According to the Wall Street Journal, Google is poised to make some of the biggest changes ever to its search page behavior. That's Google's core business, and it's garnering a lot of attention because Google essentially is Internet search (and advertising, of course, more on that soon). This is a little like NASA saying it's going to rethink how it uses rockets. The move is all about bringing semantic technology to the search experience, making Google smarter at answering questions. Except, when you get down to it, it's really not at all about this.
What Google's going to do is basically trying to interpret a user's query as a fact-seeking question, and then surfacing context-sensitive answers to the top of the results page rather than a mere list of links. The semantic aspect is about using smart algorithms to try to divine a user's meaning, rather than simply a hunt for words that match the query text. Apparently Google's been busy for a couple of years putting together a database that'll power this system--a list of facts about people, places, and so on that is "hundreds of millions" entries in size. The idea is to deliver better content to a user early in the search, meaning they don't have to trawl through the search pages using the natural semantics-powered filter between their ears to spot the answer they need among the longer list of blue links Google serves up.
For example, the Wall Street Journal suggests that if a user searches for something like "what are the 10 largest lakes in California" Google's search results page may actually give you the answer rather than linking to sites that contain data that may be relevant (very hopefully relevent, you'd think, as Google's been honing its algorithm for years so that this is how it all works). If it can't find the data in its vast new database, then it'll use its clever semantic tech to find the answers in the billions of websites in the world.
It's all about keeping people on the homepage for longer, you may think (and Google has hinted as much to the Wall Street Journal), because there's more likely to be interesting stuff presented to you after the click on "search." That translates directly to eyeballs spending more time on digital ads on the page. And this, not search, is really Google's core business.
But let's pause to question all of this. Google's been trying to do this sort of thing for quite some time--most obviously because Google now searches for synonyms of what you type in as well as trying to directly match your query. Google Instant was a less language-driven but more recent attempt to surface meaningful content as you typed--even before you finished typing your full query (because you may find the relevant answer sooner this way). So in several ways Google's already doing this context-awareness, and though it's about to promote it to the top of the page, and probably marry it to a new engine to give it more power, it's nothing amazingly new.
Also, you may query if this new system will, on the whole, keep users' eyes on the homepage for longer. Surely getting an instant list of the 10 biggest lakes in California will satisfy your needs at a quick glance--perhaps eating up much less of your time than scouring down the current list of search results and, heaven forfend, actually scrolling down the page to see more links (for which read: "more adverts").
Speculatively, we may also wonder if this better understanding of context is gleaned from that big privacy-threatening, lawsuit-beriddled move Google made recently to consolidate all its data about each user in one place. Surely a compound database containing everything Google understands about your likes, historic searches, preferences, and habits will help it better understand the context of each search.
Then there's Bing, Google's ever-more interesting Microsoft competitor. Bing made a big play early on to say it was going to surface answers and facts sooner in the search results--things users may need, based on the context of their search, like the phone number or address of a company name they're looking up.
And ultimately all this PR bluster is about a different Google search rival. It's really about Apple's Siri. The voice-powered system gets some hard press attention because it promises much, but seemingly fails to deliver what you may expect it would just yet. Some of this is deserved, and Apple's done little to expand the international availability of Siri's powers for local business search and so on. But some of it is most definitely not deserved, as Siri is probably the most reliable voice-powered search system the average user has ever encountered. Statistics also show Siri now makes up over 25% of queries to Wolfram Alpha, the semantic-powered, context-aware intelligent answer engine. This is because Siri's a filter on your voice searches, and it's aware that some facts are best served by looking at a weather database, others will be solved by WA, and others by a websearch on Google.
That positions Siri as a gateway for search that is definitely shuttling traffic away from Google and its ad-sporting answers. And sure, Siri's only on iPhones for now. But Apple's expected to sell tens of millions more iPhones this year, and it's completely plausible it'll expand Siri to the new iPad and perhaps OS X soon enough. That's an ever-growing threat to Google's lifeblood--traffic. And it's not just from Apple, as Siri's so clever it's spawned clones like Evi. As Netizens, particuarly when mobile, get used to this idea of new search tech, they may choose Google less. And the company would find that idea, well, evil.
Unless it keeps up. Aha!
[Image: Flickr user Katie Weilbacher]