Clustering

Which settings are you using for clustering?

Last updated ·
August 25, 2021

So, your question is actually what is the magic sauce? And, can I add my own ingredients to it? In short, no you may not (yet).

What keyword clustering settings do we apply?

First of all, we gather SERP data based on the language, location, domain and device you chose when uploading your keyword file. This is something you might do for a few keywords by entering them in the search engine and taking a look at the overall consensus of the results. You probably look at the intention, the titles of the results and even the competition.

We do that too of course, however we do it automatically and insanely fast. So sit back and relax or continue your work while you wait.

Next up, we compare the data for each keyword and determine which keywords are similar based on the SERP results. We actually let a search engine tell us which keywords are alike. This comparison method is home brewed and part of the secret sauce. Then there is something called "Outlier Detection", this is something you do in the supermarket. When you're looking for the perfect apple, you quickly scan which ones did not make the cut.

How many clusters do you choose?

The result of our comparison step is an input for our clustering method. We created a smart clustering method that finds the best number of clusters at each depth based on this comparison method.

Want to learn more? Check out our blog.

Was this article helpful?
😊
Yes
🥳
No
Have more questions?
Submit a request