Digg plans more categories and recommendations; could better targeted ads be next?
Speaking via video link, Digg-founder Kevin Rose gave a brief appearance at last week's NextWeb conference held in Amsterdam (see Read/WriteWeb's excellent report) in which he hinted at the social news site's plans over the "next six to twelve months".
In a further move away from just featuring technology news (the site has already added categories for Politics, Entertainment, and Sports) Digg wants to expand its remit so that users can submit and vote on almost anything. Examples mentioned by Rose include restaurant and product reviews, as well as the much requested 'images' section. Whilst it's very understandable from a commercial point-of-view why Digg would want to move further into the mainstream (it's tech-dominated user-base are renowned for not clicking on ads), simply strapping on new categories as a means to expansion -- aside from those where demand is explicit -- hasn't really worked in the past and I'm doubtful it will in the future.
The problem the site faces is that its community was originally built around a common interest in technology-related content, which then creates a self-fulliling prophecy (update: see Mahammad Saleem's theory of self selection). As the community grows, new users who submit and vote on content quickly learn Digg's cultural code and tailor their behaviour to meet that code. So for example, there is little point submitting a story outside of the core userbase's interests if it has no chance of attaining enough votes to hit the front page.
A far better strategy for Digg would be to roll out discrete social news sites which can harness the company's technology and brand, but in a way that breaks the link with its original user-base of tech geeks. As an example look at Techmeme's sister sites specialising in celebrity news and gossip and baseball.
The second element of Digg's plans -- personalized recommendations -- makes much more sense.
As R/WW reports:
With so many stories being submitted to Digg at the moment, they feel that it is getting harder for their users to find interesting articles that don't make it to the front page. They hope to look at what kind of stories you have previously dugg and suggest similar stories to you that they think you will like. This is in contrast to the current system of showing you stories that everyone likes.
This is exactly the kind of feature that I suggested in my post 'Five ways to make Digg more social', as it seems kind of a waste to not make better use of the tens of thousands of submissions every day, most of which don't make it to the front page. However, as Digg only employs broad categorization -- no free-form tagging for example -- I'm not sure how they'll achieve personalized recommendations. One way would be by looking for patterns of similar users i.e. user x regular votes on similar stories as user y, and to use that as the basis for recommending stories regardless of their front page status. It will be interesting to see how Digg solves this problem, as despite the omission of tagging, the site still has a lot of data from which to mine. This of course brings me to another potential feature. If the site can recommend content based on aggregated voting patterns, why not better targeted advertising and marketing? Digg's tech-savvy users might not click on the best that AdWords can deliver, but might they respond to personalized special offers? I'm not sure, but I think it's worth testing -- especially on an opt-in basis.
Related post: Five ways to make Digg more social