Search Insider: Does More Relevance Always Equal Less Privacy?
Does More Relevance Always Equal Less Privacy? by Kaila Colbin , Monday, March 16, 2009
You cannot have been surprised, in any way, by Google's announcement last Wednesday that it was launching a behavioral targeting program. After all, the company has masses of historical data to work with, has unfettered access to an extensive content network of proprietary properties and AdSense sites, and is the proud owner of a company that earned its $3.1 billion price tag via -- you guessed it -- behavioral targeting.
Again unsurprisingly, the pro vs. con tennis rally began almost immediately following the announcement. Google's affirmative defense is transparency: we show you what categories we put you in, we let you create profiles and we let you decide whether or not to participate.
Privacy advocates, on the other hand, believe that this defense is undone by the opt-out model being employed. Jeffrey Chester, the executive director of the Center for Digital Democracy, calls opt-out "a very incomplete and flawed safeguard." Ari Schwartz, of the Center for Democracy and Technology, concurs, saying "the opt-out is based on a failed premise... these plug-in options just serve to confuse users about what they need to do to protect themselves." Others have been less tactful. Marc Rotenberg, executive director of the Electronic Privacy Information Center, calls the program "a disaster."
In reality, the core of the problem isn't the opt-in or the opt-out; it's what gets collected, what gets stored, what gets known about you. Opt-out is only an issue because of the sensitivity of the information that can be collected if you remain "opted-in." What makes it worse is the feeling that an analysis of historical data isn't the best method to predict behavior; it's simply the most widely accepted method right now. As one commenter over on PC World put it, behavioral datamining is "studying the corn kernels in somebody's p00p, then trying to guess what they will decide to eat for lunch."
Behavioral prediction based on historical data must produce results or nobody would be interested, but it's not the only option. Along with Google's BT program, this week saw the launch of VortexDNA's Web Genome Project, a movement to map the Web. (Disclosure: I blog for them am and a minority shareholder. Are you listening, David Koretz?)
The WGP is built around an interactive search tool, but it isn't aiming for the search market. Its key proposition is non-historical personalization: relevant content based on an aggregate and continually updated "genome," rather than on a collection of data about which sites you've visited.
Google is not the only one offering behavioral targeting, and no doubt the Web Genome Project isn't the only endeavor to come up with an alternative. But I do get the sense that the current BT efforts of the big players are based on a flawed axiom: the more we follow you, the better. Given that, the importance of the WGP and others like it is to open the door to the idea that there are alternatives.
We've been constrained for too long by the assumption that More Relevance = Less Privacy. Come on, guys! Instead of a tug-of-war over that win-lose proposition, let's focus instead on a new equation: More Relevance + More Privacy = Happier Netizens.
After all, we're the Internet Generation. Solving problems like this is in our DNA. Don't you think we can do it?
You are receiving this newsletter at firstname.lastname@example.org as part of your membership with MediaPost. If this issue was forwarded to you and you would like to begin receiving a copy of your own, please visit our site - www.mediapost.com - and become a complimentary member. For advertising opportunities see our online media kit. If you'd rather not receive this newsletter in the future click here. We welcome and appreciate forwarding of our newsletters in their entirety or in part with proper attribution. (c) 2009 MediaPost Communications, 1140 Broadway, 4th Floor, New York, NY 10001