Why you should be Keyword Clustering

Hello and welcome to. Seo is not that hard. I'm your host, ed Dawson, the founder of keywordspeopleusecom, the place to find and organise the questions people ask online. I'm an SEO developer, affiliate marketer and entrepreneur. I've been building and monetising websites for over 20 years and I've bought and sold a few along the way. I'm here to share with you the SEO knowledge, hints and tips I've built up over the years the SEO knowledge, hints and tips I've built up over the years.

Hi, this is episode 88 now of SEO is not that hard, and today I'm going to talk about why you should be keyword clustering. Now, obviously, a little bit of a vested interest here. We've recently launched and updated our keyword clustering tool our keywords people use, but we've only put all this time and effort into building this because it's actually really valuable process to go through wouldn't be worth was building it if it wasn't valuable and it wouldn't be worth your time if it wasn't valuable. So I'm going to go through what keyword clustering is and why you should be using it, and they're looking at the various different types of keyword clustering is and why you should be using it. And they're looking at the various different types of keyword clustering that are available and the pros and cons of each. So let's start with what keyword clustering actually is, and what it is essentially is where we take a whole set of keywords when it can be any number of keywords and we want to group them into clusters of related keywords so we can serve all those keywords from an individual page. So each cluster we create can be considered an individual page, and this is so that we can be efficient in our content production and also make sure that we best serve each page by making sure it covers all the keywords that Google is expecting to appear on a page that covers that topic in depth.

So this is the reason we do it, because what we don't want is, like I talked about in the previous podcast was a situation where we create a new page for every single keyword we want to target, because clearly Google's demonstrating that that's not what it wants. Also, if you do that, you encounter a situation called keyword cannibalism, and this is where, if you have two or more articles that are tackling the same topic, the same keywords, there might be variations of those keywords, but what will happen is they'll start to cannibalize each other, so each page could be ranking for some shared terms and this is inefficient and it also means Google can have issues on deciding which one of the pages for your site is going to try and consider as the best version to use for those keywords where they're starting to cross over and cannibalize each other. Now, with keyword clustering, you bring all those related keywords together so they don't cannibalize across multiple pages. They should all get covered on a single page, so that's why it's valuable in that respect. So there's very many reasons why it's worth doing now.

Historically, to do this it was actually quite challenging because you've got to do an awful lot of analysis before being able to decide which keywords should logically be grouped together. Now, what we'll do is we'll look at a few different ways of doing it. Now, obviously, ai is popular at the moment and I see a lot of people out there, a lot of influencers, saying that you can just use ChatGPT to create clusters. And you know we've actually got experience in this because we tried for a long time to get ChatGPT OpenAI-based APIs to create clusters for us that actually were good and actually served a good purpose and gave accurate results accurate results. Now, at first glance the output that ChatGPT can give you. Say, you give it 30 keywords and say, put these into clusters, it'll often come back with what looks at first glance to be quite a reasonable set of clusters. They kind of make logical sense and they kind of looks like they've got the right keywords in the right places. Now, if you scale this up to hundreds or thousands of keywords, what starts to happen is ChatGPT kind of loses itself and you can end up with it creating just one or two massive clusters with hundreds and thousands of keywords in them, which completely goes against what you're trying to do with clusters, because they're now starting to put all unrelated keywords into into huge cluster groups and it just is rubbish.

Basically, it's one of those things I think more and more people find with the eye that it looks good at first glance but then, if you delve down into the results, it just doesn't work. So please just ignore everybody that is suggesting you can do it just using simple chat gpt prompts, because it doesn't work. We've tried it by um. We spent a good few months trying it where we were sort of chaining prompts, taking data and then re-chaining and chaining and, and you know, using ai with our own algorithms on the back end to try and make it work and we just couldn't. We just couldn't get results that we were happy with, so that's why we moved on.

Another way of doing it that you'll see is NLP, and there are some free NLP-based tools out there that use NLP libraries and NLP dictionaries to try and create these clusters and again, like AI, they kind of can have a good guess. Then they're not brilliant and as soon as you again get past a certain number of keywords, you'll start to see real things that aren't related are being clustered together and you know, while it's, it's a good start. You look at it and you think, actually I can't build off this. There just is too much, too much, too many cases where we've got keywords in there that shouldn't be there you can and clearly aren't related and other times, where you know it, it it doesn't cluster keywords together. You think they should be clustered because of some of this human intuition can tell you what's right and what's wrong.

Now the third way of doing this is by reverse engineering Google and to do this what you need to do. It sounds like a simple process, but it's very data heavy. So the basic process is it's called cert clustering is the title of it before we're going any further. Yes, cert, clustering this is the third and in my opinion I think many other people's opinion is the very best way of doing it. And with this one, you take all your keywords that you want to cluster and you go to google and you do a search for each keyword and you make a note all the top 10 organic ranking urls so what sites, what pages are ranking in the first 10 pages organically and you, so you do this and repeat it for every single keyword, which then gives you a list for every keyword, all the 10 urls that are ranking organically in google for that keyword.

And then, when you've got this big data set, you then compare every keyword against every other keyword and look for that they both rank for that match and you make a note of how many matches there are between each keyword pair and then, once you've done that for every keyword pair, you then create a map of all the connections between all of the keywords and you'll set a level and you'll say I want to map every keyword paired together that has got three or more ranking URLs in common, and you can adjust that level from two to 10 if you want. If you do it, you want keywords that got 10 exactly the same URLs. You're going to get very few matches. If you do it down to two or even one matching URL, then you're going to get too many. We find that kind of the best sort of level in most cases is three or four URLs in common and then from that you can then cluster all the pairs together that have enough in common. This gives you then brilliantly constructed clusters of keywords and if you just look at them visually and you look at all the different keywords in each cluster, they just make sense. And this is because Google, with its knowledge of the content graph and the link graph and the usage, how people actually use and search, have created the clustering data for you and it's there exposed in the search engine result pages. But you have to do all this work to surface that data and then it just gives you the very, very best clusters. It absolutely destroys any kind of clustering data attempt that ai will do and it again just is so many times better than nlp based algorithms that are doing it just purely based off the words. With, with this cert based clustering, you get such fantastic results and you can also visualize these results in cluster graphs which give you such incredible amounts of detail on how keywords relate to each other.

Because even within clusters there are going to be it's like a gravity between different keywords. Keywords which have got many URLs in common have got sort of closer, denser, gravitated, a closer gravitational pull to each other than those further down that maybe only have two or three. So say you, say I want to have clusters to have at least three urls in common, each keyword pet, at least three in common. You'll find some in there that'll have nine, maybe even ten in common and they're the kind of the core of your cluster and they give you the very core um topic that you need to work on in that cluster page and um you then see the gravity sort of weakened to the ones that go down, sort of four, three in common, and you can see that they are still important keywords for this cluster. But they're not going to be the ones you focus on completely, the ones you've got to make sure you cover. But you have got to focus in on the densest part, the gravitationally most dense part, and the visualization graphs that you can create help you see this and work out how to formulate your content that you're going to build off this.

Okay, so that's the power of keyword clustering done well, and I've mentioned cannibalization is one key reason to use it. The other key reason is to essentially create less content, better content that covers the intent of all the keywords that Google is expecting. So obviously, if Google is ranking common URLs for related keyword pairs, then obviously Google's considering the intent is the same as well. So it helps you split out intent and make sure that you don't cross your intent wrong on the pages and keywords when you're sticking them together. So the real way of seeing how powerful this is is to actually try it and to try some keyword clustering.

Now, if you want to try it, you can go to keywordspeopleusecom and there's a search there. One of the options on the homepage is keyword clustering. Just click to try that out. All you need is a single seed keyword. I know that creating keyword lists can be very time consuming. If you give us a single seed keyword, we will go and pick up, do a deep search on People Also Ask or Google Auto Complete in various places and create you a good seed list of around about 100-ish keywords to start with, and it will then run the clustering process and takes a few minutes and then it'll tell you really money when it's done and you can then see the visualization and the cluster data and also the ranking data, so what the actual domains and URLs that are actually ranking across a group of clusters is, and you will see just how powerful this is. And we've done it because obviously I've helped out loads of people and shown them how to use this. And I say to them you know, give me a topic that you know about and we'll run some cluster data, get some cluster data together on your topic.

And it's amazing how much I've actually learned just by looking at clusters that have been created off single keywords. So there was a guy the other day I was helping out and he was using heat pumps, pool heat pumps, which I didn't know anything about, pool heat pumps, and just creating a cluster and then exploring the clusters that were created off, I think, a couple of hundred keywords to that one and all of a sudden I just knew so much about heat pumps and the questions people had about pool heat pumps and what were the important clusters and what the important topical areas were to be able to create topical authority. In the whole, it showed you all the individual topic areas you got to cover and that was created in, you know, less than 10 minutes. So from 10 minutes, from having a seed keyword something I was interested in I had a complete overview of the topic, how one would need to tackle that topic to create the content that Google's looking for. That matches the intent that Google is looking for, that avoids keyword cannibalization and gives you the best chance of ranking for that content. And even if you already got content, it's great to do this across your niche and see which cluster areas you are not performing in or you're missed completely. It opens up so many ideas for new content.

And, yeah, I'm just. It takes the whole idea of starting with people to ask questions and autocomplete data and the questions people are asking and getting it to a level where you know what to create, you know what gaps need to be filled and you know what your opportunities are. It's just next level stuff. So, yeah, I really strongly suggest that it's worth trying. It's not going to cost you anything to try because even on free accounts, there's enough credits on there to do this a few times before having to commit to anything at all. Um. So, yeah, you please come along and try it, and I'd love to hear any feedback. Um, and, yeah, just just do try it. Get in touch. It's a new, you know. As long as you've never done keyword clustering before, then this is something you must try. It'll really open your eyes.

Before I go, I just wanted to let you know that if you'd like a personal demo of our tools that keywords people use, that you can book a free, no obligation one-on-one video call with me where I show you how we can help you level up your content by finding and answering the questions your audience actually have. You can also ask me any SEO questions you have. You just need to go to keywordspeopleusecom slash demo, where you can pick a time and date that suits you for us to catch up. Once again. That's keywordspeopleusecom slash demo and you can also find that link in the show notes of today's episode. Hope to chat with you soon.

Thanks for being a listener. I really appreciate it. Please subscribe and share. It really helps. Seo is not that hard. It's brought to you by keywords people usecom, the place to find and organize the questions people ask online. See why thousands of people use us every day. Try it today for free at keywordspeopleusecom To get an instant hit of more SEO tips. Then find the link to download a free copy of my 101 quick SEO tips in the show notes of today's episode. If you want to get in touch, have any questions, I'd love to hear from you. I'm at channel5 on Twitter. You can email me at podcast at keywordspeopleusecom. Bye for now and see you in the next episode of SEO is not that hard.