Influentials and the Social Graph

Image courtesy

In amongst all the speculative noise surrounding Microhoo (or is it Yahoosoft?), there was an announcement from Google which barely got a mention on ad blogs yet represented a significant step towards empowering communications which capitalise on connections on the social web.

Google launched their Social Graph API (under the heading URL's are people too) which uses the same algorithms from the search engine to discover how people are connected across the internet. The API works by searching for connections between people based on how everyone is linked on social networks and via publicly available profiles and pages.

Based on open standards, and the indexing of two types of commonly used link tags - Friends Network (XFN) and Friend of a Friend (FOAF) markup - the API looks for two types of publicly declared connections:

1. All public URLs that belong to you (your blog, your Twitter account, your LiveJournal page) and that are interconnected.
2. Publicly declared connections between people (blogrolls, social network 'friends' etc)

At it's most basic, the API allows websites (and brands) to make it easy for users to locate and add their friends when starting up at a new social application. Meaning that when signing up for a new social service, it not only allows me to easily feed it links to my social presence on the web (blog, twitter, dopplr, delicious), it will analyse the onward connections and network to suggest to me friends that I might want to add on the new service.

The API uses publicly available data but that doesn't mean that there aren't issues to be ironed out around how the data is applied. But there's no doubt that the ability to identify and tap into the social infrastructure of the web across multiple sites and applications is a very powerful thing. It's not only a major step forward for data portability, it's a major step forward for social media. And it sets us down a path where it becomes ever more possible for users to allow easy access into their social graph. Tim O'Reilly called it a "huge step forward" towards his vision of an internet operating system (his original vision of a web built on open architecture where simple programs could be connected to accomplish more complex tasks). He even called it "REALLY cool". 

For brands, I think this has the potential to transform communications that use social media. But I also think it throws a whole new light on the debate and controversy surrounding Duncan Watts' work around influentials. Billed as anti-Gladwellian, I actually think that the views of Watts and  Gladwell are not wholly mutually exclusive. I don't know why anyone ever expected word of mouth to be easy. Given the complexity of human nature, the effective spread of ideas is inevitably going to be dependent on a highly intricate and interdependent set of multiple features including the idea itself, it's portability, the environment into which it's born, who seeds it. And the nature of influence is probably even harder to pin down.

But I do buy into the general concept that some people, for some things, will inevitably be more influential than others. Finding them is the (extremely) hard part. The social web may well provide a small part of the answer.  If the foundations of the web are sharing and openness, then the currency of the web is very much about links. As Scott Karp says, the link is the principal driver of network effects and influence:  "Influence on the web is all about connectivity - the larger the network, the more powerful the links". If we are better able to understand and tap into the social infrastructure created by these links, brands will be better able to use social media in a constructive and useful way. In launching their new ad platforms, Facebook have attempted to allow brands to harness the power of our social connections. The difference with Google's API is that it has the potential to do this not just for one social network, but for the whole web.

Comments

[Stan Lee]

Missed that announcement completely. Thanx for the post and for explaining it so clearly Neil.

[neilperkin]

No worries Stan. Glad it was useful :o)

[Dino]

Excellent post Neil.
One thing that gets me is that the word-of-mouth industry insists on overly simplifying how networks work. As you rightly point out, whoever said that this would be easy?!