On-page is the granddaddy of all metrics. Before Google, before Yahoo, the W3 catalog, all of them were built against an algorithm called the LDA or the latent dirichlet allocation. You guys don’t need to remember that, it will not make you, you know, richer in the future, but suffice it to say the LDA is what we should refer to in layman’s terms as the keyword density algorithm. It’s the keyword density algorithm.
Now, the LDA is a very complex algorithm, but it works in a very elegant way. What it does is it tries to understand what the content is about based on a mathematical breakdown of what words appear where in what frequency. So, what words appear where in what frequency. The reason keyword density and the reason keyword presence still matters across all content inside the website is because of the LDA and it still sits there at the core of the algorithm. Currently it’s complemented by more sophisticated algorithm, like the Caffeine mitigates it, the LSI mitigates it, but the LDA is there.
It relies on the presence of terms in order to identify relevance and this is why content will always be king. Now, one important thing to realize when it comes to talking about the LDA or the keyword density algorithm, it isn’t all about the frequency of what words appear where. In fact, the “where” is very important.
According to the LDA, a word is more powerful – a keyword, a phrase is more powerful – the closer it is to the top and to the left. And that sort of makes sense, because most of the time when we talk about topics, the topic is the subject and everything else is a predicate, right? So it follows natural human behavior. The LDA attributes more weight to content that’s present above the fold in the top hundred words to the top and to the left of the page.