Welcome to the Q&A Forum

Hi Felipe,

You raise an excellent point about paid keyword traffic in GA. The reason SEOmoz hasn't offered this until this point is that:

1. Integration with Google Analytics is still pretty new to the PRO platform - but new features are being added all the time.
2. Most of our historical focus has been on promoting organic traffic - but we could definitely do a better job of making room for paid traffic as well.
3. This fits most of SEOmoz customers needs.

That said, there's a lot of validity in wanting to look at paid search volume and it would make a great feature request. (I encourage you to submit and vote for this feature)

Part 2 - If I understand correctly, you want to combine/import paid traffic data from Google Analytics with the keyword report csv from your PRO campaign. The good news is this is simple task in Excel and easy once you get the hang of it, but it takes a little learning to get there.

The full answer would require a short book (or really long blog post) detailing each step - but fortunately by learning the basic functions, you can do all this yourself and more . It might take and hour or two to get the hang of it, but you will be armed with the knowledge to tackle all kinds of common Excel challenges.

To combine data like this, what you are looking for can be achieved in Excel using VLOOKUP or INDEX/MATCH functions. The best place to learn these is in Distilled's excellent Guide on Excel for SEO. I worked myself through this guide about six months ago, and now I tackle "advanced" Excel tricks with a bit more confidence.

Hope this helps! Best of luck with your SEO.

Hi Max,

As you know, SEOmoz uses a CDN (Content Delivery Network) to host our static content. This greatly improves the load time of our pages by distributing our content across a cloud network, and results in an improved experience for users.

If I understand your question correctly, you have set up a CDN and have created duplicate content issues.

To solve this, it's important to set up your CDN only to serve static content, like images, stylesheets and javascript. That is what a CDN is designed for. Do not duplicate your entire site - your HTML - as this will cause duplicate content issues.

If for some reason you need to replicate your entire HTML, then there are some steps you can take to mitigate the damage, although it's going to depend on your exact circumstances.

For example, you can set full URL canonical tags so that all your mapped CNAMES point to your primary URL.

To revert back to one copy of your HTML, you might want to put 301 redirects in place on the duplicated content (pointing to the original) before removing them from the CDN.

But even these aren't ideal solutions. It's best just to serve your static content, and only one version of your HTML.

Hi Eugene - It sounds like you are doing things right.

Open Site Explorer diplays Link Metrics for every single page in it's index as if it is independent - even if it 301 redirects or points a canonical link towards another page.

For example we can see that the inbound link metrics for the SEOmoz.org/blog differ from the metrics for SEOmoz.org/blog/ (note the extra backslash).

One advantage of this method is that we can see if our external links are pointing to the right page, and correct them if necessary. At this time, OSE will notify you if the URL you are checking 301 redirects to another page, but doesn't tell you if it has a canonical. OSE will also calculate redirected metrics into calculating scores such as MozRank and Domain Authority, and I believe there are plans (if they haven't been implemented already) of including canonical metrics into these calculations.

But back to you - what you really need is a way to make sure you've implemented your canonical tags correctly. You can check the duplicate title/duplicate content and the canonical reports in the SEOmoz PRO platform. You can also use the MozBar to check canonical tags, or a tool like URI valet to make sure your canonical tags are pointed to the right URL.

This is important because you want to avoid canonicalization errors that can lead to a loss of rankings and search engine visibility. That said, from your brief description, it sounds like you're doing everything right.

There are different names around this concept - search visibility, web visibility, keyword visibility etc. Ideally you want a tool that will quantify your website's "visibility" for a particular set of keywords or niche.

Any such tool or report would take not only keyword volume into consideration, but also weight based on rank. Because of what we know about click-through rates, positions 1-3 are much more valuable that position 25. In fact, SEOmoz doesn't rank keywords past position 50 because the return on investment for webmasters past that point is almost nil.

Google Analytics tries to provide some of this data in their SEO Reports, but the information doesn't include competitor data, and is an estimate at best.

The best "out-of-the-box" tool, to my knowledge, is the visibility score offered by Advanced Web Rankings.

Perhaps a better, and free, way to get this information is laid out by Ben Estes in this post about Defining Your True Competitors. Requires a bit of Excel wizardry, but it's an excellent method.

SEOmoz is constantly updating it's Rank Checking capabilities. If this is a feature you would like to see, I encourage you to fill out the survey located at the soon-to-be-replaced Rank Tracker, or submit a vote at the feature request forum.

Further Reading: Search Engine Visibility as a Metric

Hi Mark,

I used to troubleshoot these types of problems (mysteries when I worked on the SEOmoz help team.

The best thing to do would be contact the Help Team (help@seomoz.org) and include information both your account, url and campaign. They can take this information and see if there is anything odd about your website, or if there is a bug in the crawling software, or finally if there is some strange quirk of incompatibility causing this behavior.

If you would rather, you can PM me with the info and I can try to troubleshoot it myself, but the Help Team has a few more tools and access to engineers, so they might be the better choice. Either way, let us know if you have any trouble.

I really like Ryan's answer, and I'm going to endorse it.

If you are only looking for ONE metric to judge links, then DA or PA wouldn't be a bad choice. Certainly better than the old PageRank. But I wouldn't advise using just one metric.

DA takes all of SEOmoz link metrics into consideration, including MozTrust and MozRank. Of all the metrics out there, it is best indicator of a given domain's (or page's) ability to rank for a particular phrase. We know of no other publicly available measurement that does as good of job.

That said, it shouldn't be the only thing you should look at when evaluating link prospect. Relevance, Context, number of links on the page, position and other considerations should all be brought into play.

Best of luck in your link building!

Hi Kelly,

Thanks for letting us know. Could be a couple of things right off the bat. Is this your exact robots.txt file? If so, it's missing some formatting like proper spacing to be perfectly compliant. You can run a check of your robots.txt file at serveral places.

http://tool.motoricerca.info/robots-checker.phtml

http://www.searchenginepromotionhelp.com/m/robots-text-tester/robots-checker.php

http://www.sxw.org.uk/computing/robots/check.html

Also, it's generally a good idea to put specific inclusions towards the bottom, so I might flip the order and put the rogerbot directives last and the User-agent: * first.

Hope this helps. Let us know if any of this points in the right direction.

Unfortunately, this might possibly an attempt by a spammer at what is called a 302 hijacking. Google says 302 hijackings haven't worked in years, but many webmasters disagree wholeheartedly.

(on the other hand, it might not be this at all)

Basically, a spammer points a temporary redirect (check the redirection through a tool like URI Valet to see if it's a 302). Often if the site has a slightly higher authority, it can fool the search engines into hijacking your traffic. A deeper explanation here.

Another possibility is that your site has been hacked. Check for spyware, check Webmaster Tools and check to make sure that your domain is not the one being redirected. In Webmaster Tools make sure there hasn't been a change of address, and do a fetch as Googlebot to make sure your pages show up correctly.

Finally, you can fight back.

1. Contact the offenders hosting company and let them know
2. (a scary letter from a lawyer might help
3. File a spam report with Google
4. Post the issue on Webmaster Central

If it is an actual 302 hijacking, Google should take care of it fairly quickly.

Often, these aren't 302 hijackings at all, but are caused by internal site issues. Do a thorough site audit first before taking further steps.

First of all, great question. The problem is that anchor text optimization isn't tested nearly enough by search engine professionals, and that Google evolves quickly, so it's hard to know what the best practice is or what benefits can be gained by this type of tweaking.

To address your individual examples, there is conflicting evidence.

1. This is correct

2. This experiment showed that both anchors would count in this example. But this experiment showed the opposite as did this one - indicating Google would index both anchors

3. A similar situation with this example. My best guess is all anchors would count, but it's hard to say.

There are some nuances in your question worth exploring. We can't really say how Google will flow link juice through these 4 links. (my pure guess is that pure PageRank will be spread equal among the 2, but this is only speculation.) But more importantly, there are other factors that will determine the value of these links, including:

• Position on the page - Links higher up tend to pass more weight than links further down
• Navigation vs. Content links - Where links placed in the body of context carry more weight than boilerplate (repeated links)
• Context of the links. For example, a link contained within an advertising block may not carry very much weight.
• And more... This article by Rand lays it out pretty good.

These factors have the potential to far outweigh traditional PageRank sculpting techniques you may consider. It's a fascinating area of study, and I hope we can quantify it better in the future with more experimentation.

Good news, you can set up the On-page reports to grade any URL you desire.

By default, the On-page function generates reports for keywords ranking in the top 50 of your given search engine, for the URL showing in the search result. But sometimes this is not the URL you wish to grade.

For example, your "Contact Us" page ranks #5 for the keyword phrase "Yellow Shoes" and the Web App grades it a B. What if you really want to target your homepage for this keyword phase? Here’s how:

1. Hit Report Card at the top of the On-page summary.

2. Choose your Keyword you want to grade. The keyword must already be included in your   campaign. Select Manage Keywords if you need to add keywords.

3. Enter the URL of the page you would like to grade.

4. Hit Grade My On-Page Optimization to generate your report.

To stop a particular keyword/URL combination, you can hit "Stop Running Weekly" button on the report card page itself. It's also possible to grade multiple URLs for the same keyword.

Read more here....  http://seomoz.zendesk.com/entries/20034412-on-page-reports

OSE does allow you to export backlinks to a particular subfolder, but it can be a little tricky.

1. On the OSE homepage, enter the exact directory path you want OSE to explore. This is the screen you want to enter it on, not the advanced report screen. Hit search.

2. The directory needs to end in a backslash "/". Sometimes OSE will try to redirect you to a directory without a slash. For example, it tries to redirect "seomoz.org/blog/" to "seomoz.org/blog" (no slash). But it gives you the option to see your original entry. Always choose the entry with the backslash.

3. After this, on the Advanced reports tab, choose the "Links that Link to;" option that says Any page in this subfolder. Choose any other options you want on this screen.

Viola. You are done!

This should return links only pointing to the subdirectory you specified in step 1. If you have any trouble, feel free to contact customer support.

I endorsed Russ's answer because of his technical nuances.

Let me put it another way. To put it simply, the value of the link is can be diminished by the number of links on the source page. Unfortunately, there's no easy formula to calculate this. The truth is, different links have different editorial value based upon where they appear on the page.

Matt Cutts has famously said to put your most important links towards the top of your content, meaning within the body of text and not in sidebars or footers.

Other factors that play into determining the value of a link include the topic of the URL/Root Domain that the link is found on, mozTrust of the site, context of the link, and probably 100's of other factors. We don't know the exact formula, but if we look at the links like humans and ask ourselves "how valuable is this link?" we can usually get a pretty good estimation.

Both Ryan and Marie provided good answers, but let me elaborate further.

This is not a php thing. You can verify this yourself by visiting these pages.

Does the page content change when you visit /?lang=en&limit=5&limitstart=20 and switch to /?lang=en&limit=5&limitstart=20? If these look like the same page, then you have a duplicate content/title issue on your hands.

Google's take on the matter is simple...

"Provide one version of a URL to reach a document

To prevent users from linking to one version of a URL and others linking to a different version (this could split the reputation of that content between the URLs), focus on using and referring to one URL in the structure and internal linking of your pages. If you do find that people are accessing the same content through multiple URLs, setting up a 301 redirect from non-preferred URLs to the dominant URL is a good solution for this. You may also use canonical URL or use the rel="canonical" link element if you cannot redirect"

You have several options to deal with this, depending on the content_._

• Use the canonical tag to point the different versions of the page towards a single, definitive URL. Best to use this option only if the pages are actually duplicate.
• If the pages are not duplicate in content, but duplicate in title, then make sure your CMS writes unique title tags for every unique page of content, as Marie describes.
• If these extra pages are producing lists of only slightly varying entries, like a list of blog archives, you may consider adding a Meta Robots tag to NOINDEX, FOLLOW those superfluous pages.
• If dealing with pagination, here's an excellent article from Adam Audette to help.

This is a tricky area, and you must be careful when dealing with duplicate content, because it's easy to make a mistake and have Google de-index your content. That said, by correcting these errors, you just might see an improvement in your indexation and traffic stats.

This refers to the difference between Page Authority (PA) and Domain Authority (DA)

Where Domain Authority measures the predictive ranking strength of entire domains or subdomains, Page Authority measures the strength of individual page. The same is true for metrics such as MozRank and MozTrust.

It's completely normal to have a homepage PA that is higher or lower than your overall DA. If your homepage PA recently dropped while everything else remained constant, it could indicate that Linkscape didn't find some links pointing to your homepage that it found previously. Normally, this isn't anything to worry about, unless you've seen a corresponding drop in traffic. In this case, I'd check your backlinks.

It's also normal for both PA and DA to fluctuate from update to update. It's a comparative metric, so these metrics are best used when comparing yourself to your competition, instead of tracking your metrics in isolation.  For example, if your DA rises 5 points, this in itself means nothing unless you are comparing it to another domain.