bring over all the connections I have in other sites but, aside from sending out a spam email, there’s no way to do it.
Making My Life Easier
In both of the above cases, we have nice web sites that I’m liking using. To be fair, they’re also both fairly new and haven’t quite entered the luxurious position of adding in gold-plating.
That said, this problem of disconnected web applications has existing forever. When it comes to making connections transportable, there’s even a spec/microformat, XFN that’s been around for-equally-ever. For some reason, sites don’t seem to use it. Not having worked directly on that kind of code, I don’t know why. Is it not useful enough? Does no one care? Is it the old cliché of customers not asking for it?
OpenID has been a promising bundle of technology and standards over the past year as well. I’ve been delighted to see sites of like LiveJournal and claimID adopting OpenID and I’m waiting to see the attribute (or “profile sharing”) parts evolve and get more use by other sites.
Mega-money sites whose core business is providing its users with a social networking service, like MySpace, have almost no motivation, ostensibly, to open up a user’s information. Sites like eBay that relies on keeping reputation data in a roach motel have little motivation as well.
Smaller sites like the ORG, Twitter, and even flickr would be so much more the awesome if I could bring in and export not only my profile information, but also the existing networks I have. Other sites, like Netflix, whose core business has little to do with the social networking aspects, would benefit as well. I have a limited set of people added as my “friend” in Netflix, but I know there’s a ton more users out there who I could hook up with. And, with more friends comes more movie recommendations, which brings more use of NetFlix.
In addition to the initial import, I’d like to see these sights unexpectedly delight me by finding new “friends” that come along. As an analog, the music event site Tourb.us monitors my Last.fm data for bands I like and sends me an update (via RSS or email) when one of them is coming to town. Speaking of Last.fm data, their neighbor’s feature is a great example of the unexpected delight I’m talking about.
Still, there’s the issue of locking down the community by not allowing other sites to poach the profile and social networking data. At the moment, I have nothing but blind faith in Metcalfe’s law that expanding the networking beyond the site’s domain name will, ultimately, increase the over-all value for both the users and the company/web app. If users could glide between different sites would there really be the danger of “loosing” users to other sites?