In response to my post from earlier today about whether AOL's unfortunate release of identity-revealing search date could spark a response from the plaintiffs' bar, Raul Valdes-Perez, CEO of enterprise search solution provider Vivisimo, wrote to me with the following e-mail (edit):
I read with keen interest your cited ZDNet article....The lamentable release of AOL user search data, the recent wrangling between several search engines and the U.S. government over search data, and the less-noticed Google Trends for searching user query logs, are just three examples of a trend that deserves serious scrutiny...The problem is that user data is profitable, e.g., for targeting ads ever more widely & finely, so data gets collected in ever more detail, leading to "personalized search" which I've argued is a dead-end for web searchers....One recourse is legislation, or lawsuits as you point out. Another is for web search engines to take a stand on privacy. In January of this year, we at Clusty.com [humoroursly] took our stand [on privacy].
I called Valdes-Perez because, after reading his letter and clicking through to some of the links, it seemed like he was saying that not only are these databases fool's gold in terms of the personalized search results they're designed to deliver, but that that fool's gold probably cuts both ways. In other words, if the databases are failing to deliver the improved search results that supposedly justifies their existence (according to the search engines that keep them), then they must be failing at targeting from an advertising perspective as well. I thought the last comment on John Battelle's blog about Valdes-Perez's position paper on the issue summed it up rather perfectly if you consider the way our TiVo boxes might one day be narrowing down the "TV guide" we see while at the same time delivering more targeted advertising based some database analytics being performed somewhere:
"My TiVo thinks I'm gay"
It's a great point. Just because I may end up recording or "on-demanding" Queer as Folk and Brokeback Mountain, does that mean that my cable TV provider can start to make certain assumptions about me, the additional programming I might want to see, and how to target advertising into that programming (not to mention how search on those titles could influence future search engine results)?
Valdes-Perez said he definitely intended to say that personalization is a Holy Grail not worth chasing but said he'd need to think about the ramifications for contextual advertising before he could respond to that part of the question. He noted that search companies are making a profit off of targeted advertising. But maybe that's just because the targeting, as sub-par as it is, is the best we've got so far. More to Valdes-Perez' in the context of the AOL situation, his position is basically that if search companies give up on keeping huge databases in search of an unattainable Holy Grail, then those search companies get the added benefit of not regrettably releasing data that shouldn't be released.
Valdes-Perez practices what he preaches. Vivisimo uses its own solutions to power the Internet search site Clusty.com (a meta-search engine). I tried it. Not only does meta-search engine usage provide a degree of insulation from working with a MSN or a Yahoo (if privacy and subpoena's are the sort of thing that concerns you), Clusty is very cool in the way that it clusters results. Clusty is equally notable for what it doesn't do. According to Valdes-Perez, it doesn't track users. According to a privacy abstract on Clusty.com;
Now search engines want to know you very well indeed: your queries, the pages you visit, the books you buy, the email you send, your age, sex, zip code, etc. etc. This new world is an inviting target of snoops or agencies that want to analyze, censor, or monitor you....We at Clusty don't track you. Our toolbar doesn't track you. We don't want to know your email address....Just search, all the time. No questions asked.
But Clusty's official privacy statement offers more specifics on the data the site collects saying:
When you use Vivísimo Web Search Services, our server logs may include information such as your web request, Internet Protocol (IP) address, browser type, browser language, the date and time of your request and cookies. We use such information to process your inquiries, respond to your requests, and improve the Vivísimo Web Search Services.
According to Valdes-Perez, the data clusty collects and the cookies it sets is not for the tracking of users and improvement of the search results themselves, but rather, just the sort of stuff needed to preserve the end-user's preferences when working with the search engine through a browser. That said, there are other meta-search engines that cater to the privacy-conscious, perhaps with less monitoring overhead than what Clusty.com does. Earlier this summer, meta-search provider ixquick.com announced that it would be permanently deleting all personal search details gleaned from its users from its log files including user IDs and IP addresses. According to an ixquick press release,
[Users] will be able to search using the 11 best search engines without their personal data being recorded....If the data is not stored, users privacy can't be breached...
Outfits like Clusty and ixquick are not alone. I'm sure there are tens if not hundreds or thousands of meta-search sites that offer an added measure of privacy to Internet searching and one thing is for sure. Now that the AOL watershed data revelation is behind us, you can expect more of these guys to come out of the woodwork (just like both did today, in my e-mail).