Welcome to the Q&A Forum

Take a look at this video from Matt Cutts outlining their position on translated content.

In addition, as to the question of canonical element and hreflang you'll see that Google removed this portion from their guidelines with an update saying 'to simplify implementation, we no longer recommend using rel=canonical.' Check out the piece for the current position.

I also had advice direct from Christopher Semturs from Google who said directly to me "the golden rule I personally recommend to everybody using hreflang: In doubt, don't use rel-canonical."

Why not use a cross-domain canonical whereby you reference the pages on your primary website as the canonical version on your secondary websites, thereby eliminating the duplication.

For example on each page that is duplicate on your secondary website you would add the following to the head to reference the primary pages:

I'm presuming that because you are willing to largely re-write this content that you actually want to rank for it yourselves. As such a sizeable re-write would be required especially the blog titles, and you may even want to to target different keywords and long tail keywords. All the while bear in mind that quality and uniqueness is essential.

Alternatively, you could just use the content from the US-based blog directly on your Canadian blog but specify the US blog URL as  the canonical version. This would avoid any duplication issue and yet give you some useful content on your site, even if you would never rank for it in SERPs.

From my own experience, I would say the images that send us the most search traffic are those where the file names have been labelled correctly rather than autogenerated such as 33456789.jpg. I would certainly suggest there is some correlation between filename and search performance, although of course there are other factors to considered. As a recommendation, I'd certainly suggest file names are important although I couldn't put a measure of quite how important. There's certainly no reason I can see for us to stop labelling filenames correctly.

You definitely have the right approach. It's amazing how many people think they ought to rank for content even when it is duplicated on other sites. By adding some value to the content and making it more or less unique will only pay dividends.

It may be as simple as the fact that Google doesn't always display the page title that you want to display. As pointed out in this article Google can change the title it displays in SERPs depending on the terms actually searched for.

Presumably all your domains redirect to abctravel.com and this is the domain presented to users, in which case just add the main domain to Webmaster tools.

I agree, this is not an issue. It merely tells search engines which page is the 'canonical' version you want displayed in search. In fact I've seen it recommended that sites can use a self referential canonical in order that it  affords some 'protection' for your pages from content scrapers which automatically steal content.

One of our sites dropped 3 points for DA at the last Mozscape index so you are not alone. And this occurred at at time when our search traffic has been growing considerably and we are performing better for the majority of our keywords than we have for some time. For this reason I'm not unduly worried and considering it just to be a fluctuation which will probably improve in due course.

In fact, having looked at our competitors which I've been tracking for comparison, I've also seen that all of them have also experienced a similar fall in this latest index update.

Have you created a site map for each of your sub-directories and set a geographic target for each in GWT? In addition, depending on your content, and if you are duplicating it across the three countries, this may be an opportunity to make use of hreflang in order that you can tell Google which pages are alternatives of one another in order that the correct ones in served in SERPs.

If you have access to the site analytics I would advise that you take a look at the habits of users coming to the site. Have a look at the reasons they are coming to the site, what pages are they visiting, and what pages are they exiting on. Only by knowing what the user intentions are can you truly ensure that the site is designed with the user in mind.

Make sure that you make accessible as possible the very things that visitors are wanting to see. Of course making a site responsive will pay dividends, but you still have to make sure that the site infrastructure makes it easy for users to find the very things they are seeking.

Hi Scott, depending on the volume of new pages being created you could always make use of the 'Fetch as Google' contained in the 'Health' tab facility in your Webmaster account to make sure that your most important new pages are indexed.

Periodically you could add a new sitemap but on a daily or perhaps weekly basis submitting the important pages for indexing through your Webmaster account could be the way to go.

Article length should not matter. Just make it attractive and readable to the user. There are many sites that provide good lengthy content but break it up into readable chunks. I particularly like this example on the Verge site whereby a lengthy piece is broken up into neat segments and there is a useful 'jump to' feature on the left hand side which acts almost as a teaser to the content.

Is it possible that some of the pages have inadvertently had a noindex meta tag added to them or that certain pages are blocked in error by your robots.txt? Without your site URL it is very difficult for any of us to ascertain the issue.

Hi Philip,

By all means add the last part if you wish to give you some consistency in the series, but make sure you append it at the end of the title. One thing to mindful of is not to make the title too long or it may end up being truncated by search engines. This SEOMoz guide should help.

http://www.seomoz.org/learn-seo/title-tag

301 redirects are fine. They pass the majority (but not all) of the link authority. It is 302 temporary redirects that you should avoid as these are intended as temporary and therefore link authority is retained by the original URL.

In my view there is no issue with you using your default for en-us. It may not be necessary to reference en-us in the code as Google will know that www.mysite.com is your default for English speakers in the US in any case, but there is certainly no harm in leaving that line in there.

As to your second question, I always reference every alternative version of a page in the hreflang including the page itself, so in the example you have included  code on the page on the UK site. This is perfectly fine and is how I have always implemented hreflang when dealing with pages that are similar but intended for different countries/languages.

I'm afraid I can't offer any proof as such. I know there is an alternative way of implementing hreflang through xml sitemaps so you can avoid having to add the extra lines in the code within your pages. It's not something I've ever implemented but you can find more about it in the Using hreflang in sitemaps section of this article. Might be worth looking into.

I agree very much with Tom's view - consistency brings you brand awareness and users are drawn to authors they recognise. I know that I generally am.  I can't really see any benefit in using different images for your social media profile so would suggest using the same ones.

Have you submitted your URLs in a xml sitemap to the major search engines? In addition, check your robots.txt - it may be that you preventing certain pages from being crawled.