Individual Link Value

[MBayes]

We understand the PA, DA, trust and all of that.  My question is, is there a process or formula anyone uses that shows an individual links value as to the link juice it passes. The old Domain Juice seemed to be that, but after further investigation (And Rand setting me straight)  I now understand it's not a good metric.

Today, we use PA  divided by the number of external links on that page to get some sense of an individual links actual value to the site or page we link to. I understand this is a very sloppy system, but seems to be the only choice we have?

It's based on this simple thought.  If you get a back link on two different pages, and both are equal in every way,  except one has 3 outbound links and the other has 30, the link from the page with 3 will be significantly stronger as far as passing juice.

So...  anyone using something to determine an individual links value?  I did ask the SEO staff, and they do not current have it.

[EGOL]

I believe that the problem is.... people are spending too much time worrying about the value of a link and not enough time producing something worth linking to.

[EGOL]

Nobody knows but Google... but I think that advertising links and reference links (usually) have very different formats.

[onlinemediadirect]

you mention reference citation. My question to you is does the google algo actually look out and read titles such as references ? As it would in a similar way to spot advertising links ?

[EGOL]

Rather than use numbers I would use qualitative measure....

- How relevant is the site?

- Where is the link on the page?  (in footer?... in contextual paragraph?.... in sidebar?... above the fold?.... in reference citation?

- Is the link on a kickass domain or a dog?

I think that these are much more important than numbers.

[MBayes]

Ryan, yes there are a lot of factors.  But ....  going back two page example. Two pages of the same value. One has 3 outbound links, the other has 30.  Clearly in most cases the page with 3 is better.  In other words, if you have this situation 100 times, you might see it proving correct 70 or 80.

I doubt that figuring out several of the SEOmoz metrics is much (if at all) more difficult than the one I am describing, and a Link value would be of huge.

Every metric we see on reports is in fact a guess. PA, DA, Trust, Cblocks, Backlinks.. none can be completely trusted and we all use these tools with that understanding, and the hope that they are at least generally correct.  What would be any different in a link value report that took as much into consideration as possible?

Can you imagine the time savings and efficiency acceleration of Link building if such a tool existed and was even somewhat accurate.

I think SEOmoz attempted this with the old Domain Juice passed metric. But seeing the formula for it, I can understand why they felt it was not very helpful.

[RyanKent]

I am not aware of any solid tool that provides this information. You may find a tool which estimates or otherwise provides a link value, but the challenge is that guesses are being stacked upon other guesses.

If someone responds "yes, try the Link Valuation Tool from Company X" my questions would be:

• What metric is being used to value the link? PA? DA? PR? If PA/DA are being used, then those metrics are limited by the Linkscape crawler and the various factors concerning it's use (i.e. 1-2 months behind, issues mentioned by Carin, etc). If PR is being used, then the tool's PR is a guess and may be quite different from Google's PR

• How is decay being handled? Is the PA/DA/PR being fully distributed? Or is the natural decay being calculated, and if so how? It's another guess factor.

• How is the weighting of links being handled? The SEO consensus is that links in content are given more weight then links in footers and other site-wide links.

There are other factors such as multiple links to the same domain, multiple links to the same page, etc. I feel there are too many unknowns for a tool to provide a meaningful link valuation. I would love to be proven wrong. Such a tool would clearly offer great value to SEOs.