“Britney Spears”,
“Sex”
and “SEO”

What my Mum Knows About Britney, Sex and SEO

My Mum always told me that the best way to learn something was to watch people who were already good at it. As I was a good boy and listen to my Mum, when I wanted to learn about search engine optimization I set out to study the top 100 websites on three major keywords: “Britney Spears”, “Sex” and “SEO”.

What do Britney, Sex and SEO have in Common?

Why these three?

With “Britney Spears” and “Sex” it is simply down to traffic. “Sex” gets 350,000 1 searches a day whilst “Britney Spears” gets 55,000 a day. A website that can capture that sort of traffic can make lots of money so they represent a highly competitive arena for search engine optimization techniques. Any site that can get to the top of those keywords is worth looking at.

“SEO” on the other hand only gets 600 searches per day but it stands for Search Engine Optimization (a term that gets 14,000 searches a day). If you are a firm selling SEO, where you rank in the results for “SEO” is a pretty good indication of your ability. So despite the low traffic levels it is the target of the best and brightest in the SEO field.

SEO Techniques on Trial

Google says very little about how their engine works. To fill this silence lots of "experts" have written about how to worm your way to the top of Google’s results. Most of it is probably rubbish but several factors are persistently cited as being critical.

Page Rank	A reflection of the quality of incoming links to the page
Keywords in URL	Does the keyword (search term) appear in the page's URL? E.g. www.BritneySpears.com, www.MySite.com/Sex, or www.MySite.com/Whatever/SEO.html
Keywords in <Title>	Using keywords in the <Title> tag
Keywords in Description	Using your keywords in the <Meta> description tag
Keywords in Keyword Tag	Using your keywords in the <Meta> keyword tag
Keyword Density	How often does the keyword appear in the page?

To test these I knocked together a PHP script that scraped the first 100 search results for each of the keywords and analyzed each URL returned by Google for various factors. The results where exported to Excel where the final processing took place.

For comparison to the prime keywords  “Britney Spears”, “Sex” and “SEO” keywords, I targeted nineteen sub-prime Britney related keywords 2. These additional keywords allow comparison between the results of the big hitting keywords and more typical keywords. In total 2195 website were analyzed based on the results of 22 Google searches.

The Results

Page Rank Does Not Matter (too much)

This graph shows the average position in the results for a Page Rank (PR). There is a bias towards better results if your PR is high but for the  twelve PR 7 sites included in the results,  the average position was  41.5. Clearly a high PR does not guarantee any sort of success.

Average Page Rank (PR)	Top-Ten	All
Prime Keywords	4.83	4.24
Sub-Prime Keywords	2.21	1.75

The average Page Rank across all pages surveyed was 2.09. Over 900 sites in the survey had a Page Rank of 1 or 0 of which 132 were in the Top Twenty and 73 were in the top-ten. Of the prime keyword sites, 23 were PR 1 or 0 but only one URL of PR 1 or 0 was in the top-ten.

The advise to web masters from this is that a low Page Rank is not a killer problem. Two or three top-ten results of each sub-prime search were PR 0 or 1.

Put Your Keywords In

The four Keywords-In factors (Keyword in URL, Keywords in <Title>, Keywords in Description and Keywords in Keyword Tag) represent disputed territory for search engine optimizers. Many say that Keyword in URL (also known as having a "Keyword rich domain / path") helps where as some say it only works for uncompetitive / sub-prime keywords. Some claim that you can be punished for over-optimization by using keywords in the URL.

The general advice for the <Meta> tags (Description and Keywords) is that they are ignored by all the major search engines though some claim they have a minimal effect and are worth keeping. Only the Keywords in <Title> approach has any consistent support among search engine optimization websites. As long as the keyword in the title tag is consistent with the actual page content, then it is seen as a benefit.

That’s the talk but what are the SEO experts walking the walk?

Keyword In ...	URL	Title	Keyword Tag	Description
Prime Sites	55%	79%	56%	58%
Sub-Prime Sites	12%	13%	4%	6%

Almost 80% of sites targeting prime keywords use the keyword in the page’s <Title> tag. Over half use each of the other three Keyword-In techniques. So whatever the SEO consultants say in public, they are certainly using these techniques at least half the time. Does it make any difference? Here is the same analysis but only for the first ten sites returned by Google. The difference between the all the sites and just the first ten is shown in brackets.

Top-Ten - Keyword In ...	URL	Title	Keyword Tag	Description
Prime Sites	67% (+12%)	77% (-2%)	60% (+4%)	53% (-5%)
Sub-Prime Sites	12% (0%)	17% (+4%)	6% (+2%)	12% (+6%)

The differences between the two data sets shows that sites using Keyword-In techniques are more likely to be in the top-ten than sites not using them.  For prime sites the picture is confussed with two techniques increasing in usage and two dropping in usage. In sub-prime sites the picture is clearer because all four methods are more commonly used in top-ten sites. This suggests that for highly competitive keywords, the Keyword-In technqiues make a marginal difference that is not enough to beat a site with a higher page rank or more inbound links. However for sub-prime keywords, where there is a sea of PR 0 or PR 1 pages, the Keyword-In methods can make a real difference.

Why Use One When you Can Use Four?

How many sites use two or more the Keyword-In techniques?

Keyword In				Top-Ten		All
URL	Title	Keyword	Description	Prime	Sub-Prime	Prime	Sub-Prime
Yes	Yes	Yes	Yes	24.14%	1.20%	24.20%	0.92%
Yes	Yes	Yes	No	17.24%	0.00%	4.63%	0.29%
Yes	Yes	No	No	13.79%	3.59%	16.37%	4.02%
No	Yes	Yes	Yes	10.34%	2.40%	20.28%	1.55%
No	Yes	No	No	6.90%	4.19%	4.63%	3.56%
Yes	No	No	Yes	6.90%	0.0%	1.78%	0.0%
Yes	Yes	No	Yes	6.90%	4.19%	5.34%	1.67%
No	No	Yes	Yes	6.90%	1.20%	2.49%	0.57%
No	No	No	No	3.45%	71.26%	3.20%	77.77%
No	No	Yes	No	3.45%	1.20%	1.78%	0.80%
Yes	No	No	No	0.00%	4.79%	3.56%	6.09%
Yes	No	Yes	No	0.00%	0.00%	0.36%	0.34%
No	Yes	Yes	No	0.00%	1.20%	3.56%	0.34%
No	No	No	Yes	0.00%	1.80%	0.36%	0.69%
No	Yes	No	Yes	0.00%	2.99%	4.98%	1.32%
Yes	No	Yes	Yes	0.00%	0.00%	2.49%	0.06%

Just over 24% of the top-ten sites for prime keywords use all four techniques where as the bottom six combinations are not used in any top-ten prime sites. Keywords-In clearly offer an advantage but other methods have a greater impact as 3.5% of prime top-ten sites and 71% of sub-prime top-ten sites, use none of these techniques.

A Brief Note About Cause and Effect

With all the factors examined and particularly the four ‘Keywords-In’ criteria, the process of cause and effect has not been established. Any boost in result rankings may be illusionary. It could be that these techniques used to be important but now have no function and are simply evolutionary left-overs, much like a human's appendix. However, the fact that they are so prevalent in the keyword “SEO” is good evidence that professional search engine optimizers believe them to be useful.

How Dense Should I Be?

Like pretty much everything in the search engine optimization world, the importance of keyword density is disputed. Depending on who you ask it either doesn’t matter at all, should never go above 5% or you will be OK if you use the Once Per Paragraph Rule.

In the real world however, the numbers suggest a different story.

Average Density	Top-Ten	Outside Top-Ten
Prime Keywords	4.08%	3.79%
Sub-Prime Keywords	0.36%	0.11%

 

Prime sites are heavily focused around the 4% mark but most notable is the massive difference between prime and sub-prime sites.  This is a pretty clear signal that keyword density matters though other factors will trump it.

What Else?

Its clear that many different factors can impact on your position in Google’s results. A few of the lesser issues (or at least issues that SEO experts talk about less) are the length of the article, the use of headers ( the <H1> to <H6> HTML tags, referred to as the <Hx> tags ) and the number of outbound links from the page.

Average Word Count	Top-Ten	Outside Top-Ten
Prime Keywords	1157.67	995.17
Sub-Prime Keywords	696.57	488.62

In both prime and sub-prime keywords, having more words on the page increases the chances of being in the top-ten.

Average Word Count In <Hx> + <P>	Top-Ten	Outside Top-Ten
Prime Keywords	771.1	913.04
Sub-Prime Keywords	391.67	374.56

By looking at only the header and paragraph we can focus on the page's main content and ignore most of the incidental text such as menus and adverts. The results suggest that the optimum length is around 750 words.

Average Word Count In <Hx> Tags	Top-Ten	Outside Top-Ten
Prime Keywords	42.4	40.21
Sub-Prime Keywords	44.15	29.73

The use of header tags is recommend in search engine friendly web design but this should not be overdone. The results indicate forty-two as the optimum number but dispite 42 being the answer to life, the universe and everything, Google is unlikely to be that specific.

Average Keyword Density in <Hx> Tags	Top-Ten	Outside Top-Ten
Prime Keywords	4.9%	5.66%
Sub-Prime Keywords	0.97%	0.4%

Of those 40 or so words, 5% should be your keywords.

Average Number of outbound Links	Top-Ten	Outside Top-Ten
Prime Keywords	224.2	117.35
Sub-Prime Keywords	147.16	81.29

The number of outbound links doesn't seem to have any impact but this analysis did not differentiate between the outbound to the same domain and outbound to different domains.

From these results it looks like Google favors pages with a higher word count but prefers the main body of the text, i.e. the text in the paragraph and header tags, to be around 750 word long. Of the 750 words in the main text, forty to fifty should be in headers in which the keywords should be lightly sprinkled.

 

Why You Should Listen to Your Mum's Advice

Thanks to my Mum's advice I've learnt that there is a big difference between what the SEO "experts" say and what actual works on those high value keywords.

In the world of high-traffic keywords, a high Page Rank and large numbers of inbound Links will beat all the on-page optimizations techniques hands down. However for sites with similar page ranks and inbound links, the ‘Keyword-In’ techniques appear to make a real difference. As a large percentage of top-ten web sites employ these methods (24% use all four) it is clear that Google does not penalize web sites for using these techniques. The worst case scenario is that they have no impact.

The picture is a lot clearer for anyone happy with picking up thousands of visitors a day rather then tens or hundreds of thousands. For sub-prime keywords, most pages returned are not optimized. In fact, most of them can be described as incidental results. They are pages that mention the keyword but are not optimized for it and most have a low Page Rank. A page using one or more the Keyword-In techniques, a length of around 1000 words and with a keyword density of 4% has a good chance of making into the top-ten even with a PR of zero.

1 All traffic estimates are based on data from the great folks at WordTracker

2 They were “Britney Spears Bikini” (4482 searches per day), “Britney Spears Exposed” (4788), “Britney Spears Tattoo” (2829), “ “Britney Spears Sister”  (2783), “britney spears piece of me” (2783), “britney spears sex tape” (2400), “britney spears naked”, (2348), “britney spears gimme more” (2146), “britney spears photos” (1462), “britney spears sex” (1339), “Britney Spears in the Shower” (1279), “Britney Spears Fakes” (1188), “britney spears mp3” (1170), “britney spears peace of me”  (829), “Britney Spears Dress up Games” (702), “Britney Spears Nudity” (664), “How Much Does Britney Spears Weigh” (496), “britney spears toxic” (394) and “britney spears no underwear” (199).