Blog Post

SEO Keyword Density Is Dead: I don't think so

There was a great post on SEOMoz last month (see it here) and we've been mulling it over internally as Chad has been preparing for a panel discussion. Overall, this is a great post. It gets into many of the myths that fly around the Internet related to search engine optimization. And I want to say first-off that I enjoy and respect Rand very much. He has consistently put out high-quality materials. We cite Rand and SEOMoz often. As you know, we are not SEO consultants at HubShout. What we are really good at is consistent, transparent execution. That's what has made our SEO technology so successful in the SEO reseller / outsource SEO space. However, our SEO workflow process is only as good as the thinking behind it. And for thought leadership, we rely heavily on Rand and SEOMoz, along with a handful of other SEO communities where we are also active (such as SEOChat).

I wanted to say all that positive stuff up-front because there is an issue on the SEOMoz post I cited above that I think needs to be discussed. But I also know how easy it is to throw rocks from behind a blog screen. So I wanted to be careful to first give credit where credit is due, and then outline what I think should be a constructive dialog. As you know, we are not into controversy at HubShout. We already think too many view SEO with skepticism, and many sell it with fear - leveraging the mystery because most lay-people still think that the websites shown at the top of the search results are the ones clicked on the most... Perpetuating misinformation or selling with fear is not something HubShout endorses.

So enough preamble, what is the beef here? Well, specifically, I have trouble with the subsection "Keyword Density is Not Used - How Many Times Do We Have To Say It?" Technically, the post is correct. They point to the TF*IDF formula and offer a few citations on the math behind this approach. And from a mathematical stand-point, what most people would define as the calc for keyword density (the number of times the target word is used / total words) is not directly used by the search engines. So, let's recognize that in a very academic way, what Rand is saying is correct.

However, this headline is seriously misleading. While the keyword density calculation is a massive over-simplification of what is really happening in the TF*IDF calculation, if you dig in to the math (which we have done at HubShout) you see that TF = Term Frequency. So, the poor-man's calculation of keyword density is actually in TF*IDF. Now, of course, there is much more in this search engine formula. And from my understanding of the equations, the idea is to rule out the commonly used phrases as much as possible while trying to figure out the main keyword usage in a body of text.... But let's get real, this is a very very complex bit of math. And for the common webmaster trying to develop good rankings for a high-quality site, they need rules of thumb. And since the actual count of target phrases is in the TF*IDF equation, and most target keyword phrases don't include common words like To, From, If and The, what exactly is the harm in using the poor man's keyword density calculation??? The post on SEOMoz itself says: "Yes, adding additional instances of a keyword term or phrase to a page might indeed help your rankings".

Again - I don't want to overstate my issue here as I have great respect for SEOMoz and Rand in particular. And yes, from a very technical and academic viewpoint, the Keyword density calculation most people use is not accurate when compared to something more complicated such as TF*IDF. But that being said, the main point for the lay-person is to get the target phrases on their website, along with other supporting text that is related, well-written, and useful for end-users. Viewed in this light, the idea behind a keyword density calculation are a very helpful guideline for non-SEO's. And there is no doubt that using your target keyword phrases in high-quality, appropriate ways, increases your rankings (all other things being equal). We have loads of empirical data on this point. And for right or wrong, we call that "keyword density." Albeit an oversimplication, this is something our non-technical customers and SEO resellers can understand...

Lastly, using the poor-mans keyword density calculation certainly doesn't deserve the statement "Spreading this ignorance of math and science does little to further the SEO field's reputation - let's end it." I can think of many, many other things that low-quality SEO shops say that deserve much more ridicule than this.

here) and we've been mulling it over internally as Chad has ..."/> here) and we've been mulling it over internally as Chad has ...">