Dan Petrovic has explained how he hijacked a few pages in Google to show his copied version over the original version of the page.
For example, he was able to confuse Google into thinking a page on MarketBizz should really show on dejanseo.com.au instead of on marketbizz.nl.
How did he do it? He simply copied the full page, source code and everything and put it on a new URL on his site. He linked to the page and gave it a +1 and the result worked days later. He is a picture of Google’s search results for the page using an info command and also searching for the title of the page:
He did the same thing on three other domains with varied levels of success.
We emailed Google last week for a comment but have yet to hear back.
In some cases, using a rel=canonical seemed to prevent it from hijacking the result fully but not in all cases. There also seems to be a case where using the authorship might be prevent this as well.
Dan Petrovic was even able to hijack the first result for Rand Fishkin’s name (with Rand’s permission):
The way this seems to work is that Google’s duplicate content system feels that the new URL is the more important page and thus replaces the original page with the more important page. It is how the competitive link trick seemed to have worked as well.
Postscript: Google has taken action against these attempts with a notification sent to the webmaster for “copied content.” Those pages were removed from the index.