![]() |
|
|
| |
|
Return to home page | |
Google News Algorithm Watch - update as at November 1, 2007
Posted November 5, 2007
Newsknife's usual aim for our Algorithm Watch is to see if there are any broad
trends that might indicate a change in Google News' algorithm.
You could study any newly elevated sites to see if you could learn something
to apply to your site.
This time we know there's been a change. From September 2007 on, Google News
started including the news wires in their listings. The idea
with this particular Algorithm Watch is to see what effect that had on the performance
of news sites that were already appearing at Google News.
(The news wires concerned are UK Press Association, Associated Press, Canadian
Press and Agence France-Presse.)
The table shows for each site each month, the percentage of their total
appearances found by Newsknife, that were on the Google News home page.
For example, during November 2006 4% of the appearances we found by Forbes were
on the home page.
Look at the SEP07 and OCT07 columns and you can see the first appearances by
The Associated Press and the other news wires.
Scroll right to the last three columns of the
table...
* The column headed "Prev" has for each site, the average % found on the Google
News home page November 2006 - August 2007;
that is, before the wire services appeared. For Forbes this
averaged figure is 4%.
* The "New" column is the average % found on the home page September -
October 2007, since the wire services have been included. Forbes is 10%.
* The "Change" column is the increase in the % we found on the home page since
the wire services appeared.
Despite the appearance of the wire services Forbes and all
the other major news sites shown actually improved their home page performance.
This table is an excerpt from a list of 43 news sites evaluated by Newsknife
for this "Wire Services Special Algorithm Watch".
See the full list in the Newsknife Members'
Area. It includes MyFox Washington DC, Scripps News, CNNMoney.com, Washington
Post,
several Reuters sites, Boston Globe, SFGate.com, Los Angeles Times, CBS New
York, The Mercury News, Chicago Tribune,
United Press International, Atlanta Journal Constitution, Toronto Star, Los
Angeles Daily News, International Herald Tribune,
Times Online, Voice of America,
Economist, Seattle Post Intelligencer and Inter Press Service.
| NOV06 | DEC06 | JAN07 | FEB07 | MAR07 | APR07 | MAY07 | JUN07 | AUG07 | SEP07 | OCT07 | Prev | New | Change | |
| The Associated Press | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 8 | 7 | 0 | 8 | 8 |
| AFP | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 8 | 7 | 0 | 7 | 7 |
| Forbes | 4 | 4 | 4 | 4 | 2 | 2 | 2 | 3 | 11 | 10 | 11 | 4 | 10 | 7 |
| CBS News | 2 | 3 | 5 | 2 | 2 | 4 | 4 | 5 | 11 | 9 | 12 | 4 | 10 | 6 |
| Guardian Unlimited | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 1 | 6 | 7 | 7 | 2 | 7 | 5 |
| Reuters | 9 | 8 | 11 | 11 | 17 | 20 | 17 | 14 | 25 | 17 | 21 | 15 | 19 | 4 |
| New York Times | 7 | 8 | 12 | 14 | 13 | 11 | 14 | 13 | 13 | 13 | 18 | 12 | 16 | 4 |
| ABC News | 4 | 4 | 5 | 5 | 5 | 5 | 7 | 7 | 9 | 7 | 13 | 6 | 10 | 4 |
| The Press Association | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 5 | 3 | 0 | 4 | 4 |
| Bloomberg | 6 | 8 | 12 | 7 | 13 | 12 | 13 | 11 | 11 | 11 | 17 | 10 | 14 | 4 |
| The Canadian Press | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 7 | 0 | 0 | 4 | 4 |
| CNN | 8 | 8 | 13 | 13 | 16 | 17 | 21 | 12 | 18 | 17 | 18 | 14 | 17 | 3 |
| CNN International | 5 | 9 | 9 | 9 | 6 | 15 | 13 | 13 | 13 | 9 | 18 | 10 | 13 | 3 |
| Wall Street Journal | 0 | 4 | 4 | 7 | 5 | 7 | 9 | 13 | 9 | 7 | 11 | 6 | 9 | 3 |
| FOX News | 6 | 7 | 8 | 6 | 11 | 9 | 13 | 12 | 8 | 6 | 15 | 9 | 11 | 2 |
| USA Today | 9 | 8 | 8 | 5 | 13 | 13 | 11 | 11 | 6 | 6 | 16 | 9 | 11 | 1 |
We stress that these figures are Newsknife's opinion based on our findings.
They were not produced by Google News.
The figures have been rounded after calculation.
Treat the list with caution.
A change in Google News' algorithm is only one of several possible explanations
for the figures.
Other explanations and limiting factors include: the luck of when Newsknife
happened to monitor Google News,
the region where each month's major news event occurred, and a change of format
or content at a news site.
Copyright © 2007 Industry Standard Computing Ltd. All rights reserved