|Short lapse in redirecting domain aliases creates ongoing problems|
| 4:09 pm on Jan 28, 2013 (gmt 0)|
Like many websites, we have more than one domain name: widget-shop.com, widget-shop.co.uk, widgetshop.eu etc. etc. And, like any decent site we ensure that they correctly redirect to our principal domain with a 301 code - so www.wiget-shop.co/foo.html => www.widgetshop.co.uk/foo.html.
For a short period five months ago, we made the mistake of not ensuring one of our domains properly redirected to our common domain. So our entire site was duplicated: www.widget-shop.com gave exactly the same results as www.widgetshop.co.uk
Pretty stupid oversight. Google indexed widget-shop.com in all its glory. The mistake was spotted after six days, and the correct 301 redirects put into place. Through Webmaster tools we let Google know that widget-shop.com was an alias for our principal domain. Double-checked it was set up right using 'fetch as googlebot'. And crossed our fingers.
The first thing was saw was that our main site (widgetshop.co.uk) now had as its principal linking domain widget-shop.com. Every link in Google's cache for the wrong domain was now (correctly) redirecting to our principal domain. Result: I have over 6000 links from widget-shop.com to widgetshop.co.uk. And not only are they not going away... five months on they're still growing.
The result was a collapse in our rankings. Traffic down 50% and all sorts of search engine pain.
I think what is happening is that from time to time Google still looks at what it cached for widget-shop.com, finds new links it hasn't bothered with before, finds they're redirecting to widgetshop.co.uk and add them to the list of linking domains. It's certainly not prepared to believe that they're the same thing.
I can still see all those pages from widget-shop.com in the Google cache if I do a site search - and the canonical meta tag correctly shows widgetshop.co.uk/foo.html. At least we got that right, but it's made absolutely no difference. In fact nothing has. There's nothing more I can do about this utterly bogus domain, but it pains me to watch, every week, its links growing and growing. I can't be absolutely sure it's the cause of a rankings drop, but the slow strangulation of our organic results since this occurred makes me think it's probably connected.
So, be careful with those domain name aliases, and don't end up like us!
| 1:29 pm on Jan 30, 2013 (gmt 0)|
How short a time period did Google crawl the duplicate site?
About how many pages did Google crawl?
| 3:29 pm on Jan 30, 2013 (gmt 0)|
hi starfish, can I ask whether your pages went into 950 and were grouped at the back of the results? We too had a 301 problem across our www and subdomains where Google was indexing 3 versions of the same page across domains.
| 11:53 am on Feb 7, 2013 (gmt 0)|
I'm completely at my wits end now. The links from the bogus domain just keep growing - now standing at 8,727...
Here's the set up according to Google's webmaster tools:
widget-shop.com (which was the domain that briefly ran side-by-side with www.widgetshop.co.uk):
- Crawl Stats -- ~100 pages /day
- Total Indexed -- 56 [good! - don't really want any]
- Ever Crawled -- static at 4632
- Change of Address -- www.widgetshop.co.uk
And here's our principal domain, www.widgetshop.co.uk:
- Crawl Stats -- ~4000 pages/day
- Total Indexed -- 4063 [sounds about right]
- Links to your Site -- widget-shop.com, 8727 links
So what are these links? Here's what webmaster tools gives as an example of a link to www.widgetshop.co.uk
1309 links from widget-shop.com to /information/foo.html
[=> originating links]
- page widget-shop.com/product/0001234.html via widget-shop.com/information/foo.html
- page widget-shop.com/product/0001235.html via widget-shop.com/information/foo.html
- page widget-shop.com/product/0001236.html via widget-shop.com/information/foo.html
So it looks to me like Google is seeing links on its cached pages, it knows they 301, and is saying "hey look at all these links to www.widgetshop.co.uk!" The problem is that for some reason it can't / won't take the originating pages out of its cache, despite the redirection request etc. and the fact that they all too 301 correctly.
I wouldn't mind so much if the number of links were slowly decreasing as Google realises that the originating page 301's now. But we get more and more every week, dominating our incoming links.
And the bit I haven't said is that we had a number of these co-domains, for instance widget-shop.co, widgetshop.com -- and all of them are behaving in exactly the same way as described above. In other words, my incoming links are totally dominated by domains I own 301'ing to my principal domain.
If this site did emoticons there'd be a very unhappy face.
| 12:04 pm on Feb 7, 2013 (gmt 0)|
I once let Google crawl an alternative hostname for ~2 hours and it took more than 6 months for the damage to be repaired.
Once Google has spidered the data it's very difficult to get it altered.
Denoting the links as "via" is the same situation I ended up with. It does eventually sort itself out.
Two years on, there's still a few remnants of the effects of that upload-the-wrong-htaccess-file disaster to be found in the reports, but most of it has been fixed.
| 10:23 pm on Feb 12, 2013 (gmt 0)|
In case anyone's following this, the total number of inbound links from my bogus domain today DOUBLED to 12,048.
Traffic down again.
I don't even have anything like 12k pages. This is miserable.