MSN, Yahoo, & Google Keyword Density Comparison from goRank.com

[bobmutch]

Interesting Report: http://www.gorank.com/research/11102004_MSN_Yahoo_Google_Density_Report.php

"For this research report, we take a look at the keyword density of on page factors for MSN's current tech preview of their new search engine technology. We also analyzed both Yahoo and Google result keyword density again. Just like in the past, we analyzed the top ten results for single keyword, low traffic queries. We used the same keywords as in previous reports so you can easily do comparisons to past results."

[I, Brian]

Yes but...haven't we all seen this type of report before?

Checking pages for the keywords on page that they rank for in engines, tells you nothing at all about the effect of anchor text on those exact rankings.

And without being able to factor in that information, the actual keyword density of ranking pages tells you absolutely nothing about the actual ranking value of the keywords present.

It's like the converse argument of counting backlinks as an indicator of links number required for ranking for a particular term. It simply does not work in these terms.

Numbers do not equal ranking value.

[bobmutch]

"Checking pages for the keywords on page that they rank for in engines, tells you nothing at all about the effect of anchor text on those exact rankings." The article does came that it does.
"It's like the converse argument of counting backlinks as an indicator of links number required for ranking for a particular term. It simply does not work in these terms."
Well there are a number of things that could for ranking weight. In fact around 100 I think Google states. I think that the number of links is one of them.

I am sure if you spend some time on McDar's tool and go over a number of the
top 1000 keywords that RankPulse monitors you will see that it is not all just links, nor is it just all allinanchor.

There are lots of examples of this. For example the keyword "blinds", nobrainerblinds.com is 3rd in Ranking but only 10th in allinanchor, and have less total links than any one else (except 8th please).

So we see there that there are other factors than just allinanchor or the number of links.

In fact nobrainerblinds.com beats out windowsblinds.net that has 108k of links and is #7 for allinanchor AND has a PR6 (all the other pages have PR5 except one).

Now this is just looking at Google. I think it is a common held theory that Yahoo places more Ranking weight on online factors.

So why not take a look and see what keyword density tells us? I think its good to look at it for all angles.

[I, Brian]

So we see there that there are other factors than just allinanchor or the number of links.

In fact nobrainerblinds.com beats out windowsblinds.net that has 108k of links and is #7 for allinanchor AND has a PR6 (all the other pages have PR5 except one).

Now this is just looking at Google. I think it is a common held theory that Yahoo places more Ranking weight on online factors.


Indeed, there are a number of factors involved - what sets up my cynicism is studies that claim to try and study the individual elements - when they plainly cannot factor in other issues, such as anchor text. This just simply destroys the whole premise of the study.

For example, to re-use John Scott's old favourite - "computers" in Google. Keyword density of the top ranking site dell.com? Zero. What does keyword density tell us about that site ranking?

Even in Yahoo! it's well within the top 10. So what role has keyword density played there?

It's not that I object to keyword density, as much as studies that claim to study *any* on-page element in isolation, as there is no closed environment to study it.

And as has been covered before, number of links tells nothing on the value of those links. For example, number of different IP ranges linked from is generally held to be more important than simple numbers from one single range.

I'm not trying to sound like I'm being personally critical - I'm trying to simply make a point on is that attempts to study the complexities of ranking need to take far more into account than they usually do.

I'm just as critical reading New Scientist, btw, which often includes conclusions with the most groundless presumptions. "Dark Energy" for one. :D

[bobmutch]

I, Brian: All that the computer search tells use is that off-page is king when it comes to Google. Some thing we all know.

That than doesn't mean that on-page doesn't matter. I still think we can look at one of the items that does have some Ranking weight with out looking at the over all picture.

[I, Brian]

Read John Scott's post on competitive searches here:
http://forums.searchenginewatch.com/showthread.php?t=2633