Archive for May 18th, 2011


Keyword Density: Does It Matter?

If you’ve ever looked into SEO as a possible solution to your website traffic shortage, you’ve probably come across the concept of “Keyword density.” This is usually expressed in a statement something like:

Is your keyword density optimal? Get your keyword density to the magic number xx% and riches will rain down on you!

Run a Google search on “keyword density” and you’ll return about 1.7 million results, many of them tools designed to analyze your density so you can optimize it and become like an SEO God.

Okay. Keyword density refers to the ratio of a particular keyword to the rest of the indexable words on a given page. So, in the sentence:

Try Seafood Charlies deluxe seafood package–just right for Mom!

The density of the keyword “seafood” would be 2 out of 10 keywords, or 2/10, or 20%.  Supposedly, this tells Google, Bing, et al that the page is about “seafood” and convinces them to rank it higher than a similar page with only a 10% density for that particular keyword.

Only the story goes that too high a keyword density means you are “over optimized” and that counts against you. The gurus tell you that there is a sweet spot of keyword density. I’ve heard anywhere from 2% to 9% pretty often.

So, is it true? Do search engines really care about the ratio of keyword to text in any given page?

Um, well, maybe not so much. At least not in the simple way usually implied. Sure, Google’s famous 200+ element algorithm probably figures in density somewhere, in some fashion, to some particular end. But let’s be clear here:

There is no “optimum” keyword density percentage.

What there probably is:  a red flag that arises if the density is high enough to raise suspicions of manipulation. And the keyword density number that triggers that red flag will not be predictable at all because:

  • it will be a different value for different keywords
  • it will be a different value for different industries
  • it will be a different value for different page text word counts
  • it will be a different value for different contextual situations
  • it will depend on whether there is other evidence of manipulation on the same page

Which brings up the real reason keyword density is a bit of a straw man in a stuffed shirt chasing a red herring.

Google (and Bing too unless they’re really behind the curve) has grown smart enough to parse context. And not by counting words. The modern algorithm understands context in much the same way you do. They scan headlines and subheads, they note emphasized text, they determine what a paragraph is actually about by reading the sentence subjects and relating them to the words surrounding them. They pick out the key concepts and relate them to synonyms appearing in the immediate vicinity. Relate them to images. Relate them to the text in links pointing to that page.

But mostly, they DON’T GIVE A DAMN if the keyword density of a page is 2% or 10% or 35%, just as long as the page content makes sense and isn’t full of deceitful practices.

Proof? The number 1 page in Google for the search term “SEO” is: with a keyword density of 0.77%. That’s pretty low by any standard.

The good news? It turns out that if you write reasonably well, with your subject matter firmly in mind, with a keyword or two on a sticky note at the bottom of your monitor, you should naturally come up with a lovely keyword density. No really., with a density of 2.88% was not tweaked in any way to enhance the density. (I know because I wrote it, and I DO NOT CARE about keyword density, not one little bit.)

Write naturally. Write for a human audience. Know what you’re talking about. And it doesn’t hurt to understand what the humans are most likely to type into Google when they want to read what you’ve written.