While Google has become synonymous with searching for information on the World Wide Web, the idea of search itself has gone through massive changes over the last decade. When it first started some searches could take up to 20 seconds and today it would be shocking if it took more than a second. Not only has search become faster it has also become predictive and more accurate. Business Line met Ben Gomes, Vice-President and engineering lead for Google Search to know how search will evolve in the future.
How do you plan to make search more effective?
We are trying to create this experience where information is available to you wherever you go. Focus here is to make your mobile search experience more and more powerful. And in order to do that our overall trajectory has been that search needs to get a lot smarter in order to function well for you in this mobile context.
Knowledge graph is moving us from the world of strings on the web to mapping that to things of the real world. So if I talk about the work going on behind the scenes this technology is just the beginning and its getting a lot better. Now we're beginning to know a lot more things about the world. For example, If we take population statistics, we know about the population of countries and various factors and how those factors relate to each other.
How important is voice search in the mobile context?
The mobile device is with you everywhere but it is not easy to type in a question. In order to make it easier, we need to enable speech because you could be driving your car and cannot see your device screen. So If I am in Delhi I would like to know how far a particular destination is or what’s the weather like in Bangalore where I am going next.
How do you deal with understanding different accents when it comes to voice search?
We have made a lot of ground on understanding accents. Ten years ago, voice search, with my accent would not have been possible -now it's possible. We're working on it and you'll see large strides going forward.
How are you able to anticipate what users want to know?
Well we already know what people are used to on the web. We are leveraging that background on the web to bring answers to your questions because we know what other people needed. So that's the part we've been working on and that's the context of search and Google now- which is anticipated answers on the home screen even before you finish the question. For example my phone knows when I'm at home or work. It tells me how far I am from work, I need not type any address or ask directions. So ‘my data’ becomes an important factor. If you get your ticket on Gmail, we know you need to go to airport so we give you information on flight timings or that you need warm clothing where you're going. That's the layer of anticipation were building on the core of Knowledge graph.
How do you ensure that the information you provide is accurate?
The data we show you has very high accuracy but it may not be correct 100 per cent of time. So we constantly get feedback from millions of our users. Once we get information from users, we not only correct our database but also pass on that info to other websites. But we put a huge amount of effort to ensure data is very accurate. We do not launch anything before ensuring data is accurate to acceptable levels.
So what will be the next big thing in search?
You ask Google a question and Google will not just give the answer but also anticipate answer to next question. Google will tell you what you need even before you know you need it. It's going to be more and more about understanding ‘my data’.