SEOMOZ duplicate page result: True or false?

[AGIAutomation]

SEOMOZ say's: I have six (6) duplicate pages.

Duplicate content tool checker say's (0)

On the physical computer that hosts the website the page exists as one file. The casing of the file is irrelevant to the host machine, it wouldn't allow 2 files of the same name in the same directory.

To reenforce this point, you can access said file by camel-casing the URI in any fashion (eg; http://www.agi-automation.com/Pneumatic-grippers.htm). This does not bring up a different file each time, the server merely processes the URI as case-less and pulls the file by it's name.

What is happening in the example given is that some sort of indexer is being used to create a "dummy" reference of all the site files. Since the indexer doesn't have file access to the server, it does this by link crawling instead of reading files. It is the crawler that is making an assumption that the different casings of the pages are in fact different files. Perhaps there is a setting in the indexer to ignore casing.

So the indexer is thinking that these are 2 different pages when they really aren't. This makes all of the other points moot, though they would certainly be relevant in the case of an actual duplicated page."

****Page    Authority                                                                               Linking Root Domains

http://www.agi-automation.com/                                        43                                   82

http://www.agi-automation.com/index.html                    25                                     2

http://www.agi-automation.com/Linear-escapements.htm                                                                  21                                 1

www.agi-automation.com/linear-escapements.htm    16                                  1

http://www.agi-automation.com/Pneumatic-grippers.htm                                                                                                                                                         30                             3

http://www.agi-automation.com/pneumatic-grippers.htm     16                     1****

Duplicate content tool estimates the following:

• www and non-www header response;
• Google cache check;
• Similarity check;
• Default page check;
• 404 header response;
• PageRank dispersion check (i.e. if www and non-www versions have different PR).

[matbennett]

I always think that there are 2 questions to answer in cases like this:

1. Are the search engines seeing duplicate content?

2. Could the search engines see duplicate content?

The tools are useful for quickly highlighting potential problems, but you really want to roll your sleeves up and look for yourself.  I'll use teh Pneumatic Grippers page as an example:

The title of that page is :"Pneumatic, Grippers, Rotary Actuators, Linear Actuator, Robotics", so I'll do a search for:

intitle:"Pneumatic, Grippers, Rotary Actuators, Linear Actuator, Robotics"

That will bring up everything in google with that page title, just the one result - Good!  With regards to that page at least it seems google is only indexing one URL.

The URL that google has indexed is http://www.agi-automation.com/Pneumatic-grippers.htm - as you say, changing that case doesn't affect what page loads, so if an odd case (say http://www.agi-automation.com/PneuMAtic-grippers.htm ) could cause a problem.

What you need to prevent that is a rel=canonical in the source (I checked, you don't).  That tells the search engine what the correct address is.   Just ensure you have something like the following in your head section

http://www.agi-automation.com/Pneumatic-grippers.htm" />

There is another way to do it, redirecting the "wrong" version of URLs, but rel=canonical looks like the right choice for your site.

What I would say though, is this: If the search engines aren't picking up duplicate copies don't panic too much over this.  It would be good to have it, but it is only a big issue if duplicate pages are actually being indexed. If not it is good insurance.

I hope that helps