Sphinn - Using NoFollow as an SEO Tool

21
Sphinn It!

Using NoFollow as an SEO Tool

Went Hot: September 3, 2007 - 6:13 am
Posted By: WaltVerastegui 409 days ago
Topic Type: News Story (Jump to http://blog.searchenginewatch.com)
Category: Google SEO

Now this is a good idea: using NoFollow's as a way to control PageRank flow.

6 Comments

Who Desphunn This (0)

Comments

from LocalHound 407 days ago #

Votes: 0 | Vote:

+ -

Dan Thies has a free ebook at seofaststart.com which provides some good information about this.

from TheMadHat 405 days ago #

Votes: 0 | Vote:

+ -

Nofollow can't be used to control PageRank flow. Nofollow doesn't block the page from being indexed. The only benefit from using nofollow internally on your links is to speed up the indexing rate. PageRank is a probability distribution calculation and the probability of a user landing on a page you have nofollowed via a link remains the same. I don't know if Matt said that by accident or if it was an attempt to cast nofollow in a good light but it's not correct.

from dannysullivan 405 days ago #

Votes: 2 | Vote:

+ -

It can be used to control the flow. I mean, if you had any doubts, I was in the room with about 20 other SEOs when Rand asked Matt about this, and Matt talked about how YouTube (owned by Google) specifically uses it to control PageRank flow.

In particular, the original PageRank formula divided the amount of PageRank that could flow off a page to other pages across the number of links on that page. IE, the more links a page had, the less PageRank each got. Sort of like a big PageRank pie that you have to cut 25 ways versus 5 ways.

We don't know the current formula, and Matt could be all Mr. Misleadingly, but I think many still believe that the fewer links on a page, the more PageRank share each link gets. Nofollow means those links tagged as such don't get any PageRank share. It also and seperately means that Google shouldn't index them. But as you say, Google might find a way to those links in another manner -- that's why they may still get indexed. But there's NO way Google will find another way to credit those links with the link love from a particular page, if that page has blocked love from flowing with nofollow.

from TheMadHat 405 days ago #

Votes: 0 | Vote:

+ -

"But there's NO way Google will find another way to credit those links with the link love from a particular page, if that page has blocked love from flowing with nofollow."

I certainly agree with you on this point. However the way I understand the original iteration of PageRank is based on the probability of a user randomly finding the page in question. Based on that let's look at this scenario. Page A and B both have 5 outbound links. Page A has all five links followed and page B has four nofollows and one followed. The probability of a user finding each page remains exactly the same as long as all five pages are indexed. I can see flowing PageRank by blocking the four pages with robots.txt or meta tags because the probability of hitting the page in question will increase without the presence of the other links.

Obviously on top of that there will be a PR distribution algorithm that will determine what "share" of PR each page gets but I see it as a small subset of a percentage in the overall score. I can see the possibility of this adding some kind of benefit on an extremely large scale (IE, YouTube with 15 billion pages) but the amount of PR gained from this method on anything less than a million pages will be so negligible it might as well be zero.

I think I may of misspoke when I said it "can not" be used to control PR flow, and should have said it's a useless exercise for 99% of websites.

I'm sure Matt said it without trying to be misleading but I think it's been interpreted incorrectly. I could be completely off base but that was my initial reaction when I thought about it. It rings along the lines of "PR hoarding" or whatever it was called when webmasters were afraid to link out for fear of losing PR.

from dannysullivan 405 days ago #

Votes: 2 | Vote:

+ -

Original PageRank is interesting, but I can't stress enough how relying on a document from 1998 to talk about Google nine years later is dangerous. I agree that it's likely many of the basic principles are in play, but there's also sooooo much that has changed.

But to go back to it, I think probability was just part of how PageRank was assigned. Number of links on a page was also another factor. So fewer links, more link juice to flow out. Remember, as far as Google is concerned when doing PageRank calculations, if there are 5 links and 4 with nofollow, it only sees 1 link -- as best as anyone can guess.

And yeah, it does ring of PR hoarding and trying to control the flow in ways that Google's sort of suggested you shouldn't need to do. I asked Matt about that, because it seemed so odd. And he was pretty much saying yep, it's OK to do, and in fact you might want to do it to help Google understand what pages deserve the most credit. It hasn't be interpreted incorrectly at all, in my opinion.

from MattC 405 days ago #

Votes: 0 | Vote:

+ -

I wonder how many of us use the nofollow to control PR flow. I never really got into myself because many of the sites I work on have decent site structure and aren't overblown with useless pages (they might not be the best but i dont want to ignore them). There are situations I suppose where I could have implemented it, but that technique just seems like one of those things that will do more harm if you get carried away it.