What newspapers can learn from the iTunes Genius button

Apple’s new music recommendation tool, the Genius button is….genius.  It’s not a new idea by any means - see Pandora, last.fm or even Amazon - but its so seamless, so easy and simple (in fact, you don’t have to DO anything) that it is already more useful and more widespread than any of the original pioneers.

Note: Of course, like any collaborative tool, Genius will get more genius as it learns more about users but its worth noting that even now it works pretty darn well….usually, at least…I just genius-ed a maudlin but amazing Bonnie ‘Prince’ Billy song and bizarrely got lots of fast-paced happy music from the the Pixies, the Yeah Yeah Yeahs and (huh!?) Bloc Party.

We at Shakeup always harp on about the value of personalization and recommendation to news sites but, the truth is, I still don’t see it done on any of the mainstream newspaper sites, at least not on any large scale.

News organizations clearly understand that their greatest asset is the huge and ever-expanding store of stories, photos, videos, etc that they control.  They are already starting, thanks to many an Internet evangelist, to look at this pile of information as a database that can be manipulated (for instance the now ubiquitous Most Popular/Most Commented/Most E-mailed lists).  But they haven’t quite taken the next step - matching this undifferentiated database with very individual readers.  That is, taking what they know about readers and their habits to suggest specific personalized stories, photos, etc, for them.

Again, not a new idea.  But the secret is to make it all seamless.  To make sure, like in iTunes, that users don’t have to actually do anything for this service.  They just have to interact with the site and, before they know it, they have the nice little surprise.

Another note:  On the difference between recommendation and personalization -  Personalization is what a user chooses to see, how a user chooses to cut the information available to him (ie. asking to see the weather on the frontpage, or a specific stock).  Recommendation is what a smart computer predicts a user will like (ie. a list of stories that match the reader’s history).  In combination with an editor’s intelligence, they are very powerful information systems that provide:  a) things I don’t know about but are important (what an editor chooses) b) things I don’t know about but will probably be interesting to me (recommendation) and c) things I am already interested in but want to learn more about (personalization).

Add comment September 17th, 2008