joined:May 11, 2012
Just as recap, my website uses multiple store views with a series of international domains. I have a .ca site for Canada, .co.uk for UK, etc. All others including the USA default to .com. I was trying to use hreflang tags in the header to direct traffic but Google for whatever reason looked at them like back links. In other words, the .co.uk store had 10K links coming from the .ca store and the .ca store would have 5K links coming from the India store. This made for a huge mess with Google.com search results. In the case of search engines like Google.ca, we seemed to be fine because the .ca domain was associated with a target of Canada which I guess allowed Google to ignore any other strange requests coming from the other international domains. In the case of the .com domain, not territory was targeted so I guess that caused confusion when there were thousands of international links coming from different domains. In the end, I can search in Google.com and see search results for both my .com store, my .in store and .co.uk store. I am afraid because they are so identical minus maybe currency types, that it could be considered duplicate content.
So, realizing that the hreflang tag approach was no working for any reason, I decided to go in another direction. I have since removed all of the hreflang tags from the header as of about a week ago and used a GEO IP redirect module to move Canada traffic to the .ca store, UK traffic to the co.uk store, etc. All remaining traffic including the USA goes to the .com store. Seems simple. I have used a VPN program to check it out and it works perfectly. If I say I am in India, for instance, it pushes me right to the .in domain. I'm happy about that.
I updated my sitemaps, resubmitted them and asked Google my fetch my sites again, same of which are 1,000+ pages deep.
Right now, in Google Webmaster Tools for my .com store, it shows the following under "Links to Your Site".
Who links the most
As you can see the top three domains linking to my .com site are my own series of international domains. This is the same on all domains, but for the India domain for instance, the numbers are much smaller like 50 links. The .com store is the most out of whack of any of them as far as the numbers go.
What is strange is, when I dig deeper, I can see that one of those links coming from the .ca store is from a particular url labeled /samples/. Now, the /samples/ url in the footer of each page, so that might explain why the link numbers are so high! BUT, Google Webmaster Tools says that from my .ca store, there are a total of 691 /samples/ link going to the .com store. I know that this has to be related to the hreflang tags because nowhere else on my site do I now have links going from one domain to another. Still, if I go to cache:mydomain.com/samples/ I can see that the page was cached two days ago and then, the hreflang tags were removed. I can also scan the source code of that cache ever more and see that links going to the .com store or any other store exist.
So why does the report in Google still show outdated results as if the the hreflang tags were there? Is there a massive delay between Google results and GTW data? Should I not be alarmed?