Google Plus: A Sign Of The Future of Search

Have you noticed the plus sign within some Google SERP that started showing up next to invitations to map an address or get a stock quote a month ago?

Pundits are quick to conclude that this is a portal feature and that Google.com is compromizing their search-only philosophy to become a “portalized non-portal.” (RC Jordan) They accept the obvious explanation that Google is embedding portal content in SERPs to improve stickiness and deprive competitors of traffic while extending the Google brand.

It is easy to dismiss these UI changes as a strategy for Google to gain market share for services that are not performing well. Leveraging a dominant platform to gain market share for another product or service isn’t new or particularly exciting—even if it can be very effective. Despite Google’s successes in search and online advertising, many of their other properties are not performing well. Google Finance, for example, didn’t make the top ten according the Center For Media Research’s January 2007 data. Hitwise data from May 2006 shows Google Maps a distant third to Mapquest and Yahoo.

PlusBox is not intended to be a competitive sledgehammer, although it may serve that purpose. Google has successfully resisted the siren call of manipulating organic search in favor of its own properties and those of its partners for eight years. Search for photo editing software and Picasa doesn’t make the top 10 for organic SERP.

PlusBox is more important than bolstering finance or maps; it offers a glimpse into the future of search. Search engines have dramatically improved over the last decade, but some of the improvement in relevancy is driven by how we search. Users don’t tell the engines what we are looking for; we enter queries for key words that we have learned will help the engine differentiate what we want from other sites.

Plus box joins Onebox and Sitelink as the first steps to go beyond the user query terms and provide real relevancy. Using complex algorithms to create a statistical approximation of artificial intelligence that incrementally improves results — discovering what we are actually looking for and providing it within the SERP

A search a few weeks back for Children of Men illustrates the distinction. Google (and Ask) correctly determined the search was for a movie. The OneBox result in Google contained an invitation to get show times near me in Berkeley by entering a zip code. Ask offered reviews, show times and a link to the official site within their version of OneBox (along with an interesting assortment of suggestions in their “Narrow Your Search”). Yahoo and MSN showed a Yahoo News story followed by the official site for the movie.

Enriching SERPs with results that predict intent can be accomplished with a statistical analysis of user behavior. Google acknowledged that they monitor user click response to UI experiments. Melissa Mayer, Google VP, Search Products & User Experience, in a recent interview in Search Engine Land, describes the process, “We hold them (OneBox results) to a very high click through rate expectation and if they don’t meet that click through rate, the OneBox gets turned off on that particular query. We have an automated system that looks at click through rates per OneBox presentation per query. “

How hard is it to imagine that Google is leveraging this understanding to predict user intent and provide what we really want instead of the page that matches the search term we enter?