Sunday, 11 July 2010

SOCIAL MEDIA GROWS UP BY GOING ALGORITHMIC

A recent piece from NMA: Social media is a catch-all term for a range of disciplines unceremoniously clumped together. I divide this world into two distinct types of activity: manual and algorithmic, writes Tony Effik.

Most of today’s social media is of the manual variety. All that activity on Twitter, those blogs, all those Facebook pages and updates, and all those sites that are open to comments and even votes are generally all manual, and usually co-ordinated by a community manager.

Social media in its current form is now over six years old but still lacks a killer business model. The goldrush currently around community management is interesting, as all brands will need some form of community management, which means there’s still more growth to come. Community management will be a key part of the killer business model, but it’s not the full answer.

During Web 1.0 it took nine years – from Netscape’s IPO in August 1995 to Google’s in August 2004 – before the claims made in dotcom era investment prospectuses where made true. Google’s innovation was moving search from a human-based indexing model to an algorithmic one, and combining it with a killer business model in AdWords.

There are a few great candidates in social media chasing the algorithmic model Twitter is now a real-time search engine. Rumour has it that Twitter is working on algorithms that not only rate tweets with a ‘resonance’ score, but also rates the pages they link to to create a PageRank-like algorithm.

Facebook’s recent launch of Open Graph is another great candidate. Through the ‘like’ buttons now appearing around the web, Facebook is creating algorithms that help users and brands with all sorts of matchmaking.

Demand Media is the smartest firm that no one has heard of. It produces 4,000 videos and articles every day based on an algorithm that identifies gaps between what people are searching for and the content already out there. All its sites have social features, but the really smart thing is that all its content is crowdsourced from a community of freelancers that it manages with a smart algorithmic process.

The different threads that make up social media are now separating. So in the same way that every firm needs to manage communities using manual models, they should also be thinking about how they might work with these new algorithms.

No comments: