Does Anchor Text Count From No-Follow Links in Google's Algorithm?

[beerhead]

Hey everyone,

Have the following question: is it worth using nofollow links for anchor variety? I usually use commercial anchors with dofollow links, and to dillute the exact anchors ratio I throw in naked or branded URLs which are pretty often nofollow. Does it make any sense considering G doesn't take into account nofollow anchor text but still sees the link?

 

[Future State]

Have the following question: is it worth using nofollow links for anchor variety? [...] Does it make any sense considering G doesn't take into account nofollow anchor text but still sees the link?

I disagree with the statement that Google doesn't take into account nofollow anchor text. When it comes to a link, there are a ton of factors that affect ranking. To name only a few, we could list page rank flow, anchor text, and relevancy. The only thing that nofollow does is restrict the flow of page rank.

The bots still crawl through the link, they'll still index pages that are linked to exclusively with nofollow, they use the content on the linking page to help understand what the destination page is about, and they use the anchor text to do the same plus to help determine where it deserves to rank.

Yes, it's worth using nofollow links for anchor variety in my opinion. It's part of what happens naturally on the net. People post links to your site in blog comments, forums, social media, user submitted editorial posts, etc. All of those are nofollow but are still used to understand your site.

Nofollow links are undoubtedly a part of the link graph and it's possible 100+ ranking factors. The nofollow tag tells just one of those ranking factors (albeit a very important one) to not be in play.

 

[turbin3]

Hey everyone,

Have the following question: is it worth using nofollow links for anchor variety? I usually use commercial anchors with dofollow links, and to dillute the exact anchors ratio I throw in naked or branded URLs which are pretty often nofollow. Does it make any sense considering G doesn't take into account nofollow anchor text but still sees the link?

Yes, include some nofollow links. Generally speaking, it's a good idea to have at least some form of variety in:

• Link type
  • Text, image, redirect
• URLs
  • Case-sensitivity: example.com/Stuff vs. example.com/stuff
    
  • Canonical status:

    http://www.example.com/  vs.  http://example.com/

• Link status
  • Follow, nofollow
• Anchor text
  • Brand: Just the brand name, and no money keywords. This gets tricky if your "brand" is an exact or partial match domain.
    
  • URL: (aka "naked" or several other nicknames) purely the URL and nothing else.
  • Money: Your primary money keywords.
  • Compound: Brand + Money anchor.

I'm simplifying of course. There are many other little variables, and even each of those categories above could be segmented into many variations. The thing that's important to keep in mind is, just make sure you have a bit of variety in some of these areas.

Link Distributions
One thing I didn't mention yet, is distribution with any of those factors. By now, with the advanced nature of search algorithms, worrying about percentage points of distribution is a waste of time.

Also, much of the conventional advice people have on the subject, usually ends up being broad recommendations for everyone. Stuff like, "You should have these percentages for these items."

The thing is, the way search works now, particularly with Google, is that everything is varied down to the niche level. Sometimes it even varies down to individual keywords. This is the machine learning age.

For example, in certain niches the average anchor text distribution among leading competitors might be as high as 90% exact match, short tail, money, etc. keywords. Conventional wisdom will say that's a guaranteed penalty, yet there are many multi-million (if not billion) dollar industries where similar trends are seen.

The point I'm getting at is, I wouldn't worry too much about the fine detail, trying to hit specific percentages of distribution in some characteristic of your links, or anything along those lines. In fact, even some of the factors and link types I mentioned above (case sensitivity, canonical vs. non-canonical, etc.) are a bit extraneous. For small sites and relatively small link profiles, some of those things may not even matter as much, if at all.

 

[bernard]

It has always been that links sending traffic were the best links for SEO, no-follow or do-follow.