Welcome to the Q&A Forum

Hi there,

This is one of the reasons why the hreflang tag was created. You can learn more about this here: https://moz.com/learn/seo/hreflang-tag

Also I would advise you to read Dave's post about it: https://moz.com/blog/hreflang-behaviour-insights

I hope this will clear it a little bit.

Keszi

Hi Russel,

My advice would be to go and check up the webinar section. There has been a quite good webinar last Thursday pointing out dos and don't for Google Places.

I believe it will give answer to all of your questions.

I hope that will help,

Istvan

Hi Chris, Maybe opensiteexplorer.org could be a solution? You could export the info and check anchor texts and links also. Gr., Istvan

Hi Salvyy,

I always told my team, that if we gain links from bad websites, that is in a some way harmful, BUT pointing out to such sites can flag us. (The same is told by Matt Cutts who pointed out that you cannot influence who is linking to you, but you can choose where to link to).

Therefore my advice would be, if it is not in your niche then don't link to these sites.

Gr.,

Istvan

Hi,

You can block RogerBot from Robots.txt

Check for further instructions on: http://www.seomoz.org/dp/rogerbot

"Please note: Adding this code will prevent our crawl test tool from being able to crawl your website."

Gr.,

Istvan

What I would add to Lewis's answer, just to make it clear, canonicals do pass link juice. If you are interested, you can check this article from Dr. Pete: http://moz.com/blog/an-seos-guide-to-http-status-codes

But for sure the 301 would be the best practice in your case. Because there would be no good use to have the pictures on 2 urls (the kws and the non kws versions).

I hope this helps.

Keszi

Hi Jackie,

What I'd advise you to do, is to create a Screaming Frog Crawl so you can compare the two crawl data. If both show missing meta descriptions on the same pages, then there might be something on your side.

It is always good to double-check the information you are provided. So you are sure that the issue isn't a bug.

Gr., Keszi

P.S. I made a quick crawl with Screaming Frog on the first domain: sundancevacationsblog.com and found quite a big number of Missing Meta Titles.

Hi,

I'd quote from Moz: http://moz.com/learn/seo/domain-authority

"How do I influence this metric?

Unlike other SEO metrics, Domain Authority is difficult to influence directly. It is made up of an aggregate of metrics (MozRank, MozTrust, link profile, and more) that each have an impact on this score. This was done intentionally; this metric is meant to approximate how competitive a given site is in Google.com. Since Google takes a lot of factors into account, a metric that tries to calculate it must incorporate a lot of factors, as well.

The best way to influence this metric is to improve your overall SEO. In particular, you should focus on your link profile—which influences MozRank and MozTrust—by getting more links from other well-linked-to pages."

So answering your question, Yes, you can improve the DA, while working on your SEO.

In general, in my belief, the 301 redirect could have influenced the score, but I would rather check the rankings and organic evolution of the website after the redirect instead of checking only the DA. If the organic results improved, I wouldn't worry to much about the DA itself. (maybe it is just only a temporary glitch until Moz recalculates it's value based on the new information they get).

Keszi

Hi there,

Let me quote Google for this:

• Average position: The average top position of your site on the search results page for that query.

  To calculate average position, we take into account the top ranking URL from your site for a particular query. For example, if Jane’s query returns your site as the #1 and #2 result, and David’s query returns your site in positions #2 and #7, your average top position would be 1.5.

Source: https://support.google.com/webmasters/answer/35252?hl=en#details

I believe that they do take in consideration also the local results. (but that is my opinion)

Gr., Keszi

Hey Miranda,

I would go for one of the following:

1. usage of canonicals (link rel canonical to domain.com/jobs-in-london/)

2. the dev team could use cookies to keep trackings (I am not a developer guy, but this issue has been solved many times with cookies) - Developers please correct me if I am wrong in this.

I hope it hepled,

Istvan

Hi there,

I underlined it many time to others and I will underline this fact to you too: before you are trying to "hire" someone to get a job done, try to get to know, what that specific person is going to do for you.

In your case I would advice you to go and check the following two videos:

1. www.seomoz.org/blog/which-link-metrics-should-i-use-part-1-of-2-whiteboard-friday
2. www.seomoz.org/blog/which-link-metrics-should-i-use-part-2-of-2-whiteboard-friday

These two take up 30 minutes, but they will give you the basic knowledge of how to measure the link metrics.

BTW: if you really want to climb in "SEO" try to deliver something real good and you will achieve the goal.

Regarding the SEO company you have hired: my advice would be to sit down with them for 1 hour meeting and ask the following question:

1. What is the strategy they are going after?
2. Why did they choose that?
3. How will this point you out from the crowd?

Taking rough decisions like Hire of Fire is not specific to SEO. There might be strategies which might not show up too much in the start, but on a long run you will benefit from their campaigns. A good SEO company knows that your goals are their goals too! If you are happy as a customer, you will stick up with them.

Another advice: SEO consultant and Client relationship should be compared to marriage. I know it sounds ridiculous, but if they hide things from you, they are going after sneaky goals, it will end up with a divorce. And a divorce is not good for neither of you.

Final thought: Go ask this question question to your SEO too, ask them why is it good to have you submitted in those directories, if it is delivering any plus value, etc. etc.

Good luck,

Istvan

Hi Geraldine,

So, meanwhile just jumped in what would be most cost efficient implementation for the HREFLANG on your website.

Give the following article to your developer: https://support.google.com/webmasters/answer/2620865?hl=en

Basically he will create a sitemap for each of the websites, which will indicate what is the alternate version of every single page in other languages.

You will then submit this sitemap to EACH of the Google Webmaster Tools accounts for the websites (.fr, .de, .com, etc.)

This should be the fastest and most cost efficient in your case.

Reminder: Remove the Hreflang link from current versions. It is the best to indicate alternate via only once channel.

Gr., Keszi

Hi Pedro,

You will find your answer in this article: http://www.seomoz.org/blog/canonical-url-tag-the-most-important-advancement-in-seo-practices-since-sitemaps

I hope it helped,

Istvan

Hi Alex,

Related to Image Optimization Kasy Allen wrote a very nice post back in august 2011. Read it through, it will provide answers to your question

Here is the link: http://www.seomoz.org/blog/is-optimizing-photos-more-important-than-you-think

I hope that helped,

Istvan

Hi!

Like Jamie has mentioned it, you have a HREFLANG issue on the site.

Try implementing for each of the domains the following code:

<- default for international
                  <- default for US
              <- default for Aus
                             <- default for italian language!
                             <- default for french language
                          <- default for german language!

I would implement it like this.

The difference is compared to Jamie's implementation idea, is that I wouldn't put the .com version as the x-default.

All of the hreflang markups have to appear on each of the pages! For example:

Read through the articles Jamie provided, they will be a huge help to kick-start the implementation.

Gr., Keszi

I would underline something related to your question: I take myself as an example. - just so it is clear for everyone

I had talks with other mozzers who told me they appreciate that I am investing time in answering questions here in Q/A. That felt like... GOOD! That was the best feedback ever for me! (and if you read this through - "You know Who" - Thank you! ) - I am not pointing out with a finger, that is not the point. The Feedback is!

Ok, now back to Alan's question: Why it is important to put thumbs up/thumbs down, mark it as a good question or not: with every question I read and I try to answer here on SEOmoz Q/A I am learning something new (if a client comes and asks the same question I would need to answer it in more detail, so I am testing, preparing and constantly learning in here) - every one of us is a man, we can make mistakes, but we learn from these mistakes only if we get the feedback. So please-please-please don't hesitate to give us a feedback, we learn from that (even if it is a negative one).

@Alan: I am glad you brought it up. I was thinking on writing something similar too.

Cheers,

Istvan

Hi there,

It can take quite some time until you will see effect of the disavowed links file. For example in one of our projects I've seen it took almost 6 months until we have seen some recovery after submitting the disavow file.

Also what I'd advise you to check if your rankings were only effected by the lower quality links you had in the past. One tool that might help you with that is Barracuda's Panguin tool: http://barracuda.digital/panguin-tool/

Basically it will use your Google Analytics data, and overlap with the known google updates. That could also highlight if your website was hit by another update meanwhile.

I hope this helps a little-bit.

Keszi

Hi Daniel,

Having a friendly/readable url structure will surely help you in several ways.

You can read more about it here:

• http://moz.com/learn/seo/url
• http://moz.com/beginners-guide-to-seo/basics-of-search-engine-friendly-design-and-development

In the second article check the "Url construction guidelines" part.

But before you start implementation on such a development, try to check how you could avoid an url structure redesign failure. (Such as forgetting about 301 redirects from the old structure to the new one).

I hope it helped. Gr. Keszi

Hi Erik,

In the campaign setup you have 3 different possibilities:

1. Subdomain (Examples: www.seomoz.org, guides.seomoz.org, pro.seomoz.org)

2. Root Domain - Track at root domain level. Track all the different subdomains within this root domain.

3. Subfolder (Examples: www.seomoz.org/tools, guides.seomoz.org/beginners-guide-to-search-engine-optimization)

Now, if you want to track the www. version of you site, go and use the first possibility and track www.yourdomainnamehere.com

Hope this was helpful!

Istvan

Hi,

Would you be able to post a URL? Maybe there is something you have missed with the implementation. More eyes, see more.

If you cant/wont post it publicly, send me a PM.

Keszi

Hi Luke,

I would implement a canonical in any situation where there can be parameters in the url. This way telling any search engine to only consider one (the original) version of that page.

Gr., Keszi

Hi Will,

Before taking a decision like this, maybe you should analyze why was it deindexed?

I have seen websites getting deindexed because of technical issues, it might be worth taking a look.

Greetings,

Istvan

As Alick300 has mentioned, Pagerank is a metric that we do not use anymore.

I would advise you to check the metrics that Moz is using: PA, DA, mR, DmR, and in case you are a Pro member, you could also check mT and DmT:

1. PA: Page authority: http://moz.com/learn/seo/page-authority
2. DA: Domain authority: http://moz.com/learn/seo/domain-authority
3. mR&DmR: mozRank and domain mozRank: http://moz.com/learn/seo/mozrank
4. mT&DmT: mozTrust and domain level mozTrust: http://moz.com/learn/seo/moztrust
5. spam score: http://moz.com/blog/spam-score-mozs-new-metric-to-measure-penalization-risk

Be aware that these are only metrics and in some cases increasing these metrics doesn't necessarily mean that you will rank better: Correlation <> Causation (check the article I have linked into the spam score, by Rand).

Also you could check the whiteboard Friday: http://moz.com/blog/understanding-and-applying-mozs-spam-score-metric-whiteboard-friday

I hope this will be helpful to you.

Keszi

I'd rather not use images that are not licensed or owned by my company.

That's just my personal opinion.

Gr., Keszi

Hi there!

First of all, I believe that you shouldn't use both canonical and rel=prev/next. The two techniques do not work together well: "In cases of paginated content, we recommend either a rel=canonical from component pages to a single-page version of the article, or to use rel=”prev” and rel=”next” pagination markup." (quoted from http://googlewebmastercentral.blogspot.com/2013/04/5-common-mistakes-with-relcanonical.html"

Basically you have more possibilities:

• Do not set up canonical or rel="prev/next", but use noindex, follow meta robots tag for pages 2+ (kind of old school, but still works as a charm)
• Implement Rel="Prev/Next" but without a canonical! the two elements do not work together (as mentioned above)

I think the best method for you would be to have a rel="prev/next" and have the canonical removed.

I hope this helps, Keszi

Hi James,

First of all you need to categorize these 404 pages, some may come from website sections that were deleted in the past, and haven't been addressed. Other 404 URLs could appear through domain redirects to your website. (unfortunately, these are harder to find, process and resolve).

For the first category (when sections have been deleted, moved, etc) you will have to ask yourself which is the correct way to resolve the issue? 301 redirect to most relevant URLs? Or just 410 and let search engines know that your page are deleted, and URLs should be removed from index. Please don't start redirecting every single 404 URL to your homepage, or any irrelevant page, or you will be creating soft 404s.

Regarding the /insertgibberishurlhere type of URLs, you should check what kind of domains are redirected to your domains (I have seen domains that had this kind of 404 not found errors via massive domain redirects to a project). First of all, if this is the case, you need to ask yourself, what are you redirecting to your website. If the domains are not in your hand, you could also redirect all of these to a 410 status code url on your website.

Oh, and the obvious version: crawl your website with tools as Screaming Frog, and make sure you are not creating the 404 URLs yourself. (almost forget about this version)

Let me know, if you have further questions.

One good advice: don't let the frustration make you take decisions

Work hard and you will benefit from it and over rank them.

Good luck!

Istvan

Hi Martins,

In my belief, if the page product/review is not linked from any indexed page with a follow link it makes it kind of hard for search engines to reach and index the page.

This is my opinion, but I will document a little bit and come back to you later on.

Istvan

Hi Hiroaki,

When you have problems like this, maybe it would be advised to contact the Help Team and let them know what system you are using. Maybe they will be able to rebuild the "bug" and resolve the issue faster.

You can contact the help team at help@seomoz.org

Gr.,

Istvan

Hey Nextman,

I have encountered the same issue a half year ago, with mini-site. The site was moved 3 times in one year, still there we didn't see no negative affect.

At the end of this video: http://www.youtube.com/user/GoogleWebmasterHelp#p/u/24/PfqKKquA_R0 Matt Cutts says that "if you are not happy with your current webhost, you can change, it is not the end of the world"

I hope that helped.

Cheers,

Istvan

Hi Remko,

First of all I would check for the domains that I want to buy and redirect for relevancy and their backlink profile.

The example you gave:

Fietstassen.eu - > Loodgieter.nl for me it was a little strange. (Correct me, if I am wrong, but Fietstassen is bicycle bags? and Loodgieter is repair/handy man?).

I did not check in details what kind of link profile Fietstassen has, but if my dutch isn't cheating on me, I might have a feeling that it has collected non-relevant links to your target website.

As for your first example: Glaspunt.nl -> Glas.nl if I get it right, these two seem to be in the same industry, so buying glaspunt.nl could be an idea.

As in general:

If you buy a domain, just to redirect and gather their link juice:

1. look for relevancy first.
2. check that domains backlink profile (just to be safe. if it only has a spammy link profile, I do not know if it would be a good idea to point it towards your project)
3. if the domain that you are buying would be more relevant to a category of your website, I'd redirect it towards there instead of homepage.

I hope this helped.

Gr., Keszi

Hi Sorin,

I would work this issue out trough Google Webmaster tools. It would be a possible solution to set up the parameter and tell Google to ignore it.

This way you wont have duplicate content issues because of the affiliate parameter in your links.

Another solution to this would be to insert in each and every page a canonical to itself (this way example.com/?affiliateID would have a canonical set up to example.com)

Another one: the solution that you have provided, but I wouldn't redirect all of them to the homepage. Just think as a user, you go on a link that should point to a product, then you get redirected to a homepage... what would you do? I would close the "tab".

301 and canonical will loose some of the link juice provided but still pass on somewhere close to 85-90%.

I hope this was helpful,

Istvan

Hi Emmanuel,

If you want to have a quick crawl then maybe it would be smart to use: http://pro.seomoz.org/tools/crawl-test

I am not aware of starting a new crawl faster with the campaign which you already set up.

Gr.,

Istvan

Hi Tai,

I would prefer sub-folders, as I have understood the PR flow better to sub-folders than to sub-domains.

Here are a few useful resources:

http://www.mattcutts.com/blog/subdomains-and-subdirectories/

http://www.seomoz.org/blog/subdomains-subfolders-and-toplevel-domains

http://www.seomoz.org/blog/understanding-root-domains-subdomains-vs-subfolders-microsites

I hope it helps,

Istvan

Hi there,

Basically you are saying that you'd like to host every language on different domains. That shouldn't be an issue. If the text is translated, you could even create a sitemap for language targeting and submit it. This way Google will know that Page A is a translated version to english of Page B.

You can read more about the HREFLANG here: http://moz.com/learn/seo/hreflang-tag

There is also a tool for generating the correct sitemaps, which will make your job easy. You can check it here: http://www.themediaflow.com/tool_hreflang.php

I hope it helped, Keszi

You should then open a ticket in the help desk. seoMoz staff will pick it up from there.

Sorry that is all I can do for you guys. Good luck with resolving the issue!

Istvan

Hi Peter,

Regarding agencies that are only working with link building I have terrible experience. I prefer to do link building with in-house people. This was it is a way easier to check what they are doing and how they acquire the links.

The problem with link building agencies is that they usually don't work on the quantity vs. quality . And although quantity is important, but quality is 1000x more important for me.

There are several SEO agencies that will tell you they will get x number of PRX links. The question is: how will those links affect your visitors? I am always trying to gain links that are win-win situations for both me and the "potential link partner", answering questions such as:

1. Does that link provide information, +value for the reader?
2. Is that link complementary to the content where it is placed?
3. Does the visitor, who reaches our site find what he was looking for through that link?
4. Is that link valuable for robots?

I have left robots for the last position since I believe we should add less value to this and give more importance to visitors.

But this is one personal opinion, this is how I try to "make the web a better place".

I hope that helped,

Istvan

Hi Adrian,

We had a similar problem not as long ago (changing the website extension from .html to .aspx). What we have experienced is, that the website traffic and rankings went down for aprox. 1-2 week, then it came back up without any problem.

I would suggest to go for all change at one time instead of going with partial rewrite, then again some partial rewrite. (Before you put it alive, test all your links! check for broken links and make sure the redirects are right)

After the website deploy you should resubmit a new sitemap in GWT.

The 301 should be done, so after resubmitting your sitemap in GWT you will not face any duplicate content issue.

The negative part: you will lose some of the link juice thanks to the 301. If you have the possibility contact the webmasters who are already linking to your website and ask them kindly to resolve the URL issue (at least for the highest authority links you have gained in time).

I hope that helped,

Istvan

Hi there,

Depending what are your intentions with these websites I would say that you have two situations:

1. Using the websites in order to gain "higher visibility" with several domains: now IF i "translate" this, then I would be saying that you are using the "mini sites" to gain link popularity... If this is the case, I would dump the idea. Still if your client is chasing this idea, then keep them separated. (Although Google does excellent job on detecting these private networks, and my belief is that one day it will punish it too)

2. Using the websites to target different zones (geo or niche subzones): Now if you create this to have several areas targeted, let it be geological (such as creating websites / country, region or city etc.) or niche related (such as having a website for eletronics, one for books and another one for IT products of one big shop) then I wouldn't be afraid to connect these within one account. These websites are not created in order to create private linking networks! These have as goal to create smaller websites for targeted "personas".

In general... IF you are creating websites in order to only optimize for Search Engines... Then don't connect them!

I know there will be several opinions regarding this issue... This is mine, I hope that helps you out in some way.

Gr.,

Istvan

Hi Jeff,

First of all, as previously mentioned in another question of yours I would recommend redirecting from non-www to www or vice-versa. And after that using only one version of the URL to link to your website.

Now related to this question: I wouldn't bother myself with PR.

If working with directory listings I would go to the most popular ones from my business niche. For example here on SEOmoz we have a list of web directories which were recommended by mozzers: http://www.seomoz.org/directories

I would recommend you to read through the short text on that page, it will give you on overview where to submit your listing.

Greetings,

Istvan

Hi Mof3kz,

Whenever I had to deal with a similar situation to yours I was always checking what we have done, what did change in the past few weeks.

Even, if you have been working with some great link partners, take your time and write an email. Ask them if they have experienced a fluctuation lately in the search engines on their ranking.

Possible situations:

1. The major link partners have experienced big fluctuations, maybe their links profile had a bigger change and since you are connected it will in some way effect you too.

2. They didn't experience any fluctuations: this is when you ask yourself, ok, then what the hell happened here? Stay chill! try to get on-site ask yourself if you are optimizing the right way. Try to avoid any possible issue from your site, add better content if possible, check for crawlability, etc. etc.

3. The link partners tell you they have experienced some kind of fluctuations, but not big ones. Now this part is tricky. As we all know, search engines like Google are constantly changing the algorithms, trying out new filters to show up the most relevant content. Some filter (new or existing one) has triggered out your possible loss. It might be temporary, it might be something stable. In these cases I usually advice the same thing as @ the point 2.

I hope this resolves your situation! Good luck!

Gr.,

Istvan

Hi Greg,

My advice would be the following: take 30 minutes brake and watch the following two videos:

1. http://www.seomoz.org/blog/which-link-metrics-should-i-use-part-1-of-2-whiteboard-friday
2. http://www.seomoz.org/blog/which-link-metrics-should-i-use-part-2-of-2-whiteboard-friday

After you are done watching the two WBFs by Rand, you will understand what can cause a big difference between SEOmoz's mR and Google's PR.

I know that the first impression is that this kind of sounds out of topic, but believe me it will help you a lot. (It did help me understand several things).

Good luck!

Istvan

Hi André,

First of all I would advice you to read through the following article: http://www.johnfdoherty.com/hreflang-markup-testing/

John Doherty's article will basically answer your question.

Gr.,

Istvan

Hi TBiz,

The fact with these website is the following: they have a lot of sitewide links (link that appear on more then one page of the linking website). For ex, footer links (such as "created by" or "designed by", etc.), advertising which appears on each page, etc.

Although it can drive a big number of inbound links, I would still go and ask the question of quantity vs. quality? Which one is better for a website? (If it would be me, I would choose quality).

I hope that helped,

Istvan

Hi Simon,

I usually do prefer the www. version, but I believe there no major difference between the two versions when you are building up a fresh website.

Still, it is important to underline the fact, that you need to choose one of the two versions and eliminate the other one be redirecting: www -> non-www or non-www. -> www.

This way you will eliminate many duplicates from the start (and you will help your link juice flow better).

I hope it was helpful,

Istvan

@ JenAdams

I have to say that I am quite sorry that you didn't get the idea.

No offense on this one, but Matt and Jeffrey put down some nice ideas here. Buying links is not a solution. I must admit, there are payed links in our industry(unfortunately), BUT the questions would be why not invest in something unique instead of buying links?

And I said Unfortunately above because I think that there are niches which took advantage of these.

If we were in the '90s, getting 100 more links compared to your competitor could have got you #1, but today you have to build on value, you need to invest more on content and development. Semantic search will be awesome and links, in my belief, will count less.

Good work Big G. to having a public fight against payed links

BTW: I would give Double Thumbs up for Matt's answer. I quote:

"Not true. I agree that most do, sure. I also think that most do not understand the dangers and pitfalls of buying links. They probably also think that the more links the better,., regardless of quality of link and the quality of the content on their site."

This is worth printing out, pining on my wall.

Hey there!

I think I have found what your problem is with you canonical link

In your code you have:

And probably you are somewhere forcing the URls to have a / at the end.

So basically you are confusing browsers and search engine bots, because they now cannot tell which is the real version:

SE enters the page. Then it sees that the right version should be the one WITHOUT a "/" at the end, then that pages has a 301 redirect to the version which HAS a "/" at the end of the URL (but that has a canonical which points out that the preffered version should be ). So it is a non-ending circle.

So if you add a / to the end of your URl, your problem should be solved.

Final thought: Screaming Frog is working well.

I hope this was a solution.

Cheers,

Istvan

Well, this was expected... I'd say.

I have a feeling that if 2011 had major changes in the SEO world (to focus more on the quality vs. quantity), 2012 will have a bigger impact on search results.

You should add the Meta Description to your website not to your campaign. If you need help, I have sent you a Private message.

Greetings,

Istvan

Hi there,

Regarding website.com/category/product

What you have to take into consideration: if a product is going to be placed in more than one category, then this product is going to be indexable on more than one URL paths. (Like Gaston mentioned below, in this case you need to take care of duplication, which you can do with canonical links or redirection to one path).

For example, let's say you have a product which is in both cat1, subcat1 and cat2. This way you will have minimum 3 available paths to the product:

• website.com/cat1/product
• website.com/cat1/subcat1/product
• website.com/cat2/product

This means that on product level you will have to deal with internal duplicated content. This is why usually prefer to use website.com/product url path (IMO).

Regarding the /category/subcategory/ vs /category-subcategory/ this is really a technical question. How "deep" is going to be your website? Do you want a flat infrastructure?

I usually prefer the /category/subcategory/ structure, because of the idea of contents of an e-commerce website should be build up with a structure based upon content silos, where from general you move towards the specific (indifferent of how we achieve this, with subcategories or filters within categories). I really hope it helps you answer your questions.