It's 3% when I use SearchStatus plugin of firefox.
It's 1% when I use dupfreepro 2.0.
It's 1% ? Sure ?
Sorry... It's 3%. 1% is a mathematical impossibility...
If your keyword has multiple words then each word does count as part of the calculation, as otherwise, using the example above, if you entered
coffee maker reviews coffee maker reviews coffee maker reviews coffee maker reviews...
etc, for all 500 words then you'd have 500/3 = 167 instances of your keyword. 167/500= 33% keyword density for what is clearly 100% of you word count featuring your keywords.
Keyword density should be calculated as:
the total number of combined keywords used / the total number of words.
Otherwise, you end up with mathematically incorrect skewed numbers.
HOWEVER:
Keyword density should now be used ONLY as a controlling factor - not a target. Any more than 3-5% or a few mentions is counter-productive, and more importantly will read badly and get flagged at a manual review. Content should be written with the reader in mind, as any kind of keyword stuffing can lead to content being ignored and a drop in ranking; not necessarily penalised, but just not counted.
ayteck said:
that's a myth

trust me i have already tested this few times
Sorry... incorrect.
When you say you've tested this a few times, I can only presume you mean you've gotten away with it up till now. I do extensive SEO testing for corporates and the thousands of sites I own, manage and do SEO for; we're talking analysis across thousands of sites and tens/hundreds of thousands of content submissions...
Over optimisation is a VERY real factor in Post Panda 2011 SEO; adapt to it or die.