Removing duplicate content

[AMHC]

Due to URL changes and parameters on our ecommerce sites, we have a massive amount of duplicate pages indexed by google, sometimes up to 5 duplicate pages with different URLs.

1. We've instituted canonical tags site wide.

2. We are using the parameters function in Webmaster Tools.

3. We are using 301 redirects on all of the obsolete URLs

4. I have had many of the pages fetched so that Google can see and index the 301s and canonicals.

5. I created HTML sitemaps with the duplicate URLs, and had Google fetch and index the sitemap so that the dupes would get crawled and deindexed.

None of these seems to be terribly effective. Google is indexing pages with parameters in spite of the parameter (clicksource) being called out in GWT. Pages with obsolete URLs are indexed in spite of them having 301 redirects. Google also appears to be ignoring many of our canonical tags as well, despite the pages being identical.

Any ideas on how to clean up the mess?

[PatrickDelehanty]

Hi there

Are you sure you are using all of the tools above properly? Not saying you're not but people make mistakes and it's just something to look into.

When did you implement all of the changes? Was it recently or was it a long time ago?

How is your organic traffic and rankings? Did you check if you have a manual action at all?

Let me know - thanks!

[WebElaine]

It sounds like part of the problem may be the sitemaps you're sending. By including duplicates in a sitemap, you're basically telling Google that each version of the page is valid. I would remove them and resubmit a sitemap with only the canonical versions you want indexed and see if that helps.

[Linda-Vassily]

Yes, as Patrick said, surprisingly often something like this is a result of a simple oversight because we have been looking at the same code over and over...

Do you have access to Screaming Frog? You could crawl your site and see whether redirects/canonicals are behaving as you expected.

Have you taken a look at the html of one of the incorrectly indexed pages when it is loaded in your browser? Can you see the canonical? If you try going to a redirected page, does it redirect? [I know--way to obvious, but sometimes it is good to start at the beginning again when we can't root out an issue.]

Another culprit in these cases can be internal links. Do you link internally using any of the undesirable URLs? That can send a message to Google that those URLs are still in play. Again, you can use Screaming Frog to find those strings.

[RuthBurrReedy]

Are you basing this on a site: search? It's fairly common for URLs to appear in a site: search that otherwise will not appear for any actual searches. Are the undesirable versions of the URLs getting any search traffic?

[AMHC]

Where this is  appearing the most is on cross domain canonicals. We have duplicate content across 2 websites, and we've canonicaled some pages from Site A to Site B, and some from Site B to Site A. In theory, pages that were canonicaled to the other domain should be deindexed. When I run a rankings report, I see pages for the wrong domain ranking, a month later. They are pages with parameters, or old URLs that we've changed. It's like a game of whack a mole. Every time we get a page deindexed, a duplicate with a different parameter takes its place. And this is in spite of calling out these parameters in GWT.

What I imagine is happening is that we have several URLs for the same page indexed. When Google crawls our site, it is correctly canonicaling the page it crawls. In the rankings, however, Google is probably pulling a duplicate page out of its index, and ranking it without crawling it. If it was crawling it, Google would see the canonical tag, and not rank it. So we have an ongoing battle to get Google to crawl the page it just pulled out of its index to see the the canonical tag.

The reason for all this is that when a page cross domain canonicals correctly, the rankings for the duplicate page on the other site goes up dramatically. As long as Google keeps ranking the wrong pages, we don't get the rankings bump on the other site.