Welcome to the Q&A Forum

Think of domain authority like you would the stock or bond market. They fluctuate over time. Sometimes an individual stock or bond's value will go up or down as a result of that specific stock's assessed risk / benefit. Sometimes the whole market will shift based on factors that have broader influence. Stock and bond valuations fluctuate real time. Website valuations (as measured / quantified by Moz) fluctuate roughly monthly. You've observed what equates to a "market shift" likely resulting from the size and characterization of the websites in Moz's sample. Stock and bond valuations fluctuate real time. Website valuations (as measured / quantified by Moz) are derived from a large and lengthy process that involves going out and collecting, analyzing, and interpreting data from millions of websites, billions of URLs, and trillions of links. The numbers that spit out at the end vary based on a whole bunch of factors, not least of which is the size and nature of the targets of evaluation. You've observed what roughly equates to a "market shift". It's likely not a coincidence and due to the sampling process.

If you have no links pointing to a page then it has no PR at all. It will also have no PR unless it has a outgoing link, see the original algorithm. yes put the internal links in see http://thatsit.com.au/seo/tutorials/a-simple-explanation-of-pagerank

Mat is correct, to put it short. A weak site about dogs, may beat a strong site about cats for the search term "dogs"

Michael is correct as it pertains to multiple links going to different pages, but in terms of multiple links going to the same page I'm pretty sure only the first one will count. I think this is true whether the links are internal or external. See references below... http://moz.com/blog/results-of-google-experimentation-only-the-first-anchor-text-counts Moz Experiment Results by Rand http://www.seo-scientist.com/first-link-counted-rebunked.html "OK, so to me this is a pretty solid data supporting Michael’s and Rand’s claims. The fact that the SERPs reacted to my changes in the order of links back and forth is kind of hard to debunk." http://www.addictedtoseo.com/exceptions-from-the-first-link-counts-rule.html I like this post because, like with the SEO Scientist test above, it highlights the fact that nofollowing will not help. This is why internal pagerank sculpting with nofollow links doesn't work. I'm not sure if it still works, but you used to be able to get around this by linking with hash tags that jump to named anchors on the same page. So www.mysite.com/url1 and www.mysite.com/url1#top would be considered two links and would both pass some pagerank.

There's a good chance of that - a drop like that would normally mean its PageRank has been penalised. If you can verify this with some suspicious looking backlinks via Open Site Explorer, then I would steer clear.

Hey Bond, Thanks for writing in with a great question. Both the campaign data and the data in OSE are pulled from the same Mozscape index, but OSE defaults to page metrics, so you will want to make sure that you are comparing domain metrics to domain metrics between the campaign and OSE. When I checked the domain metrics between your campaign and OSE, the data did match up correctly. I may be able to explain the comparison better through screenshots, but I wasn't sure if you wanted to share your campaign information publicly. The only thing I am seeing that doesn't show data correctly is actually the second competitor in your campaign because it is set up as ww.competitor2.co.uk, rather than www.competitor2.co.uk, which skews the subdomain metrics in the campaign link analysis. If you update the subdomain in your campaign settings, you should start getting better data for that competitor in the rankings and link analysis. If you would like me to send you the screenshots to explain how the campaign compares to OSE privately, please let me know and I will email that to you right away. I hope this helps! Chiaryn

Thanks for the reply, Yeah I actually think that personal training example was a really good one. I think your right as well, just not sure if it would work for ecommerce, services I do though.

We saw the May 5 spike across multiple data centers, but we don't track a ton, so it's always tough to tell (our larger system was down for testing that day - it's not public yet). I feel like this is more than a test, though, and some relatively big changes are coming.

Hi there, As Eric mentioned before, the solution will depend on how much unique content is there in the paginated pages (from the main category page): If there is very very little unique content, crawling and indexing them won't really help on earning more search visibility with them (usually in these cases would be additional long-tail type of keywords for that product category) but just to consume the crawlers time and effort. Being this the case then the best way would be to canonicalize the paginated pages towards each one of their appropriate main product category page. On the other hand, if the possibility exist to differentiate them: By featuring additional pages descriptions for each one of the paginated category pages or users reviews or ratings, or product descriptions of enough length, that can serve to give additional relevance value, then the best way to go would be to implement the rel next & prev annotations.

Things that will make these links look more suspicious include exact match anchor text, highly unrelated links (eg poker, viagra, attorney links on a gardening website), lots of low-quality websites that have other spammy indicators being linked to. If it's a bunch of websites that have generally high quality metrics, using branded anchor text for the links, then you're probably fine. That will just look like a normal blogroll.

Is absolutely normal Google bounce around a little bit especially in higher numbers. If you are anywhere in and above five you should expect fluctuation and it's the algorithm is so so complex that if you updated your site one day Google bot might not indexed it or it could've indexed it and seeing something that didn't like. Remember your competitors for those keywords are also being judged. So it's trying to figure out exactly where you fitting. And I wouldn't worry too much about it. Just keep holding quality content and you'll be alright the one thing that you can do and it's the right thing to do so I hope that answers your question and expect fluctuation from  Google from time to time. When Google gives you a break that you stay at a rank  for a while I think it is uncommon for somebody that is not watching to actually see the fluctuations because when you login to your Google account Gmail or something like that you see a different ranking than what I would see maybe you'll see a higher ranking and I will because you're searching your on-site more often than I am us Google will tailor the results to you if you are working. When you go on SEOmoz   and see the real Rankins then it can be frustrating. However fluctuation is normal part of the game and it's something to be expected rankings also highly dependent on geographic placement  however if you keep creating what Google wants witches he could user experience or the best they will keep cranking you hire and all you may fluctuate always feel fluctuate less and less and hopefully be in first place. I hope I've been of help sincerely, Thomas

Tom has the right of it. I would consider doing a few different things in addition to what he's talking about. Make sure you link from one to the other with something like "Check this out...." like Tom said.If it's an e-book or case study or something I would also consider either getting e-mails or using something like pay-with-a-tweet to make sure the person who comes to visit that content is going to get repeated exposure to your brand.

Yep, diversify you anchor and pray for the best.

There's this Firefox Google Semantics tool add-on -- http://tinyurl.com/buybxq8 as well as this Google Synonym search tool -- http://www.synonymlab.com/index.php

Depending on the amount of pages you have on your website depends on the way in which you should look at generating your meta descriptions. If possible and ideally every page should have a completely unique page title and description. Obviously, if your website has thousands of pages and products then you might find that an auto generated title and description is necessary for the "parent" product and then a variation of for the "child" products as you stated above, this should not really be used for a search engine ranking however if they are just product variants you would only SEO the lead or parent product and have the variants removed from the search engine crawls to prevent duplication. This is how all major e-commerce platforms work and why Magento and even the likes of Amazon operate to avoid duplication of data. Basically they create a master / lead / parent product and "no follow" all of the child products. Hope this helps.

Cheers Tom

Tyr this post too... How to add an image to google images

Nice cheers Irving, I can see how LSI could open the doors to other variations showing up in the SE's. Do you think a search engine such as Google makes the connection between terms such as 'PC Screen' and 'PC monitor' and do you think this is a factor in rankings? For example, if a page was targeting the term 'PC Screen' but the page included the word 'Montior' a few times do you think that word would benefit the term 'PC Screen' in relevance/rankings. Thanks for your time, it's something im trying understand more.

Nice come back. Don't really have anything else to say, your argument won me over on the first read. Thanks.

Well, while we're technically classified as "betting" by UK law, we're very much seen as a financial service provider by the public - and if Google's LSI and word association is anything to go by, we're nearer to trading than we are betting - so we've been able to avoid some of the pitfalls that some betting/gambling companies have when outreaching, for example.