Is there a subdomain for each language?

Is there a "default" language that exists on the primary domain (i.e. English on www.yourwebsite.com)?

Are you using rel href lang, X-Default Hrefe Lang, rel canonical, or other "tags" or solutions for this at the moment?

If you could share the website address we wouldn't have to ask these questions, by the way. I know that is especially tough on client sites, but I just wanted to say that Dave has asked some vital questions, including the one about setting each subdomain as its own language / geographic target audience. Here is what he is asking about: http://support.google.com/webmasters/bin/answer.py?hl=en&answer=62399

It is not necessary to have completely unique content for each language, as someone suggested below.

Bryan I really don't think that is what Dave is saying.

Hello Tom,

Five or six years ago this may have been a problem, but I seriously doubt it is right now. The text is clearly visible to humans and bots and you are not doing anything malicious. The div with the text appears over the image and it views fine in text and full mode when looking at Google's cache of the page.

Though both Pedram and Donford have answered your initial question well, I could not officially "endorse" them because I don't advocate keyword density ranges and do not agree that a manual review of the site would result in you being subject to any kind of penalty.

Write naturally for the topic at hand, focusing content on the correct keywords and topics with the goal of being helpful to the reader. Don't worry about manual checks for this because it will be obvious to the person looking at the site that nothing fishy is going on.

Good luck!

I like the idea of sending affiliate links through a redirect on a subdomain because you can kill off all of those links at once by removing the redirects so they don't affect the final destination page should you need to clean up your link profile of page-rank-passing affiliate links. Whereas, if you had those links go straight to the destination page you'd have to ask the affiliates to remove each and every link, and/or use the disavow tool on them. I have heard of people sending those though a URL shortening service instead of a subdomain, which should work just as well since they are 301s.

I'm not sure how much sway keywords in the redirecting URL have, but it isn't a bad idea to keep them contextual as you described above. That just seems to make sense anyway.

Since the affiliate.subdomain URL is going to 301 redirect it won't have a canonical tag. You could put a self-referencing canonical tag on the final destination page if you wanted to.

The number in the URL is fine.

This seems like a good topic for discussion so I will leave it open for you.

PS: My answers assume you know the risks of passing pagerank through affiliate links and are willing to take them.

Hello PFX111,

I checked the source code of the page, as well as the text-only version, and source code of the cached page in Google/Bing/Yahoo serps. I did not see any questionable content. If they are hacking and cloaking they're doing a darn good job of it, but I don't think that's it. Like everyone else, my first reaction was to say you were hacked, but the site appears in the SERPs without any message from Google about it being harmful to your computer, and I see no evidence that this spammy content appears on the page even when Google views the site.

Here are my two thoughts at this point:

1. It could be referral spam.

2. Someone else could be using your Google Analytics account code (it looks like UA-######-#) on their site.

In both cases you would see the keywords in Google Analytics but not Google Webmaster Tools. You can set up a new profile in Google Analtics and set it to track the domain (instead of just the relative path), which could bring up other domains if they are using your Analytics code. See this page for details on how to do that:

http://support.google.com/analytics/answer/1012243?hl=en

I love public Q&A because everyone gets to chip in, but nobody wants to share the domain in question (which is understandable) so that makes the job of answering a question really difficult.

Can you hide the actual domain name but provide some examples of URLs? For instance:

ourdomain.com/honolulu/four-seasons?rooms=4&view=0&page=1

Did you try any of Dr. Pete's suggestions? If not, I would implement one of those first, as they are still as relevant today as they were when he wrote them. Rel next/prev has received a bit more attention since then, but it only solves part of the problem if you're dealing with parameters beyond simple pagination (e.g. rooms, views, etc..).

From the information provided above I would probably go with a rel canonical tag to fix this issue.

I would not rely on a rel nofollow tag on links pointing to variants, as was suggested by Smarties, because Google is going to find those URLs regardless and a no follow tag on a link doesn't tell them not to index it.

Smarties #2 suggestion sounds good but I'd allow them to be followed. i.e. robots meta noindex,follow as opposed to noindex,nofollow. This allows pagerank from external links to flow through non-indexable URLs.

Good luck!

It can be tough to turn an exact match domain (EMD) into a branded domain sometimes. I would be careful of labeling this as any certain algorithmic filter / penalty. Back when EMDs were being devalued I'm sure there were a lot of false positives. This could be one example. It could also have to do with anchor-text over-optimization that isn't specially "penguin" related.

I guess my first question would be: Why do you think this is related to the use of the client's brand name? Is it just those keyword/s that have lost rankings? Is it limited to certain landing pages, keyword phrases, etc... ?

What is the domain / keyword? And if you can't provide that, how about a close example? For instance, is it something like:

buy-blue-widgets.com

Or more like:

bluewidgets.com

Or like:

bluewidgetsincorporated.com

or ???

The first example is difficult to turn into a brand. The second one is doable. The third shouldn't be seen as an EMD at all.

Hello James,

Why do these pages have "no SEO value"? Is it because they are AJAX pages or because you have them noindexed? Or both?

To answer your original question, using an on-click javascript event to send a user to a page other than the URL listed in the href tag is borderline. It goes beyond the risk level I would feel comfortable with on an eCommerce site, but a lot of affiliate sites do this. For instance, all of their links out to merchant sites may go through a directory called /outlink/ so the href tag might look like .../outlink/link1234 and appear to send the user to another page on their domain, when actually the user gets redirected to the merchant's (e.g. Amazon.com, Best Buy...) website. Sometimes the user is redirected from the /outlink/... URL and sometimes they never even get that far because the javascript sends them to the merchant's URL first.

It is not cloaking unless you are specifically treating Google differently. If Google doesn't understand your site that is their problem. If you have code that essentially says "IF Google, THEN do this. ELSE do that" it is your problem because you are cloaking. Make sense? There is a very distinct line there.

The bottom line is if you want to show users a certain page then you should be showing that page to Google as well. If the problem is the content on that page doesn't appear for Google (e.g. AJAX) then you should look into optimizing that type of content to the best of your ability. For example, look into the use of hashbangs (#!) as in:

https://developers.google.com/webmasters/ajax-crawling/docs/getting-started

PFX111,

Please provide the domain if you could. This will make it much easier for us to help you.

I seriously doubt this is a result of someone doing "negative SEO" but could not say for sure unless I can look at it myself. One thing I'd like to know is the landing page URL where these terms are sending visitors on your site from the SERPs.