An Optimized Keyword Clustering Algorithm by Alejandro Rioja Is Now Revealed

SAN FRANCISCO, Dec. 15, 2020 /PRNewswire/ -- Since its introduction in 2015, Keyword Clustering has been a lesser discussed topic among SEOs, and it's somewhat misused to this day. 

To keep it simple, clustering aims to group keywords that are closely related and thus return similar results in the SERPs (Search Engine Results Pages). 

Traditionally, clustering is done by introducing some sort of bias. For instance, telling the algorithm that if X many URLs are shared among keywords, then those keywords belong together. 

Enter unsupervised learning

Unsupervised learning is a type of machine learning that looks for previously undetected patterns in a data set with no pre-existing labels and with a minimum of human supervision.

Alejandro Rioja, alongside the team at Keyword Cupid, used historical AdWords, Google Analytics and Search Console data to overlay an "expected conversion value" on top of an unsupervised-learning algorithm to group keywords in a very sophisticated way. This allows the algorithm to use revenue, instead of just search volume, and much more closely tie efforts back to business impact. 

As of the time of this writing, this is the most advanced clustering publicly available in the market today. 

Since this algorithm relies on unsupervised and evergreen metrics, it is impervious to changes in the ranking algorithms since it simply takes the URLs without any bias and runs the clustering. It models the world based on the view of behemoths like Google, Bing and Yandex and the better they become, the better its results will be. 

The algorithm's output has already influenced over $150M in deals for the companies Mr. Rioja has worked with (such as Future Sharks, So Influential, Authority Daily), and given the computational power of Keyword Cupid's clusters, it can scale up without any issues. 

Media Contact
Rahul Dabir

Cision View original content to download multimedia:

SOURCE Alejandro Rioja

Posted In: advertisingPress Releases