During my (basic) research for an article discussing content, and whether we should still be calling it “King” or not (never queen, wonder why?), I noticed something interesting.
Most people who said content wasn't king were citing the fact that when people took this “content is everything” approach to SEO; they did so in a way that meant the content they created was, to put it bluntly, crap.
I remember seeing some text on a home page that went something like this:
“If you're interested in Burnley SEO, then we're the best Burnley SEO company to talk to when discussing SEO in Burnley. SEO is a complex subject, but it also works locally, so if you're in Burnley and your company is looking for SEO, we're the best company to talk to.”
It went on like that for a couple of paragraphs.
As you can probably guess, they were looking to rank for the keyword “Burnley SEO”. Good luck with that.
Content is not the way to be trying to rank if you're going to do it in such a blatently spammy way.
Google is already on to such tactics. They know that if there are too many instances of a keyword in a document, then it's probably spam.
In fact, they knew this for a while, everyone who knows anything about keyword frequency knows this, it's been written about for centuries.
There was a guy at a seminar I did years ago that kept pestering me for the “ideal keyword density”, I refused to give one, but he kept on so in the end, I said “3%”.
Of course, that's nonsense, but he was happy. Why? It's a nonsense figure that means nothing?
I don't recommend you research it too much (it’s boring as hell), but there's a famous rule that describes the probability of an event based on the conditions that may be related to that event. It’s called Bayes’ Therom.
Years ago I worked for a company that took this rule and worked to find the probability that a document was about a certain subject based on the words it found in the document. It seemed pretty bang-on a lot of the time, and when applied to a corporation’s knowledge, it was great at summarising and finding documents in a whole heap of unstructured data.
The thing is, it was easy to fool.
If those documents were written just like a person would normally write them, then all was peachy.
The software could analyse all the documents, and you find what you were after.
However, because humans are sneaky, it was very easy for anyone to simply increase the frequency of certain words, making sure to follow the rules on stemming and plurals, etc., and fox the system.
As a tool for searching documents, it was great, as a general search tool used to analyse web content, it was way too easy to fool.
So what's a good way to know what a document is about?
It's pretty easy if you have the sort of information and data processing power Google has.
The simplest way to know what a document is about is first to analyse the keywords but totally ignore the density.
If you have one word or six about “Classic Cars” then simply ignore them. Look for other, related words that might give you more insight.
The title of the document is also important. In fact, it's very important because you can work out the value of the document from just these two bits of information.
If the title is enticing and it introduces the subject clearly, and then the document itself matches the title in content and tone, it's likely that people will stick around and read it.
If they don't match, but people still stick around, then it could be that there's something in either the title or the document that relates to the subject, so there's a chance for the algorithm to update.
If they match, but nobody sticks around or shares it, then it's probably a case of dodgy keyword stuffing.
And so on.
I'm sure it's a lot more complex than that, but you get the idea – density is the least of your problems.
So what's the answer?
More now than ever, human experience is the biggest factor in any content marketing tactic.
If people like your content, then they'll spend time reading it and probably sharing it. If they don't, then they'll “bounce” off to somewhere else.
Google tracks all these interactions and can, therefore, tell the good stuff from the bad.
So, in short, make your content good.