OTTAWA – Jonathan Effrat spent the better part of four years working on various parts of Google’s groundbreaking search engine before the real solution dawned on him.
The Toronto native realized the best way to help people find what they’re searching for is by presenting them with results before they even finish typing.
And now, the company is making its Google Instant Search available to Canadians. It’s been available in the U.S. for about a month.
Effrat, 29, who grew up in Toronto, is Google’s product manager for web search at the company’s head offices in Mountain View, Calif.
His love of technology was born from a passion for science, which he pursued at an early age.
In his youth, Effrat became a fan of the Ontario Science Centre in Toronto, where he saw an exhibit about the "information super-highway." He was introduced to something called the Internet.
"I remember using the Internet for the first time (at the museum) and visiting different websites," he said. "It just seemed like a whole new opportunity and a whole new interesting source of information."
That love of science also propelled him to study technical topics. He attended Marc Garneau Collegiate Institute in Toronto, and excelled at math and science before heading off to Stanford University in California to study computers.
After graduating with a master’s degree in computer science, Effrat was snatched up by Google to help refine the company’s search engine in 2006.
Over the past four years, Effrat has been working in what he calls his "dream job," helping the company through several evolutions of its world-leading Internet search engine, including the addition of images and news.
He said Instant Search was a natural next step for the company.
"My area of focus was around how do people formulate their searches and how to we get them to the relevant searches more easily," said Effrat.
"Google Instant was an evolution out of that."
Effrat looked at how people were using Google’s search engine and realized users were largely wasting their time.
Instead of punching in a search term, then hitting the search button and then waiting for results to pop up, why couldn’t software guess what results the person needs? Everything the search engine needs to propose possible results could be deduced based on the letters already entered into the search box, thought Effrat.
Results could be provided to users in real time before they even finished typing. It’s called predictive search.
"There was sort of an ‘aha’ moment," he said. "Showing results for the top prediction, or showing results before you finish typing, was a dramatic change to the experience that would make it relevant and useful like never before."
For example, entering the letters "gu" would see the search engine begin to offer results about Guess Jeans and guitar tuners. Adding an "m" for "gum" would see the results automatically change to start showing web pages for gum tree and gumbo recipes. Finishing the search with an "my" for "gummy" would see the results change again to show the gummy bear song and gummy bear lyrics.
The results shown by Google Instant are based on popularity. Google collects information on all of its users, including where they are, what terms they have searched for in the past and what websites they are going to. Doing so allows the company to make predictions about what terms a person may be looking for, or what sites they want to see.
The new feature is currently being rolled out to Canadians. Users need a Google account, such as a Gmail address, in order to use the new search technology. The Google account is what the company uses to track a user’s search history and preferences.
Searching with predictive search technology such as Google Instant saves people time, according to Effrat. Many users of Google’s regular search engine were spending as much as 90 seconds looking for websites. Google Instant cuts that back to between four and six seconds on average.
According to Effrat, Google facilitates more than one billion search results daily. Those searches using Google Instant search save users around 11 hours of search time for every second.
"This is something we see as a huge win in terms of saving people time and making it easier than ever to get to the results they are looking for," said Effrat. "I think it goes to show that search is evolving and there is still room for innovation."
Google chief executive Eric Schmidt has often said the company is moving toward a more automatic offering of search results.
Speaking at the Tech Crunch Disrupt conference in San Francisco on Wednesday, Schmidt said he sees a future quickly approaching in which Google will simply provide information to people based on their geographical location, without any user input.
For example, a person passing by the Eiffel Tower in Paris could see information about the history of the structure, restaurants, souvenir shops or the location of the closest bathroom provided to their portable device without the user actually entering any search terms into Google.
Comments
Want to discuss? Please read our Commenting Policy first.