In October, I introduced you to Coke's Facial Profiler. Using Facebook Connect Coke was building a database of images to help match you with your twin. I sat fascinated by the sophisticated nature of the execution.

Screen shot 2009-12-06 at 11.03.49 PM

Phase II of the project launched last week. (http://apps.facebook.com/facialprofiler/)  Using the industry’s most advanced facial profiling software, the Facial Profiler matched me against a vast database of faces collected by Coke Zero (there are already over 105,000) to find my look-a-likes.  The results will were delivered soon after I accepted the FB connect prompt.

I found the details about my doppelganger based on the privacy settings for her Facebook profiles. I am still pondering whether I will connect with her.

There is a fun feature called the “Match Bank,” where users will be able to rate which matches they think are the best. By combining human feedback into the Facial Profiler equation, this will help improve the accuracy of future matches.

Check out a few of these matches:

Facial Profiler-Match 1    Facial Profiler-Match 2   

Facial Profiler-Match 3      Facial Profiler-Match 4

It's amazing how close this software gets.

While this seems to be a fun game - the use of my public information fascinates me. 

What is value of my public information to coke? Could executions like this lead to an advanced ways to connect me to the topics I love?  Could you use this type of personalization to put me closer to a product experience? Could the future of Social/CRM 2.0 lie in the a individual's public data?

Will the new digital divide be the difference between those of us with public data and those of us who have not shared their data?

ON: Coke’s Facial Profiler – Part 2 via @jpenabickley