Ok, this is something that I am still am not clear on. I know that PageRank is much less important in determining SERP rankings than it used to be, and I only use PR as a sort of quick estimator tool. But I'm still not that clear on the degree of correlation between the PR of the page a link is on, and the effect it will have on your SERPS. ~~~~~~~~ - True or false. A link that will provide a significant boost for your PR could do relatively little for your SERPs -- even if the anchor text was good. A backlink that was PR 0 from a subpage in a big authority/hub site could still provide a big boost, even though it might not help your PR a single bit, correct? If you were to get a link from a barely indexed, deep page on sites like about.com ,business.com, etc, just the fact that the link is coming from one of these big sites could still be a big boon -- possibly even more than say, a link from a PR4-6 page on some unimportant, unrelated directory, correct? ~~~~~~~~~~~ - True or false. Having links from many sites helps, even if each individual link is worth very little. So 5000 forum signature links from 1 site might not help your SERPs that much, but having 50 links from 100 different fairly large forums could, due to the fact that even though each link has little/no value, the SE still sees 100 different websites linking to you??? (No, I would never spend my time posting at 100 forums.... I'm just using it as an example. Although, I might pay a 14 year old internet geek a little bit to post at lots of different forums with my signature link -- he would be getting paid to do something he already likes to do.) ~~~~~~ Does this seem right to you guys? The numbers are just for illustrating the concept. So in this example, your SERP for a keyword is based on "total points", where the points are a combination of all the factors: Scenario A) - You could have 10,000 forum sig links from 1 forum. Each link = average of 0.0001 point, so you don't get all that much. .0001 x 10,000 = 1 points for the links themselves Scenario B) - 500 links from 50 different, fairly large/very large forums 500 x .0001 = .05 points for the links themselves .1 point on average for each separate website/IP linking to you. 500 x .1 = 50 points for having 50 different site linking to you
It's fairly important, but only when related to other factors. For example, you have a site about books, and A 8 year old PR8 sites about books links to you from within a good article that attacts a lot of other links to it. That's pretty close to a perfect link. The PR od the page helps as it shows that Google places a high importnace on the site - however other factors like it's age, the location of the link, the poplarity of the article and the theme are all important factors. True, a link from a completely empty page using no anchor text would do comparitively little when compared to the instance above. Possibly yes, if the link was on an authority in the same niche. Yes again, possibly correct. If the link within a related article on an authority site then again it would be a very strong link. Ummmm, not really, many quality links would be good. Many crappy directory links bad With links there's not really anything in black and white. People need to get away from giving links values based on PR, and as you are doing, start to look at their quality and topic. Google isn't stupid, they know that their PR metric is easily gamed, so no longer attribute as much importnace to it as they used to. I think the best metric to look at when building links is the ease of placement. I usually look at it this way, Comment and forum spam - Easily gained with no moderation, bottom of the tree Directory links - Some moderation, though still minimal on some directories Blog links - Middle of the road, a large buying and selling market to manipulate them, but some still good quality Article links - Similar to blogs, on the better quality ones can count for a lot Authority sites - Think, BBC, CNN, About.com (though there are also lots of niche sites) Can't get any better than this
I should have worded that a bit differently. I know that lots of forums links probably wont do much, unless you have some really easy keyword. I was just using that as an example. What I meant was: Having 100 links -- all of similar quality, fron 50 different sites would be far, far better than 100 similar links from 1 or 2 similar sites.... with everything else held constant, right?
I totally agree with you. IMHO, 100's of links from a single domain is little gain over one or two. If i was given 50 links, I would prefer one link on each of 50 domains than 10 links from 5, 10 or even 20 domains. I place more importance on the domains linking to us than the number of links on the domain. Am i making sense?? lol
Unless it is a high quality website. 1 or 2 links from CNN would be great, but imagine 100 links from that domain.
I'm not so sure. I've seen people theorise that only one link per IP address will count. I'm not so sure about that, but I think the truth lies somewhere i the middle.