Breaking News - New Keyword Clustering launched

Hi, it's Ed here. This isn't one of our scheduled episodes, so there's no normal intro and stuff. I just wanted to share with you the audio of a video that I've just produced and put on YouTube, which runs through our new keyword clustering tool. It's a really great tool. I'll put a link to the video in the show notes, but if you want to watch the whole thing with it, with the video as well as the audio but I think I just wanted to share this with you because I think it's something really cool, something really great that we've produced and I think you know most people might find it useful. So have a listen or find the link to the video in the show notes and have a watch. Thanks very much.

In this video I'm going to show you our new keyword clustering workflows, so you can access these at keywordspeerfulusecom and you just select the keyword clustering workflow here and then there's three different workflows you can choose. This is the first one, where we have search for questions and cluster, and this will do a deep search on people. So ask for the seed keyword phrase that we put in. Then we have search for keywords and cluster, and this will take Google to complete data and for the seed keyword and then do a cluster. Or you can provide you and cluster your own data. If you select this one, it'll ask you for a bunch of keywords, or you can upload a text file full of keywords here. But I'm gonna show you the questions and cluster first. So here we'll put in the seed keyword jigsaw puzzles. We can choose the language when the country we want to target, so we'll leave this as English and United Kingdom. Then we'll just click search and cluster. Now this will just bring us to this keyword cluster page where I'll show you that this clustering is in progress and what we're actually doing now is we're actually going to the people also ask and doing a deep search, which is going down so three levels in people also ask and is now going to gather all the questions that we can find related to jigsaw puzzles, and then we're then going to cluster these.

And clustering is where, essentially, we take all these keywords and we discover which ones are topically related to each other and can be targeted together on single pages or on groups of pages. We do this by going to Google, querying Google to get the search engine results pages for every single one of those keywords that we found and we'll make a note of all the ranking URLs that organically rank for that keyword in the top 10. We then we've got that data set. We'll then compare all the keyword pairs together and find those which have URLs in common. So we will by default look for those that have three or more URLs in common and we'll consider them clustered. We can further in the process change that so we can say we have two, or we can have as many as 10 to define what is a cluster, but we find that three, at least at the start, is a good one to work for now, as this is running. You don't have to sit and wait for it to run. You can't go and do other things. You will get an email from the system when it's completed and you come back to it. It'll take a few minutes, depending on how many keywords you're trying to cluster. Here we've probably got about 100 keywords being clustered and it might take maybe three or four minutes. So we'll come back to this when it's ready.

Okay, so our um clustering process is now run and this is now our results page for our jigsaw puzzles seed keyword, and here it gives us a few little stats. So we've got 87 clustered keywords, 27 unclustered keywords and the number of clusters is 18. So a keyword is considered clustered if it's got at least three shared links with another keyword pair. An unclustered keyword is a keyword that we've not found enough linked ranking URLs to consider it clustered with something else. And this is good to have unclustered keywords, because these are the ones that don't really belong with any others. So this helps us cut down the data, the noise in the data of anything that doesn't really belong there.

So now, if we come down and look at the visualization here, this is our galaxy of clusters. This is like our topical galaxy cluster, basically, and we can just zoom out a little there so we can see everything that we can, can move this around and you can see here each of the individual clusters, and a cluster is here. You see we've got like a little cluster here with three related keywords and then we have bigger clusters here. There's sort of about six in there. You can hover over it and you can see the links between the keywords and it will bring out and highlight all the keywords for you and the amount of linked ranking URLs they've got between them.

Now some of these clusters. We would consider a super cluster, so like this one. Here you can see how it's got two elements to it that are linked in the middle, so you can really see this like a super cluster. This could be considered two pages of content or you might decide to put it all on one page, but you can see here from these clusters, these are all the individual pages that we're going to be able to create to cover this topical content. And, yeah, and this, this visualization, you can zoom in on it and we can then sort of examine how things link and this. This gives us a really good idea of how we're going to start constructing our content.

We can see the cluster data in textual format like this, where we've got all individual clusters boxed out and with the title given for each cluster, and that title is the keyword that is the most linked, so it's the one with the most ranking URLs in common with all the other pages. So this gives us kind of the core topic for each page that we're going to want to create and with these we can press the little cluster button here and we can then just examine each individual cluster on the diagram like this, so we can start to to see and interrogate it ourselves to see how we might want to produce this content and link it together. We can also see for each cluster if we click a little bar graph here, we can see the actual domains and then unique URLs that are actually ranking for these keywords and we can show what keywords are ranking on what page. So this means we can go and check out all the relevant competitors if we want to get some ideas to see how they've produced their content and how we're gonna better that. The Nitrous CSV here you can also get this data on competitors for all the clusters together. And here we are. We've got them all together and you get the number of unique URLs for each domain, total count of keywords that each domain is ranking for. We've got a topical authority score that sort of gives you some information on how strong that domain is covering that keywords in a topical area compared to all the other domains that are ranking keywords in this area. And again, you can download that CSV and you can take that away and interrogate it.

Now, going back into the cluster data here, just to mention that these keywords are sorted in the strength of relationships I've got, so the more shared ranking URLs a keyword has, the higher it will be in this list. So we know that from the top to the bottom, it's the highest to the most clustered to the least clustered in any one list. We can decide that we want to tighten the clusters by changing the link intersects. Here, and by link intersects we mean the number of shared URLs in common. So three link intersects means a minimum of three shared ranking URLs, but we could up this to say maybe five if we wanted to, and this will recalculate the clusters and it will give us a different set of clusters which are more tightly ranked, more tightly related to each other. You see here how the visualization changes, but we tend to find that in most cases, a link intersect of three gives you really good results, but it all just depends on the topic and the niche that you're playing with.

Now why would you want to do this as compared to just doing People Also Ask search? Well, the People Also Ask search data that this one on jigsaw puzzles is based on is here. So we can see this here and we can see how that People Also Ask data relates to each other. But it's not as well-defined clustering in any respect as compared to doing full keyword clustering using search engine results. This is because there's so much noise comes in, especially the deeper and further down the people also ask levels you go are that you can end up not really defining well what those clusters are.

It's a good start and it's something that people have been using for many years, but to take it to the next level and to really find how keywords cluster together in a way that Google really wants them to be clustered together and believes them to be clustered together and it evidences that in its search engine results pages you really need to do this clustering, because then it basically, rather than having all this massively linked tree structure, you end up with the discrete pages here that you know you need to create. So each one of these clusters are the discrete pages that you need to create and you can obviously take this onto an X-level. So you could take any of these clusters and you could take the keywords in it, the keyword phrases in it, find the most dominant keyword in there, the one with the most links, and you can then repeat the process over again so you can actually dive down another level. I mean we've had great feedback from our beta testers on this functionality. We've been incredibly impressed ourselves with how, just from a single keyword phrase that we've put in at the start, jigsaw puzzles and we left it to do its stuff for a couple of minutes and then it's given us all this data, both on how keywords that actually clustered together, visualizes it for us, lets us interrogate it visually, puts all that together in the cluster data so we can see what keywords we've got to hit, what the top keywords are in any cluster. And then we get all the competitor data as well so we can see who's ranking, what our competition is and what we've got to beat.

And you know if imagine if we were in here already we could be able to go and see how do we can stack up against our competitors. But to be able to visualize this and get all this information from just one seed phrase, um, you know, I mean we've we've no people who've shown this to their clients and you know, showing them this visualization shows how you can get such a deep understanding of a topical area and what you need to do to cover that topical area in depth, to get the authority that you need. I mean, I think it's unrivaled and I will strongly suggest you have a go at this. You can try it with a free account. There's enough credits on a free account to be able to try a couple of these clusters, and we've got some demo clusters as well which you can link to in there that you can try and you can explore around. So yeah, I would just say, give it a go, you won't regret it.