How to judge link quality?

someone asked me

[AjiNIMC]

Recently some asked me this,

How do you rate a link quality(Q) on a scale of 5 (1 being bad and 5 excellent)?

It made me think and I made this formula but want your input to make it more effective and accurate (I know it is not messurable but still an estimate).

Link Quality Q(Link L1 on Page 1 from Site A) can be judged on these three basis,

1) Quality(x1) of Site A(Relevance,PR,Content,Backlinks,saturation) - rating on the scale of 1 to 5 (=x1)

2) Quality(x2) of Page 1
(Relevance of the page, PR, Backlinks, Total outgoing links,easy to load,level of the page) Rating on the scale of 1 to 5 (=x2)

3) Quality(x3) of the link L1
(Link prominence [how early in page], link description, anchor text, link inside an article or just a link) (=x3)

Q=.5(x1)+.4(x2)+.1(x3)

Waiting for your input.

TIA
Aji

[rytis]

1. SE wise

a. Pass PR and anchor text value (which includes over 30? things to check for, and you still gamble).

b. High in html, desired anchor text.

2. How many visitors, interested in your product, the link can send.

that's it imho

[AjiNIMC]

Thanks atleast someone replied, I really wanted an answer .. thanks again, I am trying to develop a matrix for it.

[Lorel]

I would also check the source code to see if there are any hidden tricks to keep the search engines from seeing the links (check robots.txt, css, metatags, etc.)

Also, I had one site accept my link in their directory which looked ok but then they made another page unknown to me just for advertising my site and put my title in their title and my site within a frame with a redirect in the link to my site (stealing my title, traffic and PR). So check site:www.TheirDomain.com to see if they have any such pages (their url with a redirect to other's URLs on the end) and if so avoid them like the plague.

[AjiNIMC]

Thanks Lorel,

the links (check robots.txt, css, metatags, etc.)

Very True, I have seen the real bad plays with robots.txt. I havent faced any problems with css or metatags, you can always check the cache of the page to ensure it.

Also, I had one site accept my link in their directory which looked ok but then they made another page unknown to me just for advertising my site and put my title in their title and my site within a frame with a redirect in the link to my site (stealing my title, traffic and PR). So check site:www.TheirDomain.com to see if they have any such pages (their url with a redirect to other's URLs on the end) and if so avoid them like the plague.

Thats Bad but very difficult to track such things.

I think the best will be is to see whether its cached after few days and also check in yahoo after a month or so for backlinks.

link:yoursite.com site:linksite.com in yahoo will give you the backlink from linksite.com to yoursite.com

Aji

[Lorel]

"Thats Bad but very difficult to track such things."

I catch them by watching my site meter referrals--just watch for links with their url with mine on the end.

[coconutz]

Some of the other factors we like to consider:

Linking page
Existing backlinks (external links are a plus)
Anchor text pointing to linking page
Page title
Page size
Quantity/quality of existing links on page

Link
Prominence
Leading and trailing links (if ordered by list)
Leading and trailing text
Anchor text, image or URL

[Imaster]

rel="nofollow type of links are not going to be important from SE point of view in futue. They may however get you a few referrals. Might as well tag them unimportant.

[incrediBILL]

Ok, if the site is a directory that generates reasonable traffic but uses a redirect for counting purposes, avoiding them is just silly unless your only goal is just PR and not traffic. Avoid them *IF* they aren't giving any of their links any traffic (easy to see if they show stats per link) as some directories have no SEPR themselves and it's just a waste of time.

Here's a small sampling of how much traffic my directory doled out to the top 10 new listings in my direcotry over the last 30 days:

1. 2675
2. 1967
3. 1893
4. 1491
5. 1329
6. 1258
7. 1092
8. 1028
9. 1019
10. 926

Of course my sponsored links get MUCH more traffic, but for a free listing you can easily get 1,000 clicks from qualified traffic in 30 days which is nothing to sneeze at unless all you want it PR. As a matter of fact, I saw the logs statistics of one of my sponsored links customers and it claims 100% of his traffic comes from my web site and the remainder of his traffic is statistically insignificant.

So far, this is all just organic SE traffic too....

[AjiNIMC]

My experience with directory is not that good as none is giving me any significant traffic. I guess its due to the financial field where I am.

I will have to go for niche directories to get traffic.

Thanks for the inputs.
Aji

[nuevojefe]

If it's google you're after you might want to consider whether they're sandboxed or not too.

[AjiNIMC]

If it's google you're after you might want to consider whether they're sandboxed or not too.

you mean If Site A is sandboxed then Page 1 of Site A wont be counted as a backlink

[nuevojefe]

Not necessarily that it won't count. It may not though, or it may not have full value.

Or, it's possible that linking between too many sandboxed, new sites (untrusted) might have consequences.

[AjiNIMC]

hmm life is no more simple with big G

[larryhatch]

There is another factor, maybe an overriding one:

Is the link genuine, as in <a href = ..... </a>
as opposed to www.scrapersite.com/links.php/site#1234 and the like?

The major engines might or might not even see that this leads to www.yoursite.com
If not, its worthless for SERPs rankings, good only for a few referrals.

If its an equation you want, I suggest a factor GN for genuineness.
Let the other factors be f1, f2 ....

Your equation becomes GN * (f1 + f2 + f3 + ..)

A totally phony link negates the other factors f1 etc. - Larry

[larryhatch]

I forgot to add that a rel=nofollow should figure prominently into the GN factor.
A nofollow should send GN way down all by itself. -Larry

[AjiNIMC]

The formula looks impressive, GN is very imp, I will surely add this to it.

Thanks again
Aji