Google Pursues Patents to Monetize Social Networks
Is it a wonder they fail?
The US Patent and Trademark Office published a handful of patent applications from Google, betraying the search giant's ambition to monetize social networks.
Recent reports find 70% of online adults use social media, and a whopping 93% of online Americans expect companies to have a social media presence. But social networks have proven difficult for advertisers to monetize. If a traditional media site generates a 2% click-through rate on a good day, "on Facebook, you're lucky if you're going to get 0.3%," said CEO Jaffer Ali of online ad firm Vidsense.
Google's three proposed patents — Open Profile Content Identification, Custodian Based Content Identification, and Related Entity Content Identification — rely on language processing and other algorithmic features to locate data patterns in conversations, profiles and friend lists. If the technology is leveraged to serve better advertising, users will be more likely to click through (and ultimately purchase), reports IEEE Spectrum.
Open Profile and Custodian present a way to mine data from both a user's social networking profile, and the profile s/he is currently visiting. For example, linguistic cues or sentiments ("enjoy," "don't like," "hiking," "gardening") may be drawn from a profile that reads, "I really enjoy hiking, especially long hikes when you can camp out for a few days. Indoor activities don’t interest me at all, and I really don’t like boring outdoor activities like gardening." Based on that data, ads can be served as appropriate.
Meanwhile, Related Entity gleans data from lists of friends, or groups to which a user belongs.
CEO Hussein Fazal of agency AdParlor observed none of the patents pursue business strategies that haven't already been tried, but the language-processing and pattern-recognizing algorithms may be key to their success. For its part, Google has not disclosed the structure of its algorithms.
No comments:
Post a Comment