Social Media, Yahoo and Data Mining

Introduction

Recently, Bambi Francisco of MarketWatch comments on the future of social media as they “search” (bad pun) out for an advertising model. Francisco sees a future in search query data and behavioral marketing:

“We just haven’t seen it work all that well because the companies that know about our search history — aren’t really sharing information with the companies, such as MySpace, Facebook and YouTube, that increasingly know us personally and increasingly occupy our time.”

Francisco’s commentary is not a revelation, but does give for a pause, especially when she casually brushes of the privacy concerns such data mining would bring:

“[M]arketers will increasingly take someone’s search history or search behavior, and use that information to target ads on social network pages they browse.

It’s not intrusive; in fact, people will come to expect it. And, quite frankly, some may even feel a little bit alive because they’re acknowledged, even if only by a lowly marketer.”

Interestingly Bambi Francisco talks about search engines and social networking sites as seperate companies, ignoring Yahoo’s powerful array of social networking and web 2.0 web sites (Flickr, Y!360, del.icio.us) – which is something we’ve been talking about at the company I work at.

Mining MySpace

While search queries presents great and freely volunteered information on the intent of the searcher, Francisco seems to pay insufficent attention in the inherent value of social network themselves.

Social networks – with all kinds of personal information disclosed freely and even sometimes displayed publicly – is the ideal data rich resource that can help take Behavioral Marketing to the next level. Shawn Gold, VP of MySpace, briefly alludes to this in a recent iBreakfast (via The Virtual Handshake):

87m stories in the database, and a lot of that content is professional. Every nightclub, every major Christian band, every celebrity brand, is in the database. They’re now slicing the database by professional type, e.g., if you want to reach all the comedians.

Sure, targeting comedians on MySpace is nothing compared to the possible social network analysis (SNA) and demographic data analysis possible. But its coming. SNA has been used in counter-terrorism experts and will soon be used to profile users based on their interests and those of their friends, with tools that automatically segment the wealth of user data on places like MySpace. This would invovle building behavioral profiles and purchasing habits of users for targeted marketing and product recommendations engine.
The Power of the Yahoo Network?

While the attention is on Google for privacy concerns and MySpace as the current king of social networking, let’s not forget Yahoo’s network of websites.

A month or so back, folks in my company discussed the power of Yahoo Network (based upon Yahoo’s Gay Pride minisite):

• Upcoming (Events)
• Flickr (Photos)
• Answer (Questions asked by Users)
• TripPlanner/Yahoo! Travel (Travel)
• Yahoo 360 (Social Network)
• MyWeb/Del.icio.us (bookmarking)
• Personals, Jobs, News, Local, etc

Between these network of websites, Yahoo could build a customer profiles allowing marketers to say, “let’s market to single, straight, Asian, Male, (Y! Personal/360) who vactioned in Europe (TripPlanner/Flickr/Y! Travel), went to a PR event in Palo Alto (Upcoming) and has been recently looking at jobs at the likes of Edelman and other PR companies (Y! Jobs)”. And Yahoo, would have all the information available to do this.

Indeed, today we hear the annoucement of Raghu Ramakrishnan joining Yahoo!, who will head Yahoo’s “social search”:

“At Yahoo you have this unique opportunity to integrate conventional search with Flickr, Del.icio.us, Yahoo Answers, Yahoo Groups and Yahoo Mail,” Ramakrishnan said, listing Yahoo’s services that center on human contributions. “How do you take all this search activity and learn from it?”

The question is: How long before Yahoo can successfully integrate its many web properties? How can they lure MySpace users away? What are the privacy implication of Yahoo having so much information on its figure tips and do they dare build a target ad network based on it?