Welcome to the Q&A Forum

Your href lang tags don't seem to be ok - you should include all tags on all versions (incl. the own language - so the es version of the page should also include the es version of hreflang:

The UK version seems to have the es tag twice - remove the 2nd one.

You can check the implementation here: http://flang.dejanseo.com.au/ - a generator can be found here: http://www.aleydasolis.com/en/international-seo-tools/hreflang-tags-generator/

Not sure if it's going to solve your problem - but with your current implementation I fear you're sending mixed signals to Google.

Dirk

I doubt that this would be causing the issue - as far as I can see the way the tags are implemented is according to the Google 'standards' - check: https://developers.google.com/webmasters/mobile-sites/mobile-seo/separate-urls?hl=en . If I do the mobile friendly test - it's confirming that the page is mobile friendly so Google is reading the tags correctly.

Dirk

Hi Sarah,

Googlebot will follow these links as well and discover these "useless" pages (the are off course not useless from human perspective but they don't add value for bots - and they will be considered as duplicates). Duplicates are no reason for "punishment" - so you could just let them be. Personally I would put a nofollow on these links or add a "noindex" tag to the login page. Normally you shouldn't use nofollow on internal links - but login pages are an exemption on this (check also https://searchenginewatch.com/sew/news/2298312/matt-cutts-you-dont-have-to-nofollow-internal-links : "Of course, there are always exceptions to the rule, and things like login pages can be the exception. He said it doesn’t hurt to put the nofollow link for a link pointing to a login page, or things like terms and conditions or other “useless” pages. However, it doesn’t hurt at all for those pages to be crawled by Google."

For the practical part - if you add an additional question to a question which has been marked as answered - only the ones who have already answered will see the additional question. To be on the safe side  - it's better open a new question if you want other people to have a look at it.

Hope this helps,

Dirk

Moz is only indexing pages it's crawler is able to find. This implies that on your production site you have links to your development site.

Don't really agree with what your dev is saying - he should correct these links first; put a noindex on these pages. Alternative - put a password on the dev site so it's only accessible with a password. If a lot of users are putting links to your dev site it could become more important than your main site. Google will try to choose the most appropriate site - but you have no guarantee that it will choose the right version. In any case - that's not the type of risk you should be willing to take.

Once this is done - you can request a removal of these pages via the search console.

If all pages are removed from the index you can adapt the robots.txt to prevent access to the Google & other bots. Do this only after all pages are removed - if not Google will never find the noindex directive.

Dirk

My guess would be that it's related to the fact that most of the links are using javascript to get to the detail pages. While Google is able to follow javascript links it's still a risky strategy especially if your product is only accessible via these javascript links (you're never 100% sure if these links are followed)

Even when these pages are accessible via "plain" html links - it could be that the static links are hidden too deep in the site structure (a quick crawl with Screaming Frog that this could be an issue with your site - you have 60%+ of pages at more than 10 clicks of the homepage.

Try to make your site accessible ("degrade gracefully") for users (and bots) that cannot read the javascript links which should solve this issue.

Dirk

You shouldn't use rel canonical for pagination - it's main use is to avoid duplicate content issues. It's possible to combine it with rel next/prev but in very specific cases - example can be found here: https://support.google.com/webmasters/answer/1663744?hl=en :

rel="next" and rel="prev" are orthogonal concepts to rel="canonical". You can include both declarations. For example, http://www.example.com/article?story=abc&page=2&sessionid=123 may contain:

=> as you can see the canonical is used to strip the sessionid which could cause duplicate content issues - not to solve the pagination issue

With rel next/previous you indicate to google that the sequence of pages should be considered as one page - which makes sense if you have like 4/5 pages max. If you have a huge number of pages in a pagination this doesn't really make sense. In that case you could just decide to do nothing - or only have the first page indexed - and the other pages have a noindex/follow tag.

Hope this clarifies.

Dirk

None of the tools on the market is able to find all links to your site - and different tools will find different links. If you want to have the best view on who's linking to your site you will have to combine several tools (ahref/moz/semrush/search console/...).
Ahref used to be the tool who was able to find the most links (not sure if it's still the case) - which is in line with what you're seeing. Moz does a lot of processing of the data it's finding to calculate the page/domain authority - so they need to limit the index they work on (even if it's huge).

Dirk

When creating a campaign by default it will track all subdomains. If you want to limit only to subdomain select 'Exclude subdomains to restrict this campaign to the domain you enter' under the advanced tab. If I'm not mistaken you have this option only when creating a campaign - you can't change it afterwards.

If need want to track all subdomains with a separate campaign - you will have to upgrade to the next tier.

Dirk

Sorry - already deleted the crawl so can't give you the exact url's. If you do a crawl with Screaming Frog - go to the tab Directives - filter on noindex (or rel next/prev) and then check the Meta robots column for the conflicting directive.

Dirk

I tried crawling your site with Screaming Frog - and it crawled just fine. Personally I never use the tree view - but if you look at your site structure it looks pretty flat - so I doubt it's not following your subnav. It's accessible without js so it should not be a problem to crawl.

Careful when using keyword site:mydomain.com - the ranking displayed is not necessarily the ranking you will get when doing a "normal" search.

Even when you have more optimised pages inside your site - Google sometimes decides to show the homepage instead. Probably related to the importance of each page (I guess your home has more incoming links than the inner pages)

You should check your directives - sometimes you are using rel next/previous to group pages into 1 together with noindex (for 2nd and following pages) - which is confusing for the bots.

You also use canonicals for pagination -example  http://newyoubootcamp.com/testimonials/page/11/ has canonical: http://newyoubootcamp.com/testimonials - for pagination however you either have to use rel next/prev or noindex/follow - canonicals is intended for duplicate content not pagination.

Dirk

Can't you do it the other way? Put the temp HTML site on www - use the www2 to put the Drupal instance. Once the other is ready - remove the HTML & activate the Drupal version? Don't really understand what you're going to do with the www version while the www2 temp version is online.

If not possible - can only agree with Egol - it's quite a risk you're running. Quite possible your rankings will take a dive. As we all know, it's easy to go down but a lot harder to regain position.

Several elements here:

1. Migrating from subdomain to main domain - Moz "official" line is that this should be beneficial (check Rand's WBF on this https://moz.com/blog/subdomains-vs-subfolders-rel-canonical-vs-301-how-to-structure-links-optimally-for-seo-whiteboard-friday)

2. If you change your url's - 301's are still your best option. It's possible that it's not transferring all link equity - but it's still the best option you get. Watch out for redirect chains however. You might find this post interesting - also check the comments below: https://moz.com/blog/accidental-seo-tests-how-301-redirects-are-likely-impacting-your-brand. From personal experience - I have migrated several big sites to new platforms - completely changing the url structure and in most case you couldn't even notice the smallest glitch in Google Analytics. So it's not a given thing that your rankings/traffic will drop after redirects.

3. You mention that you also are doing a re-design - improving both design and content. This is the big unknown. Question is always if your carefully designed new pages are going to please your visitors. In the migrations cases where we had a drop of traffic this was very often related to the redesign - lower pageviews/visit, higher bounce rate - lower time on site. Few days after migration - traffic dropped (and it took indeed a few months to regain the original traffic/rankings)

4. Incoming links. I agree with Dmitrii: in an ideal world you should reach out to all the webmasters and they would kindly update the old links to the new ones. Realty is that some of them heard some SEO stories that outgoing links to other sites could get you punished by Google - and that it's always better to put non follow links or remove all the links to commercial parties. I fear that the risk of losing links here is bigger that gain.

Hope this helps.

Dirk

That's a literal reproduction of the post from John Mu

Similar to Ubersuggest: keywordtool.io

No - according to Google the value of the links is transferred to the https domain:

Check post from John Mueller on Google Plus: https://plus.google.com/+JohnMueller/posts/PY1xCWbeDVC

"#Do I lose "link juice" from the redirects? No, for 301 or 302 redirects from HTTP to HTTPS no pagerank is lost"

Dirk

The reward is not important - it's solving your issue which is the thing that matters. I don't know Wordfence - but I'm getting suspicious if a tool from Google (Pagespeed Insights) renders a 503 error. This could indicate that Google bot also might encounter troubles when checking your site. The best way to check is go over your logfiles & check for the responses you serve to the Googlebot.

Removing the country blocking now seems to enable pagespeed insights to work properly. Score is not great (50/100 - mobile) so more or less confirms the test by Webpagetest.org. Try to enable caching for static resources like images and enable gzip to compress data transfer. On top of that - compress your images (some of them are really heavy - check http://www.lesliekays.com/wp-content/uploads/2014/09/brake-hand-position.jpg . The other elements are probably more difficult to do - but these things could already get your score around 70%.

I would try to keep the Wordfence disabled for a week or two - to see if situation improves. Without the log files it's difficult to be sure  - but my gut feeling is that this is the main reason why the site isn't ranking. Working on site performance can only help with this - the suggestions above shouldn't be that difficult to implement.

Some other issues - your internal links are not optimal - you have about 600+ links going to 404 (a lot of them = images) - 7 with wrong formatted url (without .com) - 600+ url's which are redirected (in your internal linking you are mixing both www & non-www versions) - so might want to clean up a bit on these ones as well. It shouldn't be a reason for not getting indexed but it's something you should have a look at - especially if the site has some issues before.

I ran a crawl - but it seemed to go on forever - so it's possible you have some kind of eternal loop as well. I stopped at 1781 URI's crawled.

I would recommend you to invest the 99 GBP in Screaming Frog - which would really help you in identifying / solving these issues.

Dirk

A penalty is not irreversible - but with the effort you will have to put into it you might as well start on a new domain from scratch. This is what I personally would do. It's not that the domain is so powerful by itself - an exact match domain might give you a small advantage but on the other hand you will have to put much more effort to re-build the site reputation.

To be 100% sure - check the search console of the site -  using a "new" Google account not related to your current one; if you want to be extremely careful - do it on some external network - not on your own network.

Dirk

Great to post this at the top of the Q&A - noticed the first questions appearing today - probably many more to follow (hopefully fellow Moz'ers will read this post before posting the question).

While I don't disagree with the remarks of Donald - I don't think it's feasible to create really uniques descriptions for each type of product - especially when it's only the dimension which is changing (like the length you mention).

I think you mainly have to options:

• keep the duplicate content (which is not necessarily a cause for punishment: check https://support.google.com/webmasters/answer/66359?hl=en), especially if these pages currently are generating traffic & from technical perspective it's not feasible to use one page where you can select the changing dimensions

• use canonicals to a type of category page - which contains the general description with links to the products with different dimension

2nd would be my preferred solution - but again if not feasible I would just live with the semi duplicated content (we have some sites in this situation which are just doing fine). It's maybe a risk - but I guess it's not really a big one. You're not the only one in this situation - e-commerce sites selling common products like batteries or replacement parts are in the same situation as you. Many of them do just fine keeping the "duplicate" content.

Dirk

Sound like a good idea to avoid the duplicate content. Does it also imply that your visitors are going to land on a page http://siga-sverige.se/siga/fentrim-2 where they can't order the product? Or would that page be like a category page -listing all the different variations (which would probably make more sense)?

Dirk