This copy is for your personal, non-commercial use only. To order presentation-ready copies for distribution to colleagues, clients or customers, use the Reprints tool at the top of any article or visit:

Back to Previous Page

 Mail    Print   Share Share

So What Do You Do, Krishna Bharat, Lead Engineer of Google News?

Bharat reveals the formula for a flawless online news experience and how companies can get readers to pay for it

By E.B. Boyd - June 30, 2010
Say the words "Google News" to a journalist in this age of decreasing newspaper revenues, and you're likely to get a lot of teeth-gnashing. Interestingly, however, the site was created back in 2001 by a Google engineer with intentions any journalist could respect: Sept. 11 had just happened, and Krishna Bharat was looking for an efficient way to access a wide range of news reports on the attack.

Eight years later, Bharat is still the lead engineer of Google News (his formal title is Distinguished Research Scientist). Having grown up in India, Bharat went to Georgia Tech for his doctoral studies in the early '90s, where he first started tinkering with ideas about how to display news online. (If you want to really geek out, check out his paper on the project here.) Today, Bharat speaks frequently on the subject of what Google is doing in the area of journalism. "Google didn't sign up to solve the problems in the industry," he told, "but it feels itself inevitably drawn into it, partly because we are a premier access point for many people seeking information." caught up with Bharat to learn about the origins of Google News and about the company's other journalism-related efforts.

Name: Krishna Bharat
Position: Founder and head of engineering for Google News
Resume: Research staff member at DEC Systems Research Center. Joined Google in 1999. Founded Google News and set up Google's R&D operations in India. Received the World Technology Award for Media & Journalism in 2003.
Birthdate: January 7, 1970
Hometown: Bangalore, India
Education: Ph.D. in Computer Science from Georgia Tech
Marital status: Married, three children
First section of the Sunday Times: Front page
Favorite TV show: None: "[I] don't watch TV."
Guilty pleasure: "Researching travel destinations I may want to visit."
Last book read: Before the Dawn by Nicholas Wade
Last favorite gadget bought: iPad
Twitter handle: @krishnabharat

Did anyone in the company ever raise concerns about Google News possibly violating copyright?
We were a search engine where people came for information. We'd show the title of a document and a small snippet and source information. From what I could see, Google News was doing the same thing, albeit without a query. The query was "Show me news." So it didn't strike me as any change in the model as to how Google was operating.

Eight years later, has Google News turned out the way you expected?
Initially we didn't know how this would take off. It was a new paradigm. Trying to get people to read multiple points of view, I felt, was very important as a social good. Democracy depends on a good understanding of issues…. Whether people indeed wanted to consume news that way was a big question mark. But when we launched it, the response was tremendously good, and it has been tremendously good ever since.

The second question in my mind was: We tried this in English in the US; how is it going to work in other countries? We've now launched it in 60-plus editions. It is, in fact, applicable in all of these different locations. So that's also been very satisfying.

"In order to be successful, you need to get inside the reader's head and understand how you need to adapt your presentation to what they want."

News organizations have generally thought they understood pretty well how people wanted to consume news. What have you learned about new ways people want to receive their news?
The news industry developed a lot of their thinking around the constraints of print and the other types of mass distribution, like TV. For example, the idea that everybody gets the same thing. The Internet has shown that everybody doesn't have to get the same thing. In order to be successful, you need to get inside the reader's head and understand how you need to adapt your presentation to what they want.

Online, the news experience is a mix of [four things]. First, you have to know what everyone needs to know. Then, you have to follow your personal interests. Beyond that, you also want to understand what's popular, what the social networks are buzzing about, what are my friends suggesting to me. So the four things are: editorial "required reading," if you will, personalized news, social recommendations, and audience popularity. Those are the four components that make up a news reading experience, and I think Google News will evolve to accommodate all of those.

What else is ahead for Google News?
Being much faster. What we discovered from FastFlip [a Google experiment launched last fall] is that when people get an opportunity to read news fast, they read a lot more. Giving you an opportunity to flip through many pages increases the probability that you will find something that you will want to read in-depth. And of course you can show ads in the previews, as well. It's an interesting paradigm, and it has potential in the following sense: A publisher can monetize their content not just on their site, but elsewhere, by allowing people to host and preview it. It drives traffic to them, and revenue.

The other thing we need to address is how to make it easy and natural for people to pay for content and… how to make it work with search and social networks. On the one hand, you want people who have never experienced the content to discover it. On the other hand, once they become a fan, and if it's high-quality journalism and expensive to produce, it should be convenient for them to pay for it.

"Our challenge isn't trying to introduce the concept of payment to consumers. It's to explain to them why this is different from everything else and worth paying for."

A recent article in The Atlantic said you and your Google colleagues think free-versus-paid is an "empirical rather than theological matter" -- that it's not a question so much of whether people should pay, but of what works in the market. Do you think people will pay for content?
Yes. People pay for media in other settings. They pay for movies, they pay for music, they pay for magazines off the shelf. If it is high-quality content, and the billing is natural, I think they will pay. The issue is how do you make the value proposition to them, [by offering] something for which there is no obvious substitute. You're not going to pay for one article that is just like any other article that covers the death of Michael Jackson. The brand, the quality of the journalism, the depth of insight, the look-and-feel, the experience, the infographics -- all of that comes into play. Our challenge isn't trying to introduce the concept of payment to consumers. They understand that. It's to explain to them why this is different from everything else and worth paying for.

What's Google doing to help publishers build paywalls?
We are looking for ways in which Google can help publishers put content behind paywalls if they choose to... There is an intersection between the world of paid content and how search engines want to operate in the future. We want to make sure that we a.) help publishers achieve what they want to achieve and b.) also are in a position to make search work with that. We're providing one piece of technology that publishers could use to implement paywalls.

Last winter, Google road-tested its Living Stories idea with The New York Times and The Washington Post. What's the idea behind Living Stories?
The core idea of Living Stories is to have a persistent URL for every story. It shows you the latest update up front, so you can catch up. But if you want some background, it shows you the background. If you want to look at opinion pieces on the topic, you can look at that. If you want to see images, you can look at that. You can go back in time, in a timeline view. It is a "living story" in the sense that it's not an article that's frozen in time. It evolves all the time. You can link to it in social media and discover it through search. When you come back, you get the latest rendition [Example: Washington Post living story on Fixing DC's Schools.] of that same thing.

From an SEO perspective, it makes a lot of sense for the publisher -- one link that people will want to link to. It also allows the investment of journalistic effort today to pay dividends not just today but in the future. We partnered with The New York Times and The Washington Post to showcase what can be done, but our longer term goal was really to have the industry involved in building Living Stories -- either using our technology, which we open sourced, or building their own.

If you could wave a wand and have news organizations do three things differently than the way they do them today, what would you change?
They should realize that "value add" is something that will ultimately influence how their content [performs]. If you're producing content that is substantially similar to what everybody else is producing, it doesn't help. It would be nice if they specialize in ways that allow them to do an excellent job on where their core competency lies. So if you view yourself as a local paper, then that would mean investing a lot of resources on local news. If you're a technology-focused publication, then technology is your forte. So, rather than having every publication cover everything, trying to get more focus within a publication would be a nice thing.

The second thing is that while networks have become much faster, websites have become much slower. A lot of it is because they've piled on media elements without realizing the consequences on download time. They should really focus on making it super-fast to load content. That'll pay dividends in terms of how much content their users look at and often they come back.

The last thing is they should move into personalization as a way to engage and retain users. There is the news that you need to know. I don't think we want to personalize that. And then there is all the other 500 things the site could show you. In that space, if, for example, there are certain sports you don't follow, those sports should recede, and the ones that you do follow should come to the forefront.

I checked out your Twitter feed. You're not a big tweeter.
I tweeted for a while and then I got tired of it. But I'm a really bad example of many things. I study the Web, I use the Web, but I don't contribute very much to the Web, sadly speaking.

E.B. Boyd is a freelance journalist based in San Francisco.

© WebMediaBrands Inc. 2010. All Rights Reserved.

This interview has been edited for length and clarity. The foregoing is the sole property of WebMediaBrands Inc. The opinions and views expressed in the interviews and/or commentaries are solely those of the participants and are not necessarily the views of WebMediaBrands Inc., its affiliates or subsidiary companies.

> Send a letter to the editor
> Read more in our archives