Google is facing a threat to its dominance and market share of the search market. And it is not a trivial threat. Users of the social networks such as Facebook and less so Twitter, LinkedIn, Tumblr etc., are now spending more time within those networks than on Google and search engines, according to comscore earlier this year.
This is a recent development and one that Google saw coming, but has failed to prevent with its development of Google+, Google's social network - which, whilst gaining popularity, still cannot compete with the likes of Facebook.
In addition to this, mobile search now accounts for nearly half of all searches. (And 50% of those searches are local, e.g. restaurants, gas stations, real estate agents, and so on.
There are three issues here that Google is endeavoring to resolve. The first is that social users to date, have found searching within the social networks clunky and often useless.
Facebook has been working hard to implement better search software that functions more effectively than to date. The current efforts are limited, but could easily be improved upon to threaten Google.
Graph Search has already been rolled out by Facebook in the U.S. (shortly, worldwide) and this allows people to search and discover information from within their own social network. There is no real reason why Facebook should include web wide search beyond this, except for the fact that they have a captive audience, so why let the users leave just to search? Graph Search uses natural search, e.g. “what is the favourite book in my network?”
This leads to Google's second problem. The increasing quantity of data in the world and hence the need to use long tail terms (phrases of 3, 4 or even more keywords) to find a relevant answer on the search engines results (SERPs). This results in searches of increasing complexity in what could be termed “Unnatural Search”.
Sheetal Pinto, SEO content expert and owner of MintCopy, Inc. explains, “People do not want to string together nonsensical sentences such as “widget maker Toronto car radio MP3 supplier”. It is easier, more sensible and natural to ask “Who makes or supplies car radio and MP3 widgets in Toronto?”
Google has been aware of this increasing problem for a long time. Conversational search (which also includes the ability to speak your search directly within Chrome) was first rolled out in May 2013 and has become an integral part of the Hummingbird algorithm. Danny Sullivan tested conversational search out on Search Engine Land when it was first released.
If, however, users resort to the social networks to find an answer, from their peers or friends, then the logical way to ask is as part of a conversation or update or tweet. Even with only 140 characters to play with in a Tweet, it is still possible to sound like a member of modern day society rather than a caveman! On mobile search i.e. speaking into your phone to search, you will obviously use natural language rather than try to construct a sentence of only keywords.
And that is the third issue - the increasing number of mobile users. Despite Android, Apple still holds the lion's share of the mobile market, and with that, comes Siri. Siri previously used only Bing (as does Facebook) but with iOS7, Siri can also use Google. However, Apple could develop its own Siri Search that worked properly and Google's search share in the mobile market would be threatened.
Peter at WSI Digital Web says, “The inclusion of conversational search in Google back in May shows that Google is getting to grips with making the search engine far easier to use. After all, many searchers have never mastered Boolean operators or understood how to bring back the most relevant data for their search by using long tail terms. Now, all website owners must do is to optimize their websites for latent and conversational search.”