Everyone talks about relevance in the SEO community.
But what is relevance?
It’s a very good question with very few good answers.
Most of the time when you ask someone in the SEO field what relevance is they will inevitably talk about content. They will say that it is important to have relevant content on your site and possibly to get links from relevant sites. But again, what is a relevant site?
I guess that a relevant site is one where the subject matter is the same as the subject matter on your site. OK….I’m listening….. I know how I, as a human, determines relevance in this regard, but how does a computer robot like Googlebot determine these things?
Remember that we are talking about a computer program and computer logic is based on rules that are quite different than the rules humans brains are based on. It’s easy to confuse the two because we are making human “judgments” every time we make an evaluation. Including evaluating the question “ how does a robot think”.
The simplest way to equate robot logic is to say that it operates based on statements. The most basic of which is “if this, then that”. This is a basic computer logic statement.
So….if I want to get a link from a relevant site I would assume that a computer “bot” would follow that link through the content on the site linking to mine and ask the question “ if this site has X and the linking site has X then relevance = YES” Problem is….what is X.
Let’s say my site is about cats. How does Google know it is about cats? Oh oh….I got this one….you say. It sees the word cat or cats on the site linking to me! OK……maybe and maybe not. What is the linking site is a heavy equipment operators blog and they refer to their machines as “ Cats”. Some of you will immediately get where I am going with this and some may not.
Who is to say that the guys who own the blog are the type of earthy characters who use slang a lot and “Cats” are short for the heavy equipment brand “Caterpillar”.
I can hear it now. Lines like “ my cat plays hard in the dirt” and “ cats are the best…I just love my cat”.
Creates an interesting dilemma doesn’t it. How does a computer bot looking at code know the difference? One could argue that Google could use other corroborating language on the page to put together the puzzle. But this presents another potential problem.
If the main subject matter seems to be cats in the example provided, then Google is going to “weight” that piece of evidence higher than other supporting words and phrases. Remember we are talking about computer logic, not human logic.
And this is only the beginning of the problem when it comes to “weighting” language criteria for a computer bot.
The placement of that content on the site matters too. Let me explain. SEO and digital marketing people will blather on endlessly about content. And they are most often referring to content that equates to article or body content written on a web page. But does Google give this content the same weight rating as other content on a website? Our testing says no. And I mean ABSOLUTELY NO!!!
I can hear the objections from my community now. “Content is king don’t you know!”
You mean that 3000 word article on the site talking about “cats “ ( heavy machinery like Caterpillar Brand ) What if I said that the page title or meta descriptions content, despite it having far less words, was more important and was “weighted “ more heavily towards a websites’ relevance score than the article content. You would naturally object to this statement for what appears to be obvious reasons on the surface.
“But there are 3000 words of content and there are only 60 letter characters allowed in a page title and only 165 allowed in a meta description.
True. But who cares. And whoever said that the number of words was the main criteria when it came to scoring a SERP ranking in Google?
That’s one giant presumption you are clinging to.
I could just say” we have conclusive and irrefutable evidence that proves otherwise” and just leave you, the reader, to guess what secret I am hiding but that is a bit cagey since you have come this far.
So here is the story.
Once upon a time…… ( I like when a story starts with that….makes me feel like a kid again!)
People in the SEO community figured out that you could “game” a search engine and influence it to behave in a predictable manner by building links with specific word content and point those links at specific web targets with similar words and rank the target website on page one very easily and very quickly. It was what we in the industry call “keyword stuffing” and “exact match anchor stuffing”.
Google quickly realized what was going on and saw it as detrimental to the user experience and modified their algorithm to punish this practice.
A lot of websites got hit really hard for this practice and dropped drastically in ranking. The SEO community basically threw in the towel and said “ the jig is up and the game is over” ……”SEO is officially dead”
But Google didn’t just make a giant sweeping change to everything about their algorithm all at once. This is not something they cannot actually do because it ends up punishing a lot of websites that are doing nothing wrong when they make sweeping changes. Google has to make algorithm changes very subtly and very gradually for this reason. Obviously, a practice like “keyword stuffing”, where people would put the same keyword or phrase on their page 100 or more times to influence the search engine is an easy practice to punish.
However, as Google tightens up on certain practices like this the funnel gets narrower and narrower as they go. There was a fairly recent algorithm that they named the “Fred update”or algorithmm change. It appeared like they were cracking down on what we call exact match domains with this update. An exact match domain is one that looks like this. Let’s say you have two million people searching for “how to lose weight” every month. So I go out and buy howtoloseweight.com because I know that by having a keyword in the domain name Google is going to “weight” that much higher then if I simply have it in my content 4 or 5 times. It seemed like domains of this nature were getting punished and dropping in ranking quite drastically with the Fred update. But this poses a problem. And this is what the problem looks like.
Let’s use the example “air quality control.” Let’s also say that this is the business that you are in. So you name your company Air Quality Control and you go and buy the domain name airqualitycontrol.com. Seems like a logical thing to do in my opinion. So tell me, how does Google know the difference between air quality control the search term that people search on the internet and air quality control the brand name? Remember, we’re talking about a computer that doesn’t think like you and I!
It reads words.
So during the Fred update, they started punishing exact match domains of this nature. The problem is that by doing so a company like air quality control, used in our example, gets punished for no wrongdoing. And this is exactly what was happening during the Fred update which of course Google had to do a huge reversal on over a period of several months as we began to see exact match domains returning to their proper position in the search results.
So how does Google deal with this conundrum? It isn’t simple, because, again, we are dealing with a programmed computer bot that works on If This Then That type statements. And Google has to be very careful, as we showed in this previous air quality control example, not to punish a group of websites that are actually doing nothing wrong just to crack down on a much smaller segment that are doing something wrong. So this is what we have discovered. And it is also what we believe Google uses as a hierarchical weighting when scoring for relevancy. So here goes. Because Google is a robot that simply looks at words for the most part and comes to conclusions about the words on a website it is too easy to game their system by placing words in content because we can put unlimited content on our website.
By content, in this example, I mean the basic article type content that we put on the pages of our website. But there are other places that we put content on the site that there are not unlimited spaces for that content. Titles only allow 60 characters. Meta descriptions only allow 165. Page permalink strings tend to get chopped off at a certain point as well. And the domain name itself is a form of content. After all the domain name is a series of words in most cases.
There are also the words that constitute “inbound links” as well. What we call “anchor text”.
Because each of these things has different character limits it would make sense that it would be harder to game some of them than others. Because gaming of the system has been a problem in the past it would also make sense to control gaming of the system by weighting things that are harder to change or have less characters to cram keywords into with a higher relevance score then things like basic on page content where there are no limitations. And this is exactly what Google has done. The one thing that I won’t to give away in this article is exactly how they order this hierarchy. But if you’re smart and have paid attention to what I’ve said so far you should be able to figure it out for yourself for the most part. Now, I can hear what you’re saying. You’re calling b******* on all of this because you’ve heard that content is King. It is King, it’s just a question of which content is king?