This is a follow up to our recent article where we examined an improper use of the rel=canonical tag to see what it’s impact would be on the pages rankings in search engines, specifically Google.
As I said at the end of that article, after Google penalized our pages, we quickly removed the tag to see 1) If Google would remove the penalty on our target page and bump it back up to it’s previous rankings 2) If Google would re-index the page that contained the canonical tag.
Just yesterday I ran another report in WebPosition and got the following results (much to my relief after getting our newsletter landing page penalized).
Our target page made a jump of 52 spots in Google which put it within a couple spots of where it was before. Our newsletter feedback page was re-indexed and regained it’s PR of 2. (now we’re back to the feedback page outranking the landing page so I’ll have to use my robots.txt file to block it. No more crazy experiments with canonical tags)
In the original post I pointed out that it was unlikely we were penalized by Bing or Yahoo as none of the pages were de-indexed and only dropped in rankings very slightly. I didn’t assume that the drop was necessarily related to the canonical tag as there are obviously small fluctuations in SERP’s all the time. However, you can see in the report above that the page moved up 9 spots in Bing immediately after the tag was removed so the change might have some connection. The changes in Yahoo on the other hand were too small for me to assume any link.
All in all, it’s good to see that Google can forgive pages for making mistakes. Still, you have to be careful. I’ve read about instances of sites never fully recovering from a bad canonical usage error.