Gravity, which calls itself the “Interest Graph Company,” delivers personalized content to visitors of sites like The Wall Street Journal, TIME and TechCrunch.
Specifically, Gravity automatically reorganizes and re-emphasizes the various sections of its partner sites — in the case of WSJ, for example, Tech, Markets, Real Estate, U.S., World — to highlight those topics and stories which it determines individual users will prefer.
For the first time, the company has shared its initial user engagement statistics, comprising data from the past six months. The results, if accurate and sustainable, seem impressive.
Gravity reports that is has delivered an average of over one million personalized content recommendations per day to more than 400 million users per month over 10 sites initially. Across these partner sites, they say that click through rates have grown two to three times. Site loyalty, as measured by monthly return rate, is up three fold. Finally — and this is a hell of a claim — its partners have seen page views increase more than 40 percent on average. (Although it’s likely that certain sites in the network are skewing that number: a source with knowledge of traffic at one of the larger sites said that Gravity’s effects on pageviews had been somewhat less than this lofty figure)
The numbers may be a publisher’s or marketer’s dream, but still Gravity’s premise invokes the type of big brother sentiment that many instinctually distrust.
It’s likely that many users were, at best, only loosely aware that they were receiving personalized content. Once the word gets out, we can expect many to be upset at learning the depth of it. If you listen closely, you might even hear the cries of those fearful about privacy and excessive filtering.
Acknowledging the creepy factor, however, the explosion of information on the web is only intensifying, and the Gravity’s early results do suggest a compelling solution. It seems that Gravity not only benefits its website partners, but enables users to find and engage with more personally relevant content and they seem happy to return often for another fix. Whether the world is truly ready for this service, is another matter entirely.
The company admits that there is a fine line to be walked around these issues and has taken steps to keep personalization within users’ comfort levels, with the aim of enabling wide adoption. Every website implementing Gravity’s technology includes links acknowledging and explaining the use of personalization. Users are given the ability to opt out completely with this status applying across all sites in Gravity’s network (more granular control is on its way). Gravity also does not keep records of a user’s behavior or online activities.
Instead, the company only records the number of interactions and the respective strength of their associations with each particular interest category. All data collected is stored securely and anonymously. In the near future, Gravity plans to allow users to view and edit their personal Interest Graphs. This step, if executed properly, could be the feature necessary to persuade users accept a personalized Web.
Many will argue that this should be an opt-in only service. Based on the reactions seen to past user tracking and information filtering efforts, this certainly would have been the safe route. Nonetheless, the company reports opt-out rates in the low single-digit percentages. It seems that with a problem unpleasant enough, and a sufficiently satisfying solution, forgiveness may not be so difficult to come by.
Gravity’s technology offers its solution to the information explosion by combining knowledge gained through a user’s browsing activity within its network of partner sites, with that user’s social graph including comments, likes, shares, Tweets, and recommendations by friends. The resulting signals are filtered to determine not only what its users read, search for, and share, but interpret the meaning and significance of this information. The end result is a granular, highly personalized, real-time evolving, and portable graph of interests and preferences: the Interest Graph.
Gravity CEO, Amit Kapur, explained his belief that users will ultimately embrace Interest Graphs by way of an example from his early days as COO of MySpace. Initially, he said, users were reluctant to post profile pictures. Gradually, after finding that they resulted in more rewarding connections, most grew comfortable with the idea. Years later, we now live in a world largely unashamed to share the most personal details of our lives through social networks.
Kapur sees parallels between this cost/benefit-based adoption curve and the one experienced by users of Gravity. And in the case of Gravity, no one sees these preferences and interests but the individual user.
Gravity plans to significantly grow its partnerships across the Web in the coming months. The company is in the early stages of creating a browser plugin which users may choose to install to broaden and deepen their online personalization experience. As Gravity grows, it plans to begin delivering interest-targeted ads. With its technology and user interest data, these ads can be expected to be among the most relevant on the Web.
Lastly, Kapur teased of a future where a “Gravity portal” may flip online discovery idea on its head. Instead of the user actively searching for content, or passively discovering it through social recommendations, Gravity could one day use personal Interest Graphs to push hyper-relevant content directly to the user in real time.
It’s not quite the Matrix, yet, but if Gravity one day offers the ability to plug a personalized stream of the Internet into the back of users’ heads, there might be a long line. Don’t be surprised though, if across the street there’s also an angry mob with picket signs.