In the Skype group, we’ve had several discussions on word count and keyword densities. Recently, Ted Kubaitis posted an excellent round up of the issues faced when deciding on what is your word count and how that count affects density. After reading it, we got to thinking about different tools that people use to get their word counts and decided to do a comparison between the tools.
Test Downloads
Transcription Below
If you saw in the Skype group, Ted posted this amazing post about keyword density and Cora. And what it is talking about when doing a keyword density calculation, one of the biggest problems is you need to know your word count and depending on how you do your word count would directly impact what your keyword density is.
There are many ways to count words on a page. Most people grab the words on a page and put it into Word and come up with a number and then they go back to tools that demonstrate these things and the tool has way higher that what I have. I can look at it with my eyeballs and it is much higher. The reason is that, there is more text going on in a page that in what you are considering.
There are a lot of things to consider. Do you count metadescription in a page’s word count? Do you count the footer, the sidebar, the header? Do you count things in scripts, do you count everything within a paragraph text? Do you count text that is hidden via CSS.
Last one is a big one because it turns out on a lot of webpages, there are a lot more hidden texts that you realize. Text that you are not intending to hide, especially those that are part of the template or how the page is built. There are notes and navigation that are written in paragraph text that are hidden in CSS, not nefariously but is just is, do you want to count that?
When Ted was talking about density and how you could count it and this was a really great discussion and I recommend reading it because he talks about, are you looking at the words on the page that you could see, are you looking at the words in the page that you can’t see, and then Ted identifies that Google also said that they discount one or two letter words.
So, do you want to count those, in your word count. This is a great discussion and something that I think everybody needs to think about when they are making these calculations, when they are thinking about their onpage, and then they are doing a word count, what level of word count do you like? Do you like only visible, do you like visible plus hidden, do you like the one or two letter words on not, these are things you need to consider.
So, what I did was, as I went through several tools.
Dori: Would it not matter if you were looking at your competition with the same tool and doing your competitive analysis and setting your barometer with a single tool?
That is exactly my main point – don’t shift. Once you decide that I like this and that I am using this tool, just stay with that, in terms of your word count or how you are going to calculate things from there.
What is fun is that – Ahrefs, Website Auditor, Screaming Frog, SEO Quake, Cora with their default which is called clean and then unabridged, that adds in the single letter and double letter words. Page Optimizer Pro, SEO Review Tools, Sitebulb which has two different ways to calculate which I didn’t realize as well, which I think was very similar to what Cora was doing, Serpworx which was a very fantastic tool, and Browseo.net. These pages were all chose at random.
There is really no rhyme or reason other than that some are big, some are small, some are in different platforms, just so you could kind of see. You could see how these tools handled it. You can see they are pretty well within each other, they are not too far off, some differences, but if you go and look at these pages, you will see that they are very different from if you just eyeball it.
Two things that I find really instructive – one is ahrefs did a pretty terrible job at finding. They came up with a lot of zeros which I think mean they are following robots.txt directives. Meaning, hey if your bot doesn’t crawl us, ok you win. They might not be the best tool when it comes to that. But this bottom one, kurbanali.com, Kurban works with me and this is just his private page that he put up and this page only has 17 words on it that you can see. Browseo nailed it. So Browseo might be doing a very good job with just the visual words that you can see.
It helps you figure out how these tools are making their decisions, cause they all have to make a decision – do we count this, do we not count that, do we try to count this, do we try not to count that. You can kind of see where a page that has “17” words, Screamingfrog comes out at 31, Cora clean which is the most restrictive still comes out at 22, so it is adding 4 more words to that.
Pop is at 27, Sitebulb is at 26, but then content words, I do not know how they are figuring that exactly, is 4, which is actually less than what you can see on the page. That might actually be the words within the text itself or within the body of the paragraph text.
You should probably know what tool you are using and how they are counting, and make sure that you like that idea. And then, as you mentioned Dori, the next thing from here is you just want to be consistent with it because then, the variance between them all is going to be approximately the same.
A couple of notes from where we left off, Moz do not actually do this at all, which I did not realize. So, if you are using Moz for content, they are not going to give you a word score. Siteliner, it was really difficult to do one url at a time, as is SEMrush, they aren’t the best tools if you want to just evaluate one url to see what is going on with it.
They are not really built for that and that is why they are not in here. They want to crawl the whole site first. So, if you want to look at one url in a huge site, that is not the best tool to use. That is why those tools were left off.
Anytime you see a zero, that is where the site couldn’t crawl that for whatever reason. Most is just fine with the exception of Ahrefs.
As you advance in your SEO, these are things that you want to think about and make a decision on why you like it and why you don’t, why you can live with it and why you can’t. I think that will help make better decisions when you really need to make some tough calls on should be add content or not, you know you might have variance of say – 10%, so you might be fine with wherever you are at. Those kinds of decisions.