Background

Consider the following scenario:

Eugene has created a flawless email campaign: it has the right content, design, call to action, and links to his social media accounts. When he sends it out, he expects to get a lot of sales. He figures that even if it converts at a "low" 5% rate on his 5,000 customers, he'd still make a reasonable amount of money within hours.

But at the end of the day, only 2% of people who saw the link clicked on it, and only 5 orders were placed. It's really sad, isn't it? What did he overlook?

RFM!!!

Eugene did everything right except one thing: targeting; everyone got the same email. As you well know, different messages elicit varied responses from customers. A customer that is price-conscious will jump at the chance to save money, while a customer that buys often may only respond to news of a new product introduction.

That's how it works!

You shouldn't try to reach everyone in your audience. Instead, you should find the customer groups that will bring in the most money for your business and focus on them (the 80/20 rule).

Imagine if you had a better understanding of your customers and sent them more relevant campaigns. You will have a considerably higher success rate, make more money, and your customers will be happier and more likely to stick with you.

Business problem

An e-commerce company's goal is to segment its clients and adopt a marketing strategy based on those segments. To reach this goal, different buying habits of customers were defined, and customers were put into groups based on these habits. 

Recency (R) answers the question: When was their last purchase date?.

Frequency (F) answers the question: How frequently do they make purchases?

Monetary (M) answers the question: What is the total value of their transactions?

How to calculate RFM scores.

To generate RFM scores, we need four parameters.

• Customer ID 

• Last Transaction date (Recency)

• Total no of transactions (Frequency) 

• Total expenditure (Monetary Value)

Dataset Description

Link: https://archive.ics.uci.edu/ml/datasets/Online+Retail+II#

This Online Retail data set contains all the transactions occurring for a UK-based and registered, non-store online retail between 01/12/2009 and 09/12/2011.The company mainly sells unique all-occasion gift-ware. Many customers of the company are wholesalers.