Google's Semantic Search Changes Coming Soon

by Admin


23 Apr
 None    Search Engines


...


... It's What We've Been Saying All Along!

by Rebecca Maynes
http://www.mediative.ca

Google makes hun­dreds of changes to its algo­rithm every year, mostly minor, but some major, that can have a sig­nif­i­cant effect on traf­fic and rank­ings. For mar­keters, under­stand­ing these updates can help pro­vide insight into the changes in rank­ing and traf­fic, as well as ways of improv­ing search engine optimization.

In March 2012, The Wall Street Jour­nal reported that Google’s Amit Sing­hal had announced the impend­ing intro­duc­tion of seman­tic search into the Google algo­rithm. The change will soon be inte­grated into blended search results in order to pro­vide users with bet­ter search results.

Sim­ply put, seman­tic search is the process of try­ing to under­stand what an actual search query is try­ing to find out from the search terms used, and then deter­min­ing what the searcher is going to find to be the most rel­e­vant infor­ma­tion, not just a list of web­sites that are well opti­mized for SEO.  Rather than just rely­ing on match­ing key­words to web con­tent, Google will attempt to look for con­text, and the rela­tion­ship between the words used to make a search. Google will answer fac­tual ques­tions sur­round­ing peo­ple, places, com­pa­nies, his­toric events etc. right on the SERP. This requires a thor­ough under­stand­ing by Google of the searchers intent and the con­text of the query – made pos­si­ble by Google’s Knowl­edge Graph con­sist­ing of over 200 mil­lion enti­ties which could poten­tially “change search for­ever”.

Busi­nesses are likely to see a drop in CTRs. The Wall Street Jour­nal reports that a Google source has indi­cated that as many as 10–20% of all searches could be affected by the shift to seman­tic search, equat­ing to tens of bil­lions of searches per month.

But it’s not all bad news for com­pa­nies; they just need a change of mind-set when it comes to met­rics. As Google begins adding this addi­tional infor­ma­tion within the search results, Share of Page on the SERP, rather than CTR, becomes an impor­tant met­ric of expo­sure, indi­cat­ing that Google is serv­ing up your web­site con­tent to answer people’s questions.

My next blog post will high­light what the effects of this algo­rithm change will have on your web­site per­for­mance, and what you as a mar­keter can do to min­i­mize the effects. Media­tive will also be host­ing a webi­nar in May on the future of search and how these changes from Google are cre­at­ing a new search landscape.

BUT IS ANY OF THIS REALLY NEW? THIS IS WHAT WE’VE BEEN SAYING ALL ALONG

Seman­tic search is not a new con­cept. In fact this is some­thing that we at Media­tive have been talk­ing about for years. Back in 2007, when Media­tive was known as Enquiro Search Solu­tions, we released a whitepa­per titled “Search Engine Results 2010: What search engines may look like in the future” which included inter­views with lead­ing indus­try experts and their pre­dic­tions for search engines three years down the road. These indus­try expert include Marissa Mayer (who at the time was Google’s VP of Search User Expe­ri­ence and Inter­face Design), Chris Sher­man (Exec­u­tive Edi­tor of Searchengineland), Michael Fer­gu­son (for­merly of Ask.com), and Jakob Nielsen (the web’s best known usabil­ity guru).

Here are some key points that were made in the whitepa­per, writ­ten in 2007. While 2010 has come and gone, it’s inter­est­ing to see that many of the pre­dic­tions made 5 years ago are now part of our real­ity when it comes to search engines:

  • Smarter search engines

A major theme was not so much what search engines would look like but how they would get smarter in the back­ground. Dri­ving this would be fac­tors like per­son­al­iza­tion and tweak­ing of algorithms.

  • Per­son­al­iza­tion

As per­son­al­iza­tion starts to pro­vide the oppor­tu­nity to deter­mine rel­e­vancy not just by com­par­ing it to a key­word but also com­par­ing it to the intent of the user, this will result in a sig­nif­i­cantly dif­fer­ent level of inter­ac­tion with the search results page. The intro­duc­tion of things like per­son­al­iza­tion and uni­ver­sal search results will dra­mat­i­cally impact the entire world of search mar­ket­ing. The rules will change sig­nif­i­cantly and the strate­gies and tac­tics used by mar­keters will have to evolve quickly and dra­mat­i­cally in order to keep pace with the rate of change being seen from the engines themselves.

  • The notion of use­ful­ness being fac­tored into future search algorithms

Accord­ing to Jakob Nielsen, there is a ten­dency now for a lot of not very use­ful results to be dredged up that hap­pen to be very pop­u­lar, like Wikipedia and var­i­ous blogs. They’re not going to be very use­ful or sub­stan­tial to peo­ple who are try­ing to solve prob­lems. There may be a change and we may go into a more behav­ioral judg­ment as to which sites actu­ally solve people’s prob­lems, and they will tend to be more highly ranked.

  • The seman­tic search engine

Marissa Mayer talked about peo­ple mark­ing up search results and web pages and inter­act­ing with them in a way that indi­cated that they found them use­ful and valuable.

Michael Fer­gu­son said another thing that we are see­ing is more and more con­tent is actu­ally going to be sur­faced onto the results page when there is high con­fi­dence that is rel­e­vant. There might be a time you might see peo­ple adver­tis­ing and pro­vid­ing con­tent not just on web pages and blogs etc. but with short dis­crete self-contained video answers and audio answers that come up either as spon­sored or rel­e­vant content.

Gord Hotchkiss, SVP at Media­tive, has been talk­ing about search engines and pre­dict­ing changes that are likely to occur for most of his career. In arti­cles posted on searchengineland.com from March and April 2011 he talks about the con­cept of “mas­ter intent”, refer­ring to the use of search engines for a spe­cific task that is actu­ally part of a much big­ger and more com­plex task.

“Web search acts as a pretty sim­ply minded assis­tant in all this – going out and gath­er­ing rel­e­vant infor­ma­tion based on the words we feed it. But what if search knew our “mas­ter intent” and offered a much higher level of assis­tance, going out and gath­er­ing all the infor­ma­tion, fil­ter­ing it based on our require­ments and guid­ing us through the entire process?”

“There is a chal­lenge in our cur­rent par­a­digm of search, built on the con­cept of rel­e­vancy. It depends heav­ily on the abil­ity of a machine to under­stand human lan­guage. And that is no small chal­lenge. But, per­haps, search’s reliance on pars­ing lan­guage isn’t as crit­i­cal as it was before. As we do more things online, we leave more sig­nals to help a search engine deter­mine our intent…allowing for a more use­ful experience.”

“Search has to become more useful.”

With this lat­est algo­rithm update shortly to come, per­haps it has.


Biography / Resume : Rebecca Maynes is Mediative’ Marketing Communication Specialist, focusing on Demand Generation Communications. Using her expertise in creating and compiling engaging content, she captures and nurtures existing and new leads through the marketing funnel. Rebecca is responsible for the creation of Mediative’s monthly eNewsletter, case studies, and cheat sheets, and she also plays an integral role in whitepaper and eBook production. Rebecca began her career with Yell.com in England, and, after emigrating to Canada in 2005, she has gone full circle, joining Mediative, a Yellow Pages Group Company, in April 2009. Prior positions include Marketing for a B2B Software company. Rebecca graduated from Cardiff University in Wales, UK, with a First Class Honours BSc in Business Administration.




News Categories

Ads

Ads

Subscribe

RSS Atom