“While these case studies are good voyeuristic fodder, snooping through one user's life barely scratches the surface of this data trove,” Splunk IT consultant Paul Boutin said. “AOL's 36 million log entries might look like an Orwellian nightmare to you, but for us it's a user transaction case study to die for.”
Using the raw data that inadvertently slipped out to the public, Boutin crunched the numbers to come up with seven distinct profiles of Internet users.
“The Pornhound,” Boutin said. “Big surprise, there are millions of searches for mind-bendingly kinky stuff.”
While nobody — least of all adult webmasters — would be surprised that many surfers browse the Internet for porn, Boutin’s unique bird’s eye view revealed some interesting insights.
“I found that porn searchers vary not only by what they search for, but when they search for it,” Boutin said. “Some users are on a quest for pornography at all hours, seeking little else from AOL. Another subgroup search only within reliable time slots.”
Boutins also found that a number of porn searches suffered from numerous spelling mistakes.
“An important related discovery,” Boutins said. “Nobody knows how to spell ‘bestiality.’”
Boutins also chronicled six other types, including the “Manhunter,” who searches for names; the “Shopper,” who he believes is more predisposed to bargain hunting than window shopping” and The Obsessive, who enters the same queries over and over again.
Boutins also concluded that the so-called Omnivore couldn’t be gamed, because that profile searches too broad a base of topics. Although he concluded that the Omnivore is likely to be someone who spends an inordinate amount of time online.
At the other end of the online user spectrum, Boutins found the Newbie, whose search queries reveal a lack of familiarity with the Internet. According to Boutins, users in this group use the AOL search box to look for “www.google.”
The final profile identified by Boutins is the so-called Basket Case, who types long sentences into the search box that typically reveal more about how the user is feeling — for example: “I’m sad” — rather than what they’re looking for.
Boutins published his findings in an article for Slate.
