February 19, 2010

How Social Networks are Disrupting Search

Yesterday I posed the question "is knowledge gained through social networks decreasing the number of web searches?"  Are people not searching for things they would have 2 years ago because of what they've learned through social networks?  The Compete data which shows Facebook is now outnumbering Google  in web referrals is making waves this week, and there have been anecdotal stories about similar data for the past several months.  But the data doesn't show whether Google is actually losing searches to Facebook, or if Facebook's traffic is additive.

There are two sets of user behaviors from which to approach this question.  The most obvious way, because it's how we've become accustomed to information retrieval on the web, is when a user asks an explicit question to their social network.  In other words, by measuring the frequency of questions posed to ones social network (Similar to the question I posed on Twitter about this very topic).

Here is some interesting data on phrases that would indicate a question:
From Tweet Volume:

TweetVolume : Home
Uploaded with plasq's Skitch!

From Trendistic:

Embed Trendistic chart in your site - Trendistic
Uploaded with plasq's Skitch!

Clearly, people have become accustomed to this behavior.  There are over 253 million tweets with a "?" according to the data above.

The other way, and not as obvious, but probably more powerful, is knowledge gained serendipitously through social networks.  "If the news is important it will find me." This concept isn't new.  Here's an analysis on this very question from 2002.  But now that people are using social networking in critical mass, I think it's safe to say this knowledge transfer is reducing the need for search.  By one count, there are over 500K unique urls shared per day on Twitter alone.  That's 182.5 Million per year.

This change in user behavior is critical for all types of organizations.  This is a real trend. and probably why there is a huge influx of "social media experts."  Google has been trying for years to enter into social networks and struggling to gain meaningful traction (Buzz, Wave, Orkut, Jaiku, Dodgeball ... I'd argue Reader a minor success, but Twitter is disrupting Reader).  Google is doing a decent job of integrating your social graph into web results, and has done a nice job with real time data, but this doesn't solve their issue with serendipitous knowledge.    

This trend will be gradual.  It doesn't by any stretch mean the demise of Google, but I do think the hyper-growth in search revenue will slow.  It also poses an interesting question for online advertising.  I've argued for years that advertising needs to become socially relevant (I give kudos to LinkedIn and Oodle for models that are starting to tackle this).  There are others who argue online advertising is a bubble, and Google will pay for it.  Regardless, organizations need to start reinventing the way they conduct business and make "social" engrained into all parts of their organizations, people, data, products, and not just through "media."  This new transparency will change how we think about "brands" and be more about "reputation."   After all...."There are no secrets.  The networked market knows more than companies do about their own products.  And whether the news is good or bad, the tell everyone." 

Update:  Sebastien Provencher just wrote similar blog post with recent research by MIT and Microsoft on querying your social network.  Will digest and perhaps expand in a new post.