WebProNews

Googler Talks About Not Reindexing Pages

Ever had a problem with Google indexing your pages? If there’s no real content on them, then that’s probably why.

Barry Schwartz at Search Engine Roundtable points to a Google Webmaster Central help thread, where one person says they had over 2,000 pages indexed on Webmaster Tools, but that the number went down to 60, and eventually back up to 116.

“When I google: ‘site:www.gamez4you.com’ I see all the indexed pages correctly,” the webmaster says. “I suspect that is the reson my Adsense CPC went down by 60-70%.”

Google’s Gary Illyes responded, indicating that it was a lack of content on the pages that was the problem. Here’s his full response:

As we improve our algorithms, they may decide to not reindex pages that are likely to be not useful for the users. I took a look on the pages that were once indexed but currently aren’t and it appears there are quite a few that have no real content; for example http://www.gamez4you.com/car-games/play-crazy-mustang seems to be a soft error page, which means that even though it comes back with 200 HTTP status code, it is in fact an error page that shouldn’t be indexed. Another example would be http://www.gamez4you.com/all-games/page-71 which seems to be an empty page.

Another reason you may see the number of pages dropping in the Webmaster Tools’ Sitemaps module is that your Sitemap is referencing URLs that are not canonical. For example in your Sitemap you reference
however the currently canonical URL of that particular URL is
To fix the indexed urls count, I would recommend fixing canonicalization of the URLs one your site and to set your server to return proper status codes (e.g. 404) for inexistent URLs. You can read more about canonicalization at http://support.google.com/webmasters/bin/answer.py?answer=139394 and about soft error pages at http://support.google.com/webmasters/bin/answer.py?answer=181708
Hope this helps

Another webmaster in the thread indicates that he had a similar problem to the Sitemap referencing URLs that aren’t canonical issue.