Welcome to the Q&A Forum

As an SEO best practice, there should never be more than one version of any page URL.  Even with Canonical implementation, it's only a signal, not a directive, and not all search engines even pay attention to that.  While it's perfectly acceptable to offer filtering options for people to discover content, filters should be blocked from indexing at all cost.  Leaving it up to Google or any search engine to have to figure out which is the original content is not a best practice even though many sites do it, and even though Google says it's okay. Trying to get one page found simply by inserting different keywords in a URL is also a crap-shoot since that's the only thing that changes about the page itself.  Like buying a keyword exact match domain only to use it as a redirect to the real domain.

Hey James, Good question! First off, Open Site Explorer actually uses Linkscape for its data, so data from the two sources should be the same. I checked it out and it looks like the root domain of www.hydropoolhottubs.com - which is all of the subdomains under *.hydropoolhottubs.com - has 102,418 total links pointing to it, while the subdomain www.hydropoolhottubs.com has 2,805 links pointing to it. This is probably because you have a lot of links pointing to a subdomain other than www, like to the root domain or to retailer.hydropoolhottubs.com. I hope this helps! Let me know if you have any other questions or if I can do more to explain it better. Cheers!

Hmmm... the only info I found about the domain I want to purchase is this: Registrar History: 6 registrars with 3 drops NS History: 11 changes on 7 unique name servers over 6 years IP History: 21 changes on 14 unique name servers over 7 years Can this be a red flag? On archive.org I could not find this website at all, but neither mine... which is not flagged... is archive.org accurate? Thanks!

Well thats a pretty easy experiment. And the results appear to be that a proxy is not necessary. I searched for SEO firm on google.co.uk from my office in San Diego and from a UK proxy @ http://goproxing.biz Here are screen shots of both SERPS. While not an extensive test it appears that US based IPs can get UK results when searching google.co.uk RkH6N.jpg hMle0.jpg

Well, we just want to show less links to Google than to the user (but the links for Google are still a subset of the links shown to users). The links we'd do as JS links are those to less often applied search filters, which we don't index in order not to spam the search index. Fortunately, if Google is smart enough in decrypting the links it wouldn't do any harm. Thanks for our ideas tough! Especially the site: thing I considered myself, it really takes ages until something is de-indexed (for us, using robots.txt did speed it up by a magnitude).

I'd recommend not using the canonical tag here for the following reasons: It's not what the tag is designed for.  By using canonical tag you're saying to search engines, "this page is the same as this other page so just ignore it."  Not true in this case. It seems like the pages you're noindexing are good candidates for it: they aren't pages that would be a good experience for users to land on from a search. IFor product pages that are no longer available, I'd use a 301 redirect to point users to the home page or a similar product - that's a way better experience for users who click on links to those pages (remember, it's about the users as much as search engines), and you preserve link juice.  I'd also just double-check and see if your "view cart" and similar pages are accruing many links; my guess is they aren't.  I'd keep those noindexed via robots.txt just because it would be very odd for a user to click on a search result and land there.  A good user experience is more important than the (my guess is very small) amount of link juice you might lose by not having them indexed.

Yes MonaVie as Steve Ollington has said, once it's realised that the two are one of the same, then there is no risk of penalty!

I don't think it will hurt to do that, but there is probably no SEO value to doing it either. It could be a good idea if you want to use an easily remembered URL in offline media and advertising thought.

Thanks Brad and I agree.  By the way, we are required to state the term of this sweepstakes (by law) which is in our official rules but I don't think we'll use the counter widget. We are also offering over 10 prizes along the way whenever a certain milestone is achieved.

Jarrett, the canonical tag you have in place will suffice. I'd suggest either to have users moving between the sites in the future or to modify your CMS to auto add the canonical tag in any future posts. It might be worth speaking to your syndication partners and changing your API to only show an excerpt with a link back to your site. This will also eliminate the duplicate issues and allow more of your pages to rank. It's highly unlikely for you to get your larger media partners to add a rel=canonical to your posts on their site as they simply don't need to, as they are more likely getting the content indexed. Aaron

Some of those headers might help you serve your page faster.  They might  save you and your users some bandwidth. I guess if you think that page load time is super important (most say it's a small factor) then you could argue those are important for SEO but genrally speaking Marcus is right.

@Steven,  I would suggest that you blog about all extensible keywords and link from your blog post to the main actionable page. You can use related keywords as anchor text. For example: web design, website design and affordable website design do not need three landing pages. You need to group your keywords though do not try to optimize one page for more than three keywords; it is difficult and takes time.

Thanks for responding back and letting us know the cause. It's helpful to know to ask about this the next time this question comes up -- and it has come up before (though I believe it was a different cause).