16 replies to this topic

[BlueHorseradish]

For one of my clients, an obscure material data sheet PDF now has about 1,000 inbound links according to Google Webmaster Tools.
When I look up those links, it says 55 domains link to that PDF with 1,000 total links.
For each of these links, it says:

Via this intermediate link: http://www.abccompan...asheets/xyz.pdf
(Obviously, I substituted abccompany.com for the domain name of maker of this material.)
I've never seen "Via this intermediate link" in GWT before.

We also have that PDF on our site with a different filename, but none of the sites listed by GWT actually link to our site; they link to abc company's PDF. And the PDF on abc company's website does not redirect to our website. That PDF shows up on abc company's website with no redirect.

OpenSiteExplorer finds no links to our PDF.

Baffling!

Any thoughts on what GWT's "Via this intermediate link:" means?

Some options I'm considering:

• Change the filename of our PDF.
• Link to the original PDF on the abc company website instead of using our local copy of the PDF
• I could block that file or folder in robots.txt
• I could do nothing and hope it means nothing or goes away

Any insights would be appreciated.

Edited by BlueHorseradish, 15 June 2012 - 12:21 AM.

[haggisv]

Perhaps your client can find a way of extracting value out of all the links pointing to this PDF...

[Jill]

It's usually a redirect.

[BlueHorseradish]

As mentioned, the PDF on abc company's website does not redirect to our website.
The PDF is a data sheet on their own material, so they wouldn't ever redirect to a customer's site. I didn't check every link but spot checked a few dozen of them and none redirect.

[Michael Martinez]

Has the client received an "Unnatural Links" warning? And, if so, why do links pointing at a .PDF worry you? There are a fair number of Websites out there that try to index PDF files for people who want to find PDF files. Those links may be legitimate, non-harmful links.

[BlueHorseradish]

Has the client received an "Unnatural Links" warning? And, if so, why do links pointing at a .PDF worry you? There are a fair number of Websites out there that try to index PDF files for people who want to find PDF files. Those links may be legitimate, non-harmful links.

No, no unnatural links warning. GWT is completely clean.
I'm worried because all of a sudden there are about 1000 inbound links showing in GWT, and there are really no links like this that exist. They aren't legitimate links at all...there are no links and no re-directs. Baffling.

Any thoughts on the options I'm considering?

• Change the filename of our PDF.
• Link to the original PDF on the abc company website instead of using our local copy of the PDF
• I could block that file or folder in robots.txt
• I could do nothing and hope it means nothing or goes away

[Jill]

I would ignore it. GWMT always shows weird things that have no real meaning in the scheme of thinsgs.

[BlueHorseradish]

I posed this question in the Google Webmaster Central.
This is the link to the question:
http://productforums...lo/-64NqAm6sygJ

This was a reply, which I think makes sense:
++++++++++++++++
As the other file is an exact copy of your file, it looks like Google has associated them as duplicates.

Those listed URLs point to the other copy of the file. Maybe Google is treating your copy as the canonical version and hence those links are treated as pointing to your site "via" where they really link.

I have seen this before, perhaps sometime in late 2010.

Page A points to file B.
File B and file C are identical.
File C is the canonical version.
Links pointing to file B are treated as pointing to file C "via" file B.
++++++++++++++++
So, if that's the case, maybe it's okay to ignore, maybe it requires some action.

What would you do?

• Change the filename of our PDF.
• Link to the original PDF on the abc company website instead of using our local copy of the PDF
• Block that file or folder in robots.txt
• Do nothing and hope it means nothing or goes away

Edited by BlueHorseradish, 01 July 2012 - 09:18 PM.

[chrishirst]

I have a question.

Why does it matter??

[Jill]

Ignore it since it is obviously not a problem.

[BlueHorseradish]

It matters because I'm trying to figure out a moderate drop in traffic and Google rankings for my client. The traffic drop happened around the beginning of April. As I understand it, Penguin hit around April 24. I have a clean link profile and Google Webmaster Tools shows no other problems except for this. I'm a white hat SEO consultant so we're not doing all the stuff that caused problems for many others. This issue caused over 1000 new links that don't really exist, so this about the only thing I can find that might cause a Google penalty; maybe this is a duplicate content problem or an "unnatural links" issue. (No, we didn't get a warning.) Anyway, something seemed to happen and this might be it.

Anyway, I decided to change the link to the PDF that we did have on our website, and link directly to the supplier's data sheet, which is the exact same PDF. We'll see.

Thanks for the feedback!

[Jill]

The drop in traffic wouldn't be related to that.

Google Webmaster Tools shows no other problems except for this.

Do they show it as a problem?

[BlueHorseradish]

The drop in traffic wouldn't be related to that.

Do they show it as a problem?

Jill, thanks for your reply. I'd like to believe that these things are not related, but really, we don't know.
Not sure how it would be designated as a problem, but I don't see that indicated anywhere in GWT.
I would think a spike of over 1000 links in a short period of time would look unnatural. I just check Bing Webmaster Tools and BWT specifically shows a spike of 1000 links right at the beginning of April too.

Another thought - could this be a duplicate content problem as this PDF is a technical data sheet from a supplier, so it's exactly the same PDF as they have.

Thanks for the feedback!

[chrishirst]

I would think a spike of over 1000 links in a short period of time would look unnatural

Not really, a blog post by Matt Cutts probably has a thousand links pointing to it before he even has time to log out.

But that doesn't mean that Google will find them all in a similar time frame. You need to have a sense of scale, Google deal with tens of millions of links every hour of the day. To find a few hundred new links pointing to a URL probably isn't going to make a blip on the Google radar.

HOWEVER

[chrishirst]

Oh and of course there is the anchor text to take into consideration. If they all use the same wording, even if it is the URI is going to look a little dubious and may mean all of them or some of them are simply ignored for passing any weighting value.