If you could capture and use the names, ages, genders and demonstrated interests of the specific people who visited your website - would you? A whole lot of people providing services online would. While we've covered the movement for standards-based Data Portability a lot here lately, the newly announced
MyBlogLog API is an alternative path to similar ends being taken by a proprietary company. Announced last week just before the Yahoo! OpenID announcement, the
MyBlogLog API could end up being of even greater importance. Really.
The Significance of MyBlogLog
MyBlogLog is an incredible model of how to leverage human psychology in order to access peoples' data. The prospect of seeing the faces of people who visit your blog is so seductive that thousands if not millions of people have offered their faces and information to MyBlogLog in order to participate. Compare this to the staid pitch of OpenID, where the "utility" of single sign-on has been the most clearly articulated value proposition so far.
By acquiring this ingenious service, Yahoo! has gained access to the browsing history of scores of users and the traffic history of scores of websites. You can imagine the value to Google of offering Google Analytics? MyBlogLog is even better for Yahoo! The potential to spread the service is big; see the MyBlogLog implementation on the Yahoo! site
People of the Web, for example.
The API
For now the MyBlogLog API is in private beta and only a select few developers have been given access. Two who have blogged about it so far include Aldon Hynes and Yahoo! employee Kent Brewster.
Developers will be able to use the API to access a variety of information about MyBlogUsers, including:
- User IDs for recent visitors to a site with MyBlogLog on it.
- Those users' contacts on MyBlogLog
User IDs on other sites they've tied themselves to, like Twitter, Del.icio.us, Flickr and many, many more.
Once you've got access to that information, it should be relatively trivial to access another layer of information that includes:
- The ages, genders and locations for site visitors who have exposed that information to MyBlogLog
- The interests all around the web of visitors to a particular site, as exhibited by their tags on sites like Del.icio.us and Flickr
The other sites commonly visited by said users.
Some additional layers of information may be inaccessible if Yahoo! actively shuts down access to data extraction tools, and much of the above may have been accessible already, in theory, using some of those same tools. It's hard to know yet what will be possible. Much of the above information is probably available for human eyes for premium account holders with MyBlogLog, too, but making it machine readable with an API is taking things to the next level.
Attention Data
I really like the idea of being able to leverage my Attention Data, the information that can be gleaned from my activities on various sites around the web. I'm hoping that the growth of the Data Portability Work Group will help move the proposed standard for Attention Data, APML (Attention Profile Markup Language) forward from the early stage it's at now.
Be that as it may, Attention is hard to explain and has only just begun to move forward in any kind of portable way. Enter the MyBlogLog API. I can opt-in to expose my activities around the web to MyBlogLog and then out to other sites that leverage the MyBlogLog API. There may be little reason to use APML if a website can offer personalized content based on a user's tags in Del.icio.us, Flickr, YouTube, etc. all via one easy to use service, MyBlogLog. It's not open and it's not as democratic as a distributed, portable, standards based approach would be - but it's also for better baked and will soon be ready to use.
MyBlogLog likes to play fast and loose with cookies, their feature that lets blog owners display MBL avatars for commenters for example is very cool. There's no reason to believe that in the coming months sites around the web with MyBlogLog installed won't automatically check to see if you are logged into MBL and offer you customized in-house content based on the tags used by you and your friends (as defined in MBL) on other sites around the web. One place to watch for mashups with MBL as they come up will be John Musser's ProgrammableWeb.
This is my prediction for the significance of both Yahoo! OpenID and the MyBlogLog API: OpenID won't catch fire with mainstream web users but Yahoo! (Open) ID may and Attention Data may never be free or widely understood but the use of MyBlogLog to broker our data could catch on very easily. To be honest I hope that's not the way it goes down but it is going to be very interesting to see. I've said it before and I'll say it again: however much money Yahoo! paid for MyBlogLog, it was money well spent. Yahoo! is positioning itself to become an even bigger broker for identity and Attention around the web than it already is.
Rajnish Kumar