1 post tagged “filters”
Digital Identity
(First posted September 2005 on Bloglines)
The viability of the long tail - and thus services like Amazon, depends on the ability of people to navigate their way from popular services to niche ones without being paralysed by choice. This can be done in many different ways. Some involve recommendations from friends either offline or on blogs or through email or social networking services. However Google and Amazon (and others) use algorithms which say "you clicked on/typed this, thus you might like that" to enable people to travel down the tail to more obscure content.
There are two ways in which this is a problem. First, the recommendations are too broad. They don’t let people compartmentalise their lives; to keep their fun separate from their serious pursuits. Second, they can be invasive and can stop people feeling secure and in control of how their information is being used.
Users like it when they can take control of their digital footprints. If you give people some agency and investment in the process they tend to respond well. Witness the difference in the reception given to Amazon where users can create their own wishlist and Gmail which reads your email automatically. The former was hailed as useful and fun while the latter was branded an invasion of privacy.
Search and e-commerce services should enable users to have at least 4 digital identities with different levels of privacy. Their professional ID, their personal ID, their private ID, and a non-traceable ID. This should be built into the service such that if a service provider like Amazon or Google is saving information from clickstreams, that information will be linked to that particular ID.
Benefits:
- Reflects how people are aspirational - who we are, how we see ourselves and how others see us are three different things. This allows the info that we want out to get out, while quarantining things we find embarrassing.
- Enables people to feel empowered, not violated.
- More accurate results – the ID being used is another set of metadata for the search engine or e-commerce site to use. A person whose professional ID is full of medical clickstreams is more likely to want medical stuff when using their professional ID.
Challenges:
The idea of a unified method of aggregating the various different ID’s on the many online services is required. This is the subject of Dick Hardt’s presentation which is attached to this post.
Update: 25th July 2006
It is great to see that user content driven services like flickr, vox and videoegg are allowing people to set different levels of privacy on their content.
