Assignment Brief:

The dataset comes from an online retail company that tries to acquire customers through various means such as direct mail, organic search, paid search, etc. Each row identifies one customer that was acquired and a certain acquisition cost is incurred. Some customers make subsequent purchases, and some don’t. In order to improve the performance of the company, the marketing team decided to send out discount offers to some of the customers that were acquired through direct mail. The discount offers are included in the acquisition cost.

Please read the questions carefully and answer them within the same sheet as the question:

Question 1:

Create a table for each attribute (e.g., Gender) and record the percentage of responses for each answer (e.g., % of respondents answering “male” and % answering “female”).

Please indicate what each table represents. Avoid generating meaningless table as it would incur penalty points

Question 2:

a) Correlate as many attributes and variables as possible?
b) Find the three highest and three lowest correlations?
c) Explain what those high and low values mean? Not only in mathematical terms but also contextually.
d) What are the insights that you can generate from the correlations above?

Question 3:

Select one or two attributes to use as the basis for generating segmentation schemes. Please go beyond simple segmentations like gender or age categories, unless you have a very good reason and explanation for choosing these simple segmentation strategies.

As you generate segmentation schemes, Use the data in the “Data” worksheet to fill each cell in the tables shown below. Remember to try to create segments that are homogenous internally and heterogeneous across segments for the indicated number of segments. Not all attributes will be used.

Note: You don’t need to necessarily generate 7 segments. Use your data and your own judgment of how many segments you need.

Note: Other interesting information is important to include. This will depend on you segmentation and analysis. You can use any number of rows below the table to add futhur information about each segment.