K-means or Hierarchical clustering

Question

I have 1 million rows of customer-level data and some attributes like Revenue, Tenure, Age

Now I have to create a test & control group out of this (it is for the promotions). Should I do K-means or Hierarchical clustering?

Answer 1

You can use hierarchical clustering for this particular case as hierarchical clustering will using a bottom-up approach and consider each observation as one cluster, and will merge clusters successively together and create dendrogram and each level of it you will be able to visualize merging of all columns,