Keyword density is now a more exact science, thanks to the people at GoRank, who recently released an extensive report on keyword analysis. They chose 2000 low-competition keywords and analyzed the top 10 results for all of them. They chose low-competition keywords to insure a more accurate look at how Google looks at keyword density.

I wrote to GoRank and asked about their findings. I wanted to know if there were any surprises, and how their findings mattered to website owners. Here's what they said:

Research Goals. In this research, we set out to find the raw keyword density of results on Google and differences between the top ten results. The averages are helpful for people who like general guidelines.

General Guidelines. For those people, the data indicates that the overall average keyword density for a top ten page in our results is 2.1%. For the page title, the data shows the top ranking pages have an average of between 6 and 7 words in the title with a keyword density of about 17%. The average number of words in the link text on a page was almost 200 words, and there was an average of about 1 keyword occurrence found in the link text. All of the other page elements generally had less than one occurrence on average on a page.

A better analysis of the data is the analysis of trends within the top results.

Top Ten Results Trends. Within the top ten results, the average number of words on a page tends to increase as the ranking improves, the same happens with the density. The top result had an average page keyword density of 2.5% which decreased gradually to 1.9% for the tenth ranking result. The page titles also had a very interesting pattern. The top result had an average keyword density of 20% and decreased to 16.3% at the tenth result.

The Most Surprising Information. The most surprising information was that 37% of all of the top 10 results analyzed contained the keyword somewhere in the URL. This often seems to be an optimization technique that is overlooked, but the data shows that it is a very important element. Only 10% of the results contained the keyword in the domain name while 22% contained the keyword in the filename.

Another surprising result was that the data suggests that there is an over-optimization penalty for meta-keywords, bold text, and heading text. As the ranking falls from the first position to the tenth position, the keyword density increases for these elements. It seems the more keywords you put into these elements the lower the rankings go.

More To Come. I think that is a pretty good overview of the major findings from this first set of data. Next month there should be a lot to learn by comparing how the data changes month to month. If you or your readers have any ideas for additional research that could unlock more search engine secrets please send them my way.

George
GoRank.com
Read the complete report and tell us what you think the results mean.

__________________
Garrett French
Editor, WebProNews.com
http://www.WebProNews.com

::: Focusing On Keyword Density Is A Waste Of Time - It Is Majoring In The Minors :::

Hi Garrett and George,

We have to disagree with that statement, as there is NOTHING exact about the keyword density that can be applied to ANY low-competition keyword.

Since you did not define "low-competition keyword" how is someone supposed to know WHEN to apply this data?

We mean no disrespect to anyone, but this data as it is presented cannot be useful to the Newbie who is just learning SEO, nor can it be useful to the "Intermediate SEO" because if they have learned anything they will have learned that focusing on keyword density is UTTERLY USELESS for a single keyword or for a phrase of multiple words.

For example, we'd like to give the Newbies reading this an example to look at and compare the following data with.


Our example is for the low-competition keyword (337 total daily searches according to Wordtracker and 443 according to Overture) ...

bicycling

http://www.google.com/search?q=bicycling

The #1 ranked page has a keyword density of ...

1.35%

... And has 6 occurrences. So our Newbie SEO would need to add about 4 occurrences to reach the ...

average keyword density for a top ten page in our results is 2.1%

The page's title has a 10% keyword density and therefore our Newbie SEO would need to add another occurrence to reach the ...

top result had an average keyword density of 20%

... The page displays 407 Backward Links, which is double the 205 Backward Links displayed by the data, which means that our Newbie SEO would be disappointed when they didn't acquire a #1 ranking with their 205 Backward Links and would then be wondering how many more links they would need or how much more keyword density they would need to increase or decrease in order to achieve a #1 ranking.

In other words, this data is UTTERLY USELESS as a guideline!

We teach our students that this type of data research is known as majoring in the minors and it can be very time consuming and yet produce very little in the way of actual top ranking results, since there are so MANY other factors besides keyword density that go into Google's algorithm, not to mention all of the factors that will EASILY out TRUMP keyword density.

OUR CONCLUSION: For those who are new to SEO, the data collected for this study can and should be useful as long as they understand that keyword density is probably only about 10% of what they need to focus on when it comes to a low-competition keyword and probably less than 1% when it comes to a medium to moderately competitive keyword phrase and then of course much less than 1% when it comes to a competitive keyword.

OUR ADVICE: Always write your copy for your HUMAN VISITORS (and not for your ROBOTIC VISITORS, AKA the search engine spiders) and in a NATURAL way. Forget about keyword density completely, and just make sure that you include your "low-competition keyword" in all the appropriate places, such as the HTML Title Element, Heading Elements, etc., while keeping your page size around 400 words. Then make sure that your Site Map page and related pages that link back to this page include the keyword as its Link Reputation. Following this advice will help a Newbie more (in our humble opinion, of course) than following the advice of focusing in on specific keyword densities.

Your Friends,

Sharon and Roy Montero

__________________
Want To Learn All About Search Engine Optimization? We Teach A 6 Week
Online SEO Class - Learn SEO The Write Way - There Is So Much More To
SEO Than Just Meta Tags ::: http://www.iwon.net/seo/class-description/

Interesting study Garrett - Thanks for bringing it to public attention.

IMO some of the very basic parameters of the study call the results into question, since they used only single word search terms (not a very likely scenario these days) and studied only terms which were not often searched for resulting in only getting data on how relatively unsophisticated pages were ranked which sheds no light on how the best constructed pages on the web fare.

It may be that this was done to eliminate the effects of other optimization strategies, but trying to study only what may be a minor factor in isolation make it very hard to draw relevant conclusions. For instance a more complete study might show that the effects of inbound anchor text are strong enough to mask any effects of keyword density, or that KW density just doesn't make enough difference to matter.

Some of the conclusions that are drawn from the data may not be correct however, for instance suggesting that the keyword in the URL is of ranking importance is stretching it a bit based on the information available. It would also be quite logical IMO to conclude that the top ranking pages are better optimized and better optimized pages will tend to use keywords in the domain name as a part of the general optimization scheme, than to conclude that the ranking is due to the keyword in the domain name.

But all in all a useful bit of information. Now if they could only do the same for optimized pages.

__________________
Mel Nelson
Expert SEO