What link affinity is for Google and what it is for

My name is David Ayala, and I’m here with a guest article in which I want to talk about affinity in links.

Semantic affinity is something that hasn’t been talked about much; however, it is highly relevant in the world of SEO.

We should bear one thing in mind. Google is getting ever more precise about search intent, which means it understands language better and better and knows how to interpret it better; therefore, it also knows how to interpret the semantic similarity between one website and another.

What the semantic similarity of websites is

The semantic similarity of websites is not an absolute value. In other words, two websites are not always 100% related; they can be more or less related, just like language itself.

Let’s take an example, because everything is easier to understand with examples. If I have a “dog food” website, a “nuts and bolts” website has zero affinity with my site. However, an animal website starts to be more related to my site, and a dog website comes even closer in semantic affinity.

Now, is a dog food website 100% related to my site? Yes and no. Let me explain. If the other site covers the same dog food brands, targets the same keywords, etc., the similarity will be very high. However, if my site covers certain brands and targets some very specific keywords and the other site targets other brands and other very specific but different keywords, then the similarity, although still high, will differ a little more.

What do I mean by this? That in semantic similarity there is no black and white; there are many shades in between.

Why semantic similarity in link building makes sense to Google

So far so good, it’s more or less clear, but the big question remains of why Google really values this semantic similarity.

We should bear one thing in mind. Link building is still part of the core of Google’s algorithm, and it is still highly relevant for SEO whatever type of techniques you use.

Also, let’s think about this. What is a link to Google? Ultimately, it is a recommendation from one web page to another web page.

Starting from this premise, let’s extrapolate this to the “real” world with a case.

• A tailor recommends a surgeon in Spain
• A family doctor in the USA recommends a surgeon in Spain
• A surgeon in the USA recommends a surgeon in Spain

Which of the three recommendations carries more weight in “real” life? Obviously, a recommendation from a tailor is nice, and a recommendation from a family doctor in the USA is great, but… what about a recommendation from a surgeon in the USA? That is huge, because they are a doctor in exactly the same speciality as you, so theirs is an expert opinion on the subject.

Exactly the same happens with links. The more semantically related the websites linking to you are, the more value the link has, because the recommendation comes from an “expert” website in your exact sector.

How to get semantically similar sites

If a website is 100% about your topic, there is really little else to check; however, we won’t always find this.

Most often, if a site is 100% about our topic and belongs to our direct competition, that would be the ideal link, but it will be very hard to get, if not practically impossible in many cases.

So what can we do?

Look for sites that are as semantically similar to ours as possible. For example, a newspaper or a multi-topic portal, even if it is a general one, may have a section on our topic.

Even so, the site may still be more or less semantically related. How do we settle the question of which site is the most semantically related, so we can choose it over the rest?

To do this, we need to research the keywords they rank for with a tool such as SE Ranking, which lets us see the total number of keywords the website ranks for and the number of related keywords.

In the following screenshot we see website A, which ranks for around 16,000 keywords in total, of which around 4,000 are related to dogs.

In the following screenshot we see website B, which ranks for a total of around 10,000 keywords, of which around 2,000 are related to dogs.

It’s true that in this case I’ve done it quickly and without looking for synonyms and semantically similar variations, but it works as an example.

Note that for website A, 25% of its keywords are about dogs, and for website B, 20% of its keywords are about dogs. As well as having a higher percentage of dog-related keywords than website B (a more specialised website), website A also has twice as many keywords on the topic as website B.

So, from this example we can say that website A is more semantically related to our dog website than website B.

When related links matter most to me

If we’re talking about free links such as forums, profiles or similar, we’re not really going to go out of our way to check semantic similarity, because even if we post on sites about our topic, it will be very hard for them to have a high affinity.

However, when we talk about paid links, it’s a different story. Since we are spending money out of our own pocket and buying powerful links, it’s best to make the most of our budget.

Now, do we have to analyse 100% of the sites we want to buy links from?

If you want greater semantic affinity, yes, but there is also the option of having “someone else” do it for you.

This is where tools such as LEOlytics can help us save time, as their tool offers suggestions of related media and portals based on its own data mining and AI algorithm.

As you can see, you have both options; reviewing the sites yourself is the more costly one compared with a tool that makes the suggestions for you directly. In the end, it all depends on how much time you can devote to this work and, above all, how many links you need each month.

You may be interested in: How to create websites with Google My Business

Conclusions on link affinity

Achieving perfect affinity is very difficult, so don’t become obsessed with getting a site that is 100% related to yours.

However, achieving partial affinity is very feasible, so if you haven’t been doing it until now, start right away and compare the SEO results you get with the links you were getting before.