SezWho.com Homepage

We live in a page-rank world. Google the main organizer and cataloger of the internet, uses Page-Rank as the primary way to organize information (I know I am oversimplifying here as Google uses a number of other algorithms as well but link structure between sites and pages is still one of the most important factor). Even blog search engines like Technorati and Sphere use a derivative of page-rank algorithm to rank the content. At a high level what that means is that content on a site derives its credibility from the credibility of the site. E.g. if there is a page on cnn.com, it inherits the page-rank from the site. Now if there is a page that has same information and is on a site with a lower page-rank than cnn.com, the page will be considered less credible and show up lower in the Google search results. The idea of deriving credibility of the content from the site made a lot of sense when there were editorial boards and organizations to ensure everything was vetted, reviewed and solid. But does it still make sense in the evolving social media landscape?

Let’s take an example. Let’s say there is a video on YouTube. Should the fact that the video happens to be on a popular site make it more credible? You can be sure that the staff at YouTube has not reviewed the video to ensure the credibility of the content…In such situations does it still make sense to use a page-rank based mechanism to evaluate the credibility of the content? Clearly with user generated content the credibility of the content cannot be derived from the credibility of the site, instead the credibility has to come from other source. How about users who are generating the content? How about the consumers of that content?

Its all about the people

When sites become a two conversation (ReadWrite) and when everybody has access to the means of publishing content, and has the potential to get immediate, unlimited distribution, as is the case with social media, the ranking of the site become meaningless in determining the quality of the content. This is a change for the Internet but in the real world, that is really how things work. E.g. in a meeting, a conference or a social gathering, people take into account the credibility of the person who is speaking to determine what to make of it. In other words, who is delivering the message is almost as important as what is being delivered. Now that Internet is enabling a global conversation, we need to go back to the same people-based credibility model to evaluate the content that is generated by users.

Let’s go back to our earlier example to see how it can work. Instead of using the site based credibility, suppose there was a way to establish that a particular user has spend some time thinking about the topic and has posted some interesting thoughts on the subject on his/her blog. Wouldn’t that make you more likely to watch the new YouTube Video?

This is all good but how?

One of the key strengths of social media is that users have the means of producing and publishing content. This also means the conversation on any topic span multiple sites. While this provides a great deal of flexibility to users it also makes it really hard for any particular site to provide enough of user context to make their content credible. Even a popular site like YouTube can only show what other videos a user has published, but what if the user has only a few videos on YouTube and the rest of the context is in the form of Flickr pictures, blog posts/comments and forum discussions? YouTube will not be able to show that context for the users and the content is going to become less interesting as a result.

Another problem with building a people based credibility framework has been that there has been now way to establish people’s identity. This is an artifact of the evolution of the web where initially the focus was on sites and the organization principle was page-rank. Lack of a universal identity mechanism prevents sites, even though community clearly benefits from such context, from putting together cross-site user profiles. Right now there are a number of efforts like OpenID, card-space that are underway to establish a universal distributed framework using which applications can establish user’s identity. The issue though is that these frameworks are still in their infancy and still a few years away from critical mass. So in the meantime, how do we proceed?

Rise of Community-Rank

One of the key ingredients that has been that have not been leveraged this far is the incentives for participants to identify themselves and be known as a good member of the community. There are a number of members in each community that are serious participants and would be happy to be rewarded in terms of recognition for their value-added participation. What if there was system that enabled users to build and control inter-site and intra-sites participation profile. Such a system will have to allow full user control over the profiles and provide mechanism to users to have as many identities as they want (let’s face is – all of us have multiple identities both in real and virtual worlds). Much like real world, in such a system, community will be able to reward users for participating well and punish those users that don’t. Let’s can this system a community-rank and identity system.

Using community-rank and identity system, readers of social media sites will be able to establish participation history of a user, understand what the community thinks of the user’s content and easily find the most credible content. For search, community-rank will lessen the reliance on site context and put the focus on the community reputation of the people generating the content.

But what about privacy?  There are always risks when you start organizing information around users and their participation in communities. While a system like this benefits the community as a whole, some of the users might not want to have participation profile. To address these concerns such a system will need to provide full user control on the profile information. In addition, it should allow users to be anonymous if they want their contributions to not be a part of their profile. By addressing some of the privacy concerns, such a system can really help improve the quality of conversation in communities.

ConclusionPage rank based organization is not suited for social media site (you just have to go and search in a discussion forum to realize that things don’t work as well as you would like). A community rank and identity system has the potential to unlock huge amount of value in social media by incentivizing participation and by empowering readers.