How Google Has Screwed Up the MySpace Deal
The article cites Google's deal with the number on social networking site which allows Google to serve ads on the site through 2010, while agreeing to share-revenue with MySpace to the tune of $900 million. However, Google has to pay up regardless of whether it is able to meet its minimum ad goals.
However, given the uncharted and largely untested nature of advertising through social networking sites, some don't believe that simply offering ads alongside 'related' user-generated content will be particularly valuable to advertisers:
"You can't put up contextual ads against user-generated content," Monfried says. "It's irrelevant, and advertisers don't want to risk their brands on user-generated content."
Instead, Andy Monfried, formerly of Advertising.com and now CEO of Lotame, suggests that the best way to capitalize on social networks is to identify those who hold the most sway in the online world.
Monfried says Lotame dramatically improves targeting for advertisers by doing things such as identifying the "influencers." Monfried was reluctant to disclose too many details on how Lotame's solution varies from Google's ad-targeting platform, but says his company is focused on "verbs and actions." So instead of serving ads based on the text in users' profiles, Lotame focuses on users' actions and demographic data...
This model seems reminiscent of Facebook's Social Ads feature, but rather than simply targeting users through friends (i.e. turning peer pressure into ad money), such a system seems to indicate that it would weight users' value as advertising conduits based on the position they hold in the network.
In the article "Social Networks and Group Formation" (in boxes and arrows, available in coursebook), the author notes that social network members can be characterized by their degree (the number of users they are connected to), control (the extent to which others depend on a node for information flow/connections to other nodes), and independence (the extent to which a node is closely related to other nodes considered, such "that it is minimally dependent on any single node." Perhaps by algorithimically identifying those nodes with the greatest degree and/or centrality, Monfried and Lotame can inject value into social networking ads.
However, the article doesn't address the potential problems with targeted advertising in social networks. As Facebook discovered with its launch of Beacon, users are not always thrilled to have their activity tracked and used for advertising purposes. Moreover, how would some users feel about being the particular focus of such advertising techniques? How do you capitalize on a users' degree or centrality without compromising that users' experience on the social networking site? With facebook we now see that users have the ability to opt out of such advertising schemes... does Lotame's model hold up if the networking site decides to relinquish control?
To be fair, Lotame is currently only working with smaller niche networks, not the big three. Perhaps it is easier to integrate advertising more seamlessly when users are less diverse and less numerous.
Still, the article makes a good point: Social networking is not search, and Google would be wrong to blindly follow its search-based ad model. Facebook's Social Ads claims to be able to reach users "before they search". I'm not sure networking is in any real way antecedent to search, but it is most definitely different... and it seems likely Google will realize this as it navigates its ad deal with MySpace:
"Google does a great job of monetizing intent," says Ray Valdes, an analyst at Gartner Research. "It knows what I'm searching for and it can show me relevant ads. But social networks are not about intent."
Valdes argues that its unfair to slam Google's approach -- after all, the company's core business is search, not social media. And to be fair, Google said on yesterday's earnings conference call that they're trying new things -- including demographic targeting -- and that they've seen some progress.
No comments:
Post a Comment