|Domain name replaced in SERPS with alias domain name|
For one of my websites I have 3-4 additional domains which are 301 redirected to domain.com (they've always been 301 redirected). These alias domains are not just domain.net but also synonymA.com, synonymB.com.
Turns out that if I do a Google search for "synonymA.com" the main website will show up, but instead of domain.com in the URL field Google shows me synonymA.com instead!
Doing cache:synonymA.com gives me the cache of domain.com so we're not talking about synonymA.com being indexed.
Any ideas? Why would Google replace the original URL in the SERPS?
|Why would Google replace the original URL in the SERPS? |
For the basic reason that it actually was the query term. My assumption is that this is an algorithmic change to help improve click-through.
We know Google rewrites titles and description snippets based on the query - this is the first I've heard a report about rewriting a domain name.
I'm seeing it too....
Additionally, I just checked several domains of client sites that have .net versions which are 301ed, and they're showing the searched-for .net domain on the serp page. Mousing over the page title displayed in the serps shows the .net url in the browser status bar. Server header checks confirm that these are 301ed.
Searching a non-canonical variant on some domains will also display the non-canonical version, both in the url line and in the status bar on mouse-over.
This seems to be independent, btw, of whether the redirected domains had ever been previously promoted.
The behavior doesn't seem to be consistent with all domains that have non-canonical versions or alternate TLDs redirected.
So, Google fixed the "302 hijack" problem a few years ago (this was a huge topic for many months), only to now introduce a "301 hijack" method now?
If there is no domain ownership validation and server/IP hosting validation within the loop, there will be utter chaos.
What could be the possible benefit of this? I don't particularly want to see domains I bought in order to keep typosquatters off, suddenly ranking - especially if the canonical domain is nowhere to be seen in the SERPs.
Imagine I own hexagonal-purple-widgets.com but Google decides to list an alternative that I bought and redirected to the main site. So, Google lists hecksogonall-purpel-wiggets.com in the SERPs instead of the real domain. Users shouldn't be clicking that. It looks like it should be a malware infested scraper site. It shouldn't even be listed, but if it is a popular typo I might want to own it and 301 redirect it to the real site.
Now consider I want to visit the bank website at bank-of-podunk.com and log in.
If I am now being taught to trust that hecksogonall-purpel-wiggets.com in the SERPs will get me to the right place, why should I not click on the entry for bonk-of-podnuk.cn and give it my login details? Surely, that too will get me to my bank website, no?
It will be interesting though, to see what crazy domain names various corporations and institutions have been buying up over the years.
It will also be very interesting to see if buying acme-corp-we-suck.com and redirecting it to the corporate website can result in a situation where the company website appears in the SERPs but is listed only with the ghosted domain name and not the real one.
We live in interesting times.
Wow - this could be interesting - wonder if it will last?
I have a bunch of these situations (cause I usually have at least .com/.net/.org for important personal and client domains; and usually other permutations as well) and I'm not able to reproduce this behavior on any of them.
I think this is fairly clean (and I have no affiliation with it other than being a consumer of its services).
Our local alternative weekly here in Burlington, Vermont, is Seven Days, and its official URL is 7dvt.com. They also have a URL that's 301ed to that URL, sevendaysvt.com. When I search for "www.sevendaysvt.com" [google.com], their main listing shows "www.sevendaysvt.com/ - Cached - Similar", and there's the additional line, "More results from sevendaysvt.com »".
However, before we start wringing our hands, I'd want to know if there are other signals Google could be using to authenticate the alt-URLs a bit, e.g., IP address of the domain server. It's not clear to me yet that this is a gaping hole begging for spoofers to exploit.
[edited by: goodroi at 3:38 pm (utc) on Jun 20, 2011]
[edit reason] Fixed URL [/edit]
I have a handful of domains formerly owned by widget manufacturers, bought for the purpose of 301-redirecting into a hierarchy of widget information pages on my main site. The WHOIS info and registrar are different (e.g. the main site is a private registration at registrar A, and the 301'd sites are registered at registrar B with slightly different data). It is my understanding that Google can read private registration data (they're a registrar too, no?) so it is possible they've normalized both addresses and decided them to be equivalent.
The sites use different IP addresses, but within the same class C network.
Doing a google search for the FQDN of one of the 301'd domains, e.g. turbocharged-widgets.org, indeed shows that domain at #1, with the title and snippet of the page that users would be redirected to.
Curiously, if I search for the name without the TLD ('turbocharged-widgets'), my pages still rank highly but show the proper canonical URL and domain. IOW it is only when I search for the actual old domain, with the .com at the end, that Google rewrites the URL in the SERPs.
I'm also seeing similar behaviour over here in the UK.
Also rank highly for exact words in the domain split into individual words - but the destination URL is then show in the SERPs.
Is anyone seeing this phenomenon when doing a search that doesn't contain the 301'd domain name?
Otherwise, I don't see a problem. It's targeted at those less web-savvy people that still search for domain names.
There's a reason you or your client purchased these domains, right? If you changed domains, you still want to be found by those folks. Seeing a result they're not familiar with may turn them away. If you own a number of domains to prevent typo-squatting, it's beneficial to see the domain they typed in. They may have misremembered the exact spelling and would be reluctant to click on the actually correct result.
Is it that when google sees that several websites still link to the old domain, they decide to show it as a display url in the SERPS.
It probably exposes the dubious link building practices.
"It still remains to be seen whether Google restricts this kind of alt-domain in the SERPs only to domains that satisfy other criteria, e.g., their DNS has the same IP address."
This would make a lot of sense, or they could also check the Registration information for the domain, nameservers, or some other "Branding" algorithm.
It also may be possible they are doing this in order to maintain brand integrity from bought out brands or websites. If they are listing these 301'd domains individually, do you think this would effect the link juice value of 301 domains too? To me it would seem the Link Juice would continue to go on ranking the original domain and not so much of it being passed to the new one.
Even though it will still 301 on click, it still creates sort of confusion but also makes sure that WYSIWYG (What you search is what you get) so all the sudden a weird domain name doesn't show up in the SERPS were it wasn't supposed to originally, perhaps if what ranked before made more sense than what does now?
It's easy to learn what the canonical domain is. All those search results using all those extra domains currently will 301 redirect to the single canonical domain. The domain owner seems to be doing nothing sneaky at all - the flaw is at Google.
[edited by: tedster at 6:27 pm (utc) on Jun 20, 2011]
Ok - nuked a few examples here that were not necc examples of what was going on... (don't want to go down that he said/she said road of outtings...)
What is going on? Is exactly as the original poster said. I was sent a couple examples in pm that were surprising. It is shocking really.
I think there is a clear opportunity for hijacking here.
The 7 Days example really isn't that clean. The short form domain name, 7dvt -- the main domain -- uses the long form, sevendaysvt, as the link to the home page on every page. And there's also a a bunch of subdomains hanging off the long form name.
Looks like they simply confused the heck out of G and shows how G can be a bit buggy. Bing returns the short form home page when searching for the long, but does not confuse the two.
@jimbeetle, ahhh, good call. I'll keep lookin'. [:-)]
Still, I'm not convinced this is a real spoofer/exploiter issue unless Google does not pay attention to the IP address of the DNS doing the 301 redirect (or take some other kind of authentication measure). I.e., if we find out that Google lets any ol' 301 redirect from any domain server anywhere supplant the original domain in the SERPs, then we have a definite potential spoofing issue.
I was reflecting that this is almost as if Google were parsing .htaccess files. I was also noticing that I couldn't find any "large" prominent sites (large enough to be OK to use as an example) that were having this problem. Then a little bell went off.
Large sites generally have dedicated servers, and they most likely use httpd.conf rather than .htaccess. Could that be one of the distinctions involved here? I'm also noticing that sites I know that are hosted on VPS, even though they are using .htaccess for canonicalization, don't show these results.
I am seeing problems from sites hosted on IIS as well, but that's another can of worms.
This isn't a complete answer, but it's a start, I think, and maybe someone can build on it.
PS... I'm seeing that some sites that are hosted on VPS apparently do show these results.
If your webserver displays .htaccess files in a way that leaves them open to be parsed by Google you should have it checked. That is far from normal operation.
I run IIS and have examples - it is definitely to do with 301's
While the anomalies persist, Google appears at least to be no longer displaying noncanonical forms of a domain in the url line... ie, if you've 301ed the non-www version to the www-version, Google is now displaying the www-version. This is very different from what was showing last night. So perhaps this is a work in progress.
In some cases, Google is still displaying redirected old domains. I'm seeing a bunch of examples of domains redirected many years ago still displaying the old domain, but in the context of their new sites/destination domains. Redirected TLDs are also still displaying.
I am also seeing an interesting inconsistency, though, where there's different behavior for different aliases redirected to the same destination domains. In this case, olddomainA.com and olddomainB.com are both redirected to example.com, and...
- Google shows example.com for olddomainA.com redirect...
- but it shows olddomainB.com for the olddomainB.com redirect.
(This latter is an example from another mod, and I'm not familiar with the hosting or where or how the redirects were done... whether, eg, in the destination domain's .htaccess, or from different registrar or hosting accounts).
With the listings that display different domains...
- mousing over the page titles still shows the displayed (old) url in the task bar...
- but mousing over various types of Sitelink displays or additional results shows the new domain in fully canonicalized form.
This is almost like watching DNS changes propagate. This may be the wrong choice of words for a simile in this case (and maybe one shouldn't make technical similes), or it may be an exactly right one, but I'm not sure of the explanation.
Little by little some things may be returning to what they were. It's hard to know whether Google wants to display the original redirected domains, and I can't imagine why Google would.
I should add that the redirected domains are displayed only when you search for them, but I agree with g1smd that it's bad training to have people click on redirected alias domains. With banks, I should add, it's bad enough that many banks are careless enough to sometimes use different domains for marketing than for banking purposes.
|If your webserver displays .htaccess files in a way that leaves them open to be parsed by Google you should have it checked. |
ADovervik - thanks for that observation. These are unrelated sites at different hosting companies, with servers set up by different people, and a good many of these are not domains I work with. The simile I used here... almost as if Google were parsing .htaccess files... was probably not a good choice of words to describe server operation, but it was the best I could come up with very late last night. Shared hosting was the common feature which came to mind. It will likely turn out that these sites all do not have publicy available .htaccess files, but it's most definitely worth a look. Again, I can't imagine why Google would want to display this in the index if it were true.
I should add I've just gotten a report of at least one site on a dedicated server having this problem, so that most likely shoots that theory.
Looking at all sorts of behavior I'm seeing, I think this is a glitch, but I'm also thinking there may have been an underlying change in how Google is indexing redirects which may be useful to know about.
PS: Instead of sending examples to me, please examplify them... keyword1 keyword2, or whatever... and add an asterisk after http if you can't use example.com for reasons of description. Eg, http*//www.keyword1keyword2.com will keep these from becoming live links.
Happens with one set of domains I am looking after.
IIS, custom rewrite module, no .htaccess nor ISAPI
Shared hosting, both domains on the same IP, the old domain redirects to the newer one.
Exactly as Robert described - searching for the old domain shows title & description from the newer domain, but hover and URL in green shows the old domain. Click executes 301 redirect.
Ouch, this feels heavily exploitable...
|if you've 301ed the non-www version to the www-version, Google is now displaying the www-version |
Just a sidenote: this doesn't appear to be the case on UK Serps yet (maybe in progress).