Google Wants to Find the Perfect Format for Your Ad

Last Updated Jun 3, 2010 5:09 PM EDT

Ad powerhouse Google (GOOG) has done amazingly well applying automation to selling and delivering ads, not only in search, but in online display advertising. A patent filing made public today shows that Google has opened a new front in ad automation: using statistics to help advertisers determine the perfect ad format, including such factors as size, page placement, and color.

U.S. patent application 20100138553, titled Selecting Format for Content Distribution, discusses a way to use statistical analysis to determine not only where to distribute an ad, but how to format it.
Content is distributed in computer systems or by other technologies in different situations. For example, advertisements can be used in an attempt to inform people about a wide variety of products, goods, and services. Generally, advertisers may seek to target the content of their advertising to the intended audience or viewers.Advertisements can take many forms, such as printed material, commercials on television and radio, billboards, and online advertising. These advertisements can be placed without detailed knowledge about the potential viewers with the intention of reaching potential customers who encounter the advertisement coincidentally. In some cases, advertisements are placed to target a particular demographic group (e.g., ads for toys in a children`s` [sic] TV show or billboards for tires along a roadside) in order to increase the chances of reaching potential customers.
Online advertisers have been able to target ads to particular people through behavioral information and analysis. What they haven't had is a virtual equivalent of a format that might work best for a given target market. That's what Google's application discusses:
Among other disclosed subject matter, a computer-implemented method for performing a content distribution associated with a format includes performing analysis of an information collection reflecting network activities by a user, the network activities occurring in response to earlier content distributions to the user. The method includes selecting a format for a content distribution to be made to the user, the format being selected based on the analysis. The method includes performing the content distribution to the user using the selected format.
In other words, systems would analyze how different individuals and groups react to ads of varying formats and contexts, and then deliver the ads to the right places in the right formats. Some of the format attributes that Google mentions include:
  • color
  • size
  • particular layout
  • shape
  • specific placement on a user interface
  • time of ad delivery
  • behavioral tendencies of users
  • geographic region
Although these capabilities would be of questionable use in search advertising, according to IDC analyst Karsten Weide, thanks largely to display ads on YouTube, Google has the third largest display advertising revenue, after Yahoo (YHOO) and Microsoft (MSFT).

When I spoke with Weide today, he said that at first blush, the system sounded like Yahoo's Smart Ads. "What is not new about this is creating an ad on the fly, custom tailored to a particular user," he said. "Let's say they know you're looking for a car, so you only see BMW ads. Because you're in San Francisco, you only see ads for BMWs in San Francisco."

What is new is using data to drive aspects of the format. "It fits into the greater ad scheme of ad tailoring," he said. "We're seeing a whole lot of initiatives in that field."

Related: Billboard Image: RGBStock.com user cienpies, site standard license.
  • Erik Sherman On Twitter»

    Erik Sherman is a widely published writer and editor who also does select ghosting and corporate work. The views expressed in this column belong to Sherman and do not represent the views of CBS Interactive. Follow him on Twitter at @ErikSherman or on Facebook.

Comments