Online Content, A Day in the Life - Annual Report 2006

Google

Google News offers users an entirely different experience from the New York Times or the Web sites of other traditional news outlets.

Here the news is edited not by people but by algorithms, and the site produces no original content whatsoever. In other words, computers choose from a mix of content produced elsewhere.

“Search and browse 4,500 news sources updated continuously,” the page promises at the top. The result is less an ordering of the news than a kind of stacking it in different piles — with some 14,000 articles accessible from the front page of the site.

Yet in its computerized effort to be constantly new, the site also reveals the degree to which the continuous nature of the 24-hour, seven-day-a-week news cycle is not really so continuous. Most of the stories added to the site through the day are nearly identical versions of the same event from different news outlets. There is no really new information to report, just newer filings of stories. For instance, the plane story from the Canadian Broadcasting Corporation, which led the site at 5 p.m. had different — and arguably more interesting — information than the plane piece that was the lead at 9 p.m. from Reuters. The Reuters piece was longer and with more official response, but the CBC piece not only described who the pilot was and what he was wearing — a t-shirt and shorts, which seemed to further emphasize that he was a student pilot who made a mistake — but it included a picture of his arrest. And by 9 p.m. the CBC piece was gone from the front page.

Google News Lead Stories

The site leads with two top stories in its main, middle column and five more headlines right of that, which vary from business to sports news. Below that are a list of subjects “In The News”— on May 11 the keywords included Delta Air Lines, Riverside County, Van Gundy and the Detroit Pistons.

That is followed by three top stories under each of eight different topic headings — World, U.S., Business, Sci/Tech, Sports, Entertainment, Health and More Top Stories.

For each news event, Google News offers apparently every related story it can find. On May 11, for instance, the top story at 9 a.m., about a grenade found near the site where Bush delivered a speech in the former Soviet Republic of Georgia, linked to 1,968 other stories about the event. In all, the 31 “stories” highlighted on Google News’s front page at 9 a.m. this day were actually links to 14,228 separate stories. The depth is breathtaking. The utility of it, for an average person, is harder to fathom.

With each subsequent visit, we found new stories in most of the spots. According to the numbers, 80% of the top stories were new on Google. Again, however, usually they were about the same news event, just new versions from a new outlet. At 1 p.m., for instance, the story about anti-American protests in Afghanistan was from CNN, plus links to NPR, the San Diego Union Tribune, Radio Australia and others. At 5 p.m., there were stories from the Associated Press, ABC News, the International Herald Tribune. At 9 p.m., a story from the Scotsman (UK) plus Reuters and the Guardian on top. The actual accounts didn’t vary much, sometimes not even the original source. At 5 p.m., for instance, the story on Afghan riots came from the Guardian Unlimited, but on a closer look it was actually an AP story. The story on the riots just underneath it is from ABC News, it says, but it is also the AP story.

On Google, some topics got more prominence than they did elsewhere. Sci/Tech, for instance, comes third in their list, followed by Sports, and Entertainment. Subjects that get more prominence elsewhere, such as politics, are not in the headings here.

In exploiting the potential of the Web for multimedia and interactivity, Google fell behind, at least on this day. Only 5% of the stories on the site had links to video of the event in the news compared with the average of the sites monitored of 45%, though again since that was a result of grabbing stories from a variety of sources, the inclusion of video was ultimately the call of those outlets. None of Google’s stories had links to graphics, maps or special text boxes (the average on this day was 17% of stories). Only 5% of stories had some link by which users could manipulate or customize data, whereas the average was 18%. And only 15% of stories on Google had links through which users could follow up with queries or communicate with someone, compared with 39% in our sample over all.

The only multimedia element in which Google was above average was in the use of audio. Some 15% of the top stories had such links, versus the 6%.

How did the news agenda offered by Google’s computers compare with those of the editors of NYTimes.com? At 9 a.m., Google led with the story about the unexploded grenade found near where President Bush had spoken, a story that never was near the top of the New York Times news agenda this day. Both sites had Iraqi violence at the top. Google had the Afghan riots next, a story the Times would not post for two more hours. But the Times had exclusives, an interview with Defense Secretary Donald Rumsfeld and a story about AIDS in Africa that were nowhere on Google’s site. The United Airlines pension story that would dominate elements of the American press this day was not listed among Google’s top stories anywhere on its page.

By 9 p.m., the grenade story was again one of the “Top Stories” on Google (after being displaced by a general US-Georgia relations story at 1 p.m. and a piece on the Canadian budget at 5 p.m.) The Times also had the plane scare but continued to think violence in Iraq was important, a story now replaced on Google by the next day’s relative calm. Throughout the page, the stories also tended to differ, topic by topic. Apparently, at least on May 11, the choices made by Google’s machines differed from the choices made by editors at the New York Times.