John Mueller of Google recently said that PDF links were now being treated as ‘no-follow’. James thought that this should be investigated, to see the truthfulness of the statement.
Transcription Below
Do PDF links pass juice? This is interesting. John Mueller came out and he’s saying that pdfs are going to be nofollow. If you follow the rest of the news nofollow is a hint. It is not a directive anymore. Basically, before it was dofollow or nofollow and you want to get dofollow links because nofollow ~. The thinking was it didn't do any good.
I disagree with that completely. It was based off of the URL, the domain, the website was on and you can get nofollows from good domains and still pass juice.
DORI: Yes and we have pages with deindexed sites with nofollow link.
CLINT: And they were passing juice. So, they changed the rules or the public rules and it was nofollow are now hints so essentially dofollow. Then, it had sponsored which we tested and sponsored don't pass ranking power. Then, user-generated content is the other one and I'm not sure that we tested that one.
I think we did but I don't have to go back and look but I would anticipate that user-generated content would improve your ranking but not as much as a nofollow or a dofollow.
With that being said, James wanted to go through and test it. What John said is that there's nofollow, that all PDF links are nofollow now. In his results, you see he added the PDF test and he went through here and created it then linked using the PDF to his test page right in the middle. If you go down here, bounced down, bounced down, bounced back up, bounced down, bounce back. It's consistent and he waited a long enough time to see the power from it.
Here, it actually dropped the rankings. The question that still has to be asked is if nofollow passes juice based on the authority of the site where the PDF is at, and would that make a difference. Where was this PDF hosted, is the question. I think if James goes – he's really tedious so he was doing the linking from or linking to sites ~ can see if it was actually being valuable I bet James will be all over testing which PDF and document sharing platforms actually Google trust enough to give you ranking juice through that nofollow.
It may be that the one that James tested this one on, hosting that PDF is if it's not a no-fly zone but Google's like, yes, whatever. People who link there are from spammers versus if he found a different source maybe even a paid Google Apps Google Drive and host the PDF in there would improve it because it's a paid known entity, now I kind of "trust you". That would be really interesting.
At the end of the day, James proved that John was actually telling the truth in this case that yes, PDF links are nofollow; yes, they can hurt you. Now that he's onto something I encouraged him to stretch it out a little bit and check the other PDF sharing places ~.
DORI: This could also. If I was putting pdfs on the toxic sites and linking out from PDF ~ sites that I already know that are toxic.
CLINT: There you go. Then, you couple that with the web 2.0s tests that we got to redo for sure. These ones are good for you, these ones are bad for you. If you can recreate those lists you can probably chop down your link building to maybe 20-30 sites and then use your spam tools like your GSA's or whatever on those 20-30 sites and rank all day long, with some PBNs sprinkled in there too. Do a whole lot less work.
Hopefully, James does but if he doesn't I may have to go and do it just because it's so enlightening, not thinking about that for so long. I just have my old list of places where to go and then ~ test it to see if one would hurt me or not. I think that's something that I need to do.
DORI: Yes, it was like when I started testing the PBNs I thought a good link is a link. It ~ works more valuable than others but I didn't really know how many would be pushing sites down that were actually toxic. It was really enlightening.
CLINT: Which begs the question ~ link sellers granny buddy for that matter if you're not testing your links then how can you – let's say, you use Dori's Network (I'm just going to pick on you because we're talking about you) if you use Dori's network and you don't go anywhere, you don't go up or you don't go down, you bought ten links who is to say you didn't buy five good ones and bad ones if Dori is not testing.
CLINT: That's what I was thinking and being in the SIA now for five years it's like, "Okay. We need to test these." Actually, I have to give it to Matt Diggity too. A couple years ago he said. But he had another way of testing PBNs but it was still like, "Okay. I'm building a new network. I definitely have to test the sites I'm putting in the network."
CLINT: On his thing though. For you, I would anticipate that you would be retesting those continuously to see the patterns versus the way Diggity was teaching it. It was, "I was testing it before I put it in the network." There is no mention of "Three months later I tested it again and now it's a toxic link."
DORI: Yes, we continually do that, at least. The test sites will be up. Then, I'm going to be testing because we're shooting anchor text to them. Is it bad? Can you shoot a URL? There are different variables that I'm going to go forward and test with too, with the sites that are, say, toxic and the ones that are good. Like, good different tests.