how far can relevancy be passed?

[KevinC]

We all know that a link with the anchor "widgets" will help make the linked to page relevant for this exact keyword.

But how far can this relevancy be passed? For example:

www.siteA.com/index.html

links to

www.siteB.com/index.html
with the anchor "widgets"

and www.siteB.com/index.html links to all other pages associated with this domain will it pass on this relevancy to the other pages - even though the other pages may not be linked to with the same anchor text?

Think of a general topic directory setup - If I link to a directory with the anchor "widgets" chances are the home page of the directory is not about "widgets" but will a sub category get any benefit?

hope this makes sense - I'd love to hear some peoples thoughts on this.

[Mohamed_E]

As far as I can tell there is no general "inheritance" in Google. That is a great pity, as inheritance would simplify life a lot.

I have an index page optimized for "widgets in location". Under it I have pages on the traditional "red widgets", "blue widgets", etc. They will not match "red widgets in location" etc unless I explicitly mention location on each and every page.

[caine]

Inheritance does exist with all search engine's, its inherent in the algo's. The question that i feel you may be asking is the weight of anchor text links starting at point x then consistently being diversified out across ever larger networks of site linking.

Obviously the strongest point is at the x-1 stage with a constant weakening as it hops down.

[KevinC]

Maybe a simpler way of putting it is if:

my index page is relevant for "widgets" and I link to a sub page using the anchor "click here" will this sub page inherit the relevancy of the terms "widgets" - even though the term was not used in the anchor text?

[nakulgoyal]

I vote for caine! That's the best way and I do the same here for my clients. and everything seems to work well. So you can say that I have tested results for caine way!