Nofollowing internal links

I have just been reading an interesting article on Search Engine Land about nofollowing internal links. The author, Shari Thurow, argues that web developers should not be nofollowing links to low-priority pages (terms + conditions, privacy policy etc). Shari points out that it's odd to put a nofollow on links that search engines won't crawl, that these low-value pages are stripped from search results anyway, and that giving spiders different links to users is a bad idea. Instead of nofollowing pages, a better idea is to revamp the link architecture of the site.

Personally, I don't see the problem. While spiders might index a site not micromanaging their nofollows just fine, Google is hardly going to start penalizing webmasters for using nofollow excessively (that would be a bit rich).

I like to nofollow certain internal pages for the following reasons...

Multiple links to the same page

When you have a number of links all pointing to the same plase, I like to nofollow the links with the worst anchor text. Typically, you might have a product listing with a clickable title, image, and "more info" button. I would nofollow the image and the button, because they have poor anchor text. If Google is going to use anchor text as a ranking factor, let it use the rich anchor text from the best link we have (ie the product name / title).

Low-value or machine-generated pages

I'm talking member profile pages, product image popups, and pagination links. These pages are usually database generated and fairly similar looking. I'll often nofollow these because I don't want to pour precious link juice at them - but I'm happy to have them show up in search results if someone else wants to link to them directly (which is why I don't use a noindex tag). In this respect, I view nofollows as more a way of prioritizing than blocking - I'm effectively saying "Google, you can index this page if you want to, but don't use my link juice to do it".

Truly worthless pages

I view terms and conditions pages and the like as a truly useless entry point to a site - but they can be important for users who are already on the site.
I'll usually use a noindex tag on these pages, because I just don't want them in search results. And if I don't want them indexed, I don't see the harm in nofollowing the links pointing to them. Why send the Google spider out to look at a page when you know it's going to find a dirty big noindex tag on there? I'd rather the Googlebot used that extra click to view some of my good content instead. True, if you add the page to your robots.txt (which I also often do) then Googlebot won't even visit the page, so consider the nofollow to be a backup measure in this case.

Useless, at worst

If you accept the concept that Googlebot is only going to spend a certain amount of time on your site / or only look at a certain number of pages, then it makes sense to give the Googlebot the best instructions possible - and that means nofollowing links to pages you don't want it to find.

I compare the practice of micromanaging your nofollows similar to measuring the ingredients in a Jamie Oliver recipe. Most of the time, you can get away with estimating the quantities, and the recipe still comes out great. But being pedantic about measuring the quantities is not a bad thing, and at worst it gives you the same result, but takes a little longer.

Likewise about nofollowing the links - you can probably get away without them if you have good link authority, but I think you need every little advantage you can get. And hacking a template with some nofollows isn't a difficult job, so I'm saying it's still an important change to make to most websites.
Digg StumbleUpon del.icio.us technorati blinklist furl reddit sphinn

Tags: nofollowseonofollow seo