Friday, January 12, 2007

Sacrificing privacy for personalisation

Those recommended books that pop up on Amazon; those Just for You tracks suggested by the iTunes Music Store- they can be a little spooky at first. How did they know I'd love Richard Dawkins athiest polemic, or track 6 from the Punk Rock Power Ballads compilation?

Fans of the popular social-music application LastFM will be familiar with intelligent recommendation systems of this sort, which compare an individual user's inputs with the historic inputs of a huge database of users to generate related results that are likely to be of interest to that individual. For more on social-music, check out Johnny Dee's article in The Guardian Guide, which offers a neat overview of the movement and its key protagonists.

Recommendation services are just one example of the personalisation trend that's sweeping the web. And not before time. Left unfiltered, the Long Tail dream can quickly turn into a nightmare of excess content and choice-anxiety. Hence, as we're slowly coming to realise, the Amazons and iTunes of this world are doing us a favour. They pick strands from the Tail that suit our tastes and save our sanity. (Obviously there's a sales incentive for these retail behemoths, and we will increasingly find that they pursue us with so-called 'remarketing' tactics, but for now, at least, they're kind of on our side.)

And you know what? Those recommendations are pretty good. They're not always right, nor anywhere near. But they are right perhaps 3 or 4 times out of 10, which is enough to be useful. Moreover, this ratio will only get better as systems swell with ever more hard data on the likes and dislikes of people like you.

In his new year article for Ad Age, Steve Rubel predicts a change in how we perceive and manipulate the web. Whereas 2006 saw the continued explosion of web content, 2007 will be the year of a great implosion, Rubel predicts. Whilst the amount and variety of content available will continue to grow, increasingly sophisticated micro-chunking technologies like RSS will allow us to select just those slithers of content that really interest us.

There is a catch. In order to benefit from recommendation services and micro-chunking technologies, we consumers must give up a little of our privacy. By browsing our personalised picks, we tacitly agree to the exploitation of past click-streams. To solicit niche content one must volunteer personal details, and deal in the revealing currency of tags and keywords.

It seems that we are increasingly happy with this settlement. As eMarketer reports, both our positive view of personalisation and our hunger for help in tackling the Long Tail, even at the cost of privacy, are reflected in the results of a recent ChoiceStream survey. 79% of US adults who responded to the survey expressed an interest in receiving personalised content (83% amongst the Diginative-biased 18-34 age group, presumably due to their increased familiarity with personalisation, and their grasp of its benefits). And as the tables below demonstrate, the willingness of these adults to exchange personal details and click-streams in order to gain the desired personalised content has risen significantly since 2005:


Most interestingly for marketers, more than a third of all US adults and a majority of the 18-34 age-group profess their enthusiasm for personalisation in the realm of advertising:




This widespread pre-acceptance of personalised ads is a surprise. One might have expected the increased salience of civil liberties issues to prevent digital marketers from exploiting the inherent bidirectionality and addressability of some digital technologies to deliver personalised communications.

The special enthusiasm of Diginatives for personalisation is particularly pleasing, and allows marketers to begin planning for a super-targeted future.

No comments: