Forum Moderators: Robert Charlton & goodroi
When searching for keyword1, I get normal results. The first result is www.domain.com where the domain name is keyword1. There are also no problems when I search for keyword2. One result within the top ten is from www.domain.com. However, there are no results from www.domain.com when I search for keyword1 keyword2. Results of the form www.domain.com are completely filtered out (there are less than 1000 regular results). However, repeating the search with the omitted results included leads to a re-inclusion of the results from www.domain.com in the first places. Also, adding an additional word can result in 'normal' SERPs. The first keyword is not a real word but the domain name (copyrighted)! The second keyword is a regular word.
To give an example: Suppose searching for microsoft would show www.microsoft.com, searching for explorer would show a different page from the same domain, while searching for microsoft explorer doesn't return any page of the form www.microsoft.com.
I' ve observed this behaviour for a long time (several months) and numerous keyword combinations. (Sometimes results from www.domain.com are not completely filtered out but appear at the end of the SERPs as supplemental result.) Im my opinion it's a bug because results which doesn't violate Google's guideline are completly filtered out. They are just filtered out for special keyword combination while adding or omitting words (as well as setting filter=0) leads to a re-inclusion. I guess that the behaviour is a side effect of the algorithm and it's accepted even if it's an unwanted effect. It doesn't seem to be a duplicate content problem.
Of course, Google can change the algorithm in any way they want. If they are aware of the problem and decide that there is no need to change the algorithm, it's fine. However, if this behaviour is a bug I hope they try to fix it. (This wouldn't be the first bug [webmasterworld.com]. The last time they fixed it within 10 days.)
