google adsense by Genetic algorithm?

Google Targeting Ads Based on Previous Queries

Reuters reports from a briefing for journalists at the Googleplex, where Google’s Susan Wojcicki (pictured) talked about how Google is a bit wary to use behavioral targeting... that is, to deliver ads based on a deep user profile aggregated over time (and potentially, through different services). Instead, Google prefers targeting ads to the “task” at hand – the current search query. But Google is now also playing with targeting ads based on immediately previously entered queries (the “search session”):

Google has been testing for several weeks a new advertising feature that delivers ads based not simply on a specific search term, but also on the immediately previous search, [Susan Wojcicki] said.

A user who types “Italy vacation” into a Google search box might see ads about Tuscany or cheap flights to Europe. Were the same user to subsequently search for “weather,” Google will assume there is a link between “Italy vacation” and “weather” and deliver ads tied to local weather conditions in Italy.

Google is “very careful” about traditional behavioral targeting, though. According to Susan, “Nothing is stored, nothing is remembered. It all happens within that session.” (Not sure about the “nothing is stored” bit when you enabled web history...)

Interestingly enough, some of the problems Google points out in relation to behavioral targeting of ads also apply to personalization of search results (something which Google emphasized they believe in):

Wojcicki highlighted the problem of a user searching “video games.” Advertisers might be wrong to assume the searcher was a gamer and not, say, a grandmother, looking for a gift for her grandson, she noted.

This is a good example of why personalization, at least in the form of “user looks at A a lot, A is related to B, let’s show more of B”, can be counter-productive. Then again, some of us recently wondered if Google perhaps partially employs a different personalization strategy: “user looks at A a lot, C is unrelated to A, let’s show more of C as that’s not what the user likely knows anyway.”

I wonder, with that massive amount of ads + searches Google has, if there’s some merit in allowing the software to figure it out for itself... evolutionary algorithms, self-learning style. Search sessions are automatically grouped into general patterns, and then random ads are presented, and when an ad performs well, more ads from that ad segment will be displayed next time, and so on, causing a “survival of the fittest ad” environment. Then when Google meets the press in 2012, they can tell the journalists, “We don’t have a clue anymore how our ads work, but click-throughs are higher than ever.”

[Via Barry Schwartz.]