Identify Lookalike customers

feature_engineering

#1

Hi Everyone,

I am a newbie in this field. Below is my requirement on which I would appreciate your expert advice.

My client has a set of loyalty program customers who are profitable and he now wants to find out look alike or similar customers to them in rest of the customer base. I am thinking to solve this as a classification problem where in I would be deriving features of loyal customers and classify my customer base using that.

  1. Please let me know if my thinking is right OR there should be change in technique.
  2. Also, I have transactional and loyalty data as of now. I don’t think Social would be possible at this time. Can you please help me with feature engineering task as to how should I move ahead?

Thanks


#2

Find similar customers in marketing is called customer segmentation. For doing the segmentation, you should go for clustering, it will help you in creating a clusters of users with similar behaviour (based on feature provided) .
Here is basic informative blog link from AV https://www.analyticsvidhya.com/blog/2016/11/an-introduction-to-clustering-and-different-methods-of-clustering/
Hope you can make use of R/python to do this, which are open source tools used for data science