Welcome to the Q&A Forum

Hi JP Moulin, Thanks for your question.  Rand and The Back Burner already did a good job explaining some things about PA/DA so I'll stick to looking at the numbers. I looked at the top level metrics for these four sites in Open Site Explorer: Site                            PA           DA        # Linking Root domains         # Links www.nahoku.com      51           41               196                                    3695 www.mauidivers.com  53         43                228                                    4274 www.jared.com          59          51                 283                                     871 www.hawaiianjewelryshop.com  60  52        325                                   5890 These look pretty well ordered based on these numbers to me (more links + linking domains => higher PA/DA) with the possible exception of www.jared.com which is a bit of an outlier with 283 linking root domains and only 871 links.  In this case, the domain links are boosting it's PA/DA.  The new model prefers metrics associated with link diversity/domain signals (linking root domains, linking C-blocks, etc) over the old one and this is what's happening with www.jared.com.

Often if there's a glitch just doing a hard refresh will work. If you encounter a problem again with things looking funny, take a screenshot and send an email to help@seomoz.org. The help team doesn't actually monitor the Q&A forum unless an associate specifically flags a question, so emailing the team will help you get a response from Moz faster, and give us some information on what may be misbehaving. Thanks!

Hi Miranda, If you download the CSV of your errors, you'll see a column in the export labeled refer and you'll be able to see where these links are coming from.

Thanks anyway but no this client has never before done any structured SEO or link building. The site is a long established site with thousands of clients linking back to it and the checks I've done show that they are natural links. I'll do a bit more digging but my previous observations had been that GWT underestimated the links versus OSE or previously Yahoo Site explorer. This is so far the other way around it is odd.

Hey! That sounds like odd behavior and I don't think I've heard of that happening before. I'd love to dig a bit deeper to see what's going on. Would you be able to send me the pages you are searching? I assume you are experiencing this in Open Site Explorer? If you would prefer not post the URLs in this forum, feel free to email me directly at carin@seomoz.org! Thanks, Carin

Can you give an example where SEOmoz and OSE are different? I'm not quite sure I follow you, since Open Site Explorer is an SEOmoz product.

Both of those URL's you provided return a 404 error. However, when I do click "English", the page loads up.  And if I visit this URL (which I created myself), the product comes up, too: http://shop.samson.de/deutsch/products/hardwarepaket.html?___from_store=english It seems you have some rewrite issues as you already mentioned.

Sha has a good response here. You might also be interested in reading Dr. Pete's post about "How Many Links is Too Many" on the SEOmoz blog earlier this year at http://www.seomoz.org/blog/how-many-links-is-too-many for some more background information.

Hey Mark, Our apologies that the results weren't as you expected. Unfortunately, this is something of an intractable problem. The search engines - Google, Yahoo! & Bing - all maintain multiple indices that run across multiple datacenters. This means if you query different datacenters, you can get somewhat different ranking results. Other elements, such as personalization, geography and search history (even if you're logged out of your Google/search engine account) can also influence ranking positions. Hence, there can be a lot of variation in what different people, searching from different computers/locations might see in the rankings. Our solution has been to run searches from a variety of IP addresses and IP blocks using non-personalized, search history/location agnostic requests. In our experience, these have provided the most accurate results, showing what the "most" users see, though we know there's still quite a bit of room for fluctuation. Thanks again, and have a great day! Nick

Although I don't have an explanation as to why your SEOmoz rankings vary greatly every week, I can offer some suggestions/ideas to get over this little issue. Have you tried e-mailing the SEOmoz customer service team (help@seomoz.org)?  They might be able to look into your specific case and diagnose any technical problems. In the mean time, have you tried creating a completely new campaign to see if the same thing happens? What about using the rank tracker that is NOT part of the PRO web app?  It has weekly updates and if the same problem occurs you can try manually refreshing the rank once every day. Lastly, a little more legwork here but you could also try running your rank checker about the same time SEOmoz updates their rankings in the web app/rank checker tool.  Maybe you might see a difference there? Best of luck!

Good, glad to have helped

Yes, I did misinterpret when to properly use rel canonical. Google seemed to respond the way I was expecting though. All of my new pages went into the index while the old pages were ignored. Once a 301 is in place, is it safe to remove the old page from the server?

Hey Lawrence, The changes made to the PA model do not treat 301 redirects any different - we haven't made any updates to how we handle redirects. However, if you've seen a drop in your links, we probably didn't crawl the pages redirecting to you. This would definitely affect your PA score. I hope this helps, but let me know if you have more questions! If you want to find out more information on the new PA model, Matt Peters, the engineer who worked on the new model, has a great blog post here: http://www.seomoz.org/blog/introducing-seomoz-updated-page-authority-and-domain-authority Thanks, Carin

I've endorsed Ryan's answer but just a quick note about PageRank. We encourage folks not to use PageRank as an indicator of ranking strenght, as "PageRank has a relatively low correlation to how well things actually rank." - Rand Fishkin PageRank is good for a couple of things, such as a raw indication of link popularity, or as Ryan mentioned to determine a penalty. PageRank data is historically inaccurate, Google doesn't update it regularly enough to be useful and finally Google has asked us politely not to scrape PageRank data, and SEOmoz honors that request. So those are a few of the reasons we don't include it in the MozBar (SEO Toolbar) and other places. Rand discusses the subject in depth in this WBF http://www.seomoz.org/blog/what-is-googles-pagerank-good-for-whiteboard-friday

SEOMoz's offerings is SEO software, not consulting. If you need consulting our advice try and being more descriptive in your Q&A posts or visit the marketplace in regards to consultants for hire: http://www.seomoz.org/marketplace Good luck! - Kyle

Here you go buddy. Short answer is yes, but it will take 24 hours... http://seomoz.zendesk.com/entries/20034082-crawl-diagnostics

Its just a notive to let you know they are there, Rel canonicals are somthing you would nort want in the wrong page, so letting you know where they are is a good thing

Wow, what a great answer. Thanks bunch, actually helped alot