@Kris: Have you even read this post and watched the video that Gianluca posted on this thread earlier: https://www.hobo-web.co.uk/keyword-density-seo-myth/ ? I do agree that mentioning keywords needs to happen obviously. But what I'm completely against is measuring that against a metric like keyword density. TF-IDF is already a better way (which was also already written about in this thread).
For many reasons keyword density is a flawed metric, like I mentioned before. But let me add some more explanation to it:
- The percentage of 1 keyword mentioned v.s. all words on the page. But what does that percentage mean, you have no clue right? Is 1% good, is 2% bad?
- Keyword density takes very many things not into account: the number of words on a page, if I have a 100 word description and 10 mentions of the keyword it could potentially be fine and not keyword spammy. But if I have 1000 words and 25 mentions it could potentially be that it's absolute spam.
- Competition? What are all my other competitors doing, what is their density for a certain keyword. All of that is not being taken into account when looking at these numbers. Maybe a 5% density is high in your industry, maybe it's 0.5% you wouldn't know.
These are just some examples of why the metric in itself is not more then just a percentage and for that I would never recommend using it.