On Tuesday, Adweek posted an article which reported that Google is testing a new search retargeting feature. The story alleged that Google had begun allowing advertisers to run display ads across its Google Display Network based solely on a user’s search query, regardless of whether they had clicked on a search result or not. That would have marked a major change in Google's previously stated policy.
However, Google has informed Adweek that no such policy change has been made, and Google advertisers are not able to retarget search users on an impression basis. Adweek sincerely regrets the error.
Google has expanded the number of advertisers testing search its retargeting capabilities, which enable marketers to target ads across the Google Display Network for up to seven hours after a user has input and clicked on a search query, according to sources that are participating in a beta test.
In February 2010 Google announced  the ability for AdSense publishers to contextually target ads on their sites through the use of referral URLs that are created when a user clicks to a site through a search query result. Here’s how it works: A user might search for “Prada shoes” and then click on a search result or search ad. Later in the day (within seven hours) that user might visit any site within the Google Display Network and come across an ad from, say, Prada or Saks Fifth Avenue.
Google doesn’t break out search and display advertising revenue separately in its quarterly earnings statements, but in the first quarter advertising on Google Network sites contributed $2.9 billion in revenue, or 28 percent of Google’s total advertising revenue for the period.
Scott Jones, general manager of display for Responsys (which is not participating in the beta), said that search data is a valuable signal for targeting display ads, categorizing it among first-party signals like a brand’s CRM data. “It’s all about intent, and from our perspective, there are a lot of signals that represent intent. One of them is what term were you looking for or what content were you specifically looking for,” he said.