Sale of a New Technologically More Advanced Laptop

P1 of 4

A nationwide retail department store is planning a promotional sale of a new technologically more advanced laptop. Promotional literature describing the new laptop and upcoming price discount coupons are to be emailed/mailed to customers. To send out promotional literature to all the department store’s email/mailing list is quite costly. The more cost-efficient approach would be to mail promotions to those customers who are more likely to buy the new advanced laptop. You are in charge of this project. Provide an end-to-end data analytics plan for this project. Include all the necessary steps from the beginning to the end. Make any assumptions and notations (if needed). Give a comprehensive description of the algorithm(s) as well as the related formulas you will use for this project (this is the fundamental part of your role in this project). Provide a detail description of the algorithm and how does it work. Please put your answer in the format of Step 1, Step 2 …

Solution:

The Retail Department is planning for a promotional sale of a new technologically advanced laptop and wants to send the promotional literature and discount coupons to its customers. However, it would not be very cost efficient to send the literature to all the consumers and hence it’s decided to take a more cost-efficient approach by sending it to consumers who are more likely to buy the laptop.

Being the in charge of this project, it would be my responsibility to analyze the complete consumer data of the store and find out consumers who are more likely to buy the laptop.

My approach here would be to follow CRISP – DM methodology (Cross – Industry Standard Process for Data Mining)

The main phases of this method are as follows, and returning to earlier phases anytime between the phases is possible.

[pic 1]

Step 1→ Business research and understanding phase

• Nationwide Departmental stores usually have many customers. As the store sells variety of products, the consumers are usually of all age groups and sex.
• The consumer’s purchase list also varies from season to season and during other sale days.
• Its not cost efficient to send promotional literature and coupons of a specific item to the complete consumer list as from the age, previous purchase data and region it would be very definite that they won’t by this product.
• Technology does interest few people more than others. There are consumers who have a fetish to own all new technologically advanced products and buy them as soon as they are launched. Hence these customers are first to be targeted to send the promotional literature.
• Secondly, we need to target consumers who buy more products related to school. This gives an idea that the consumer is more likely to be a student and would require a laptop.

Step 2 → Data understanding

• The data shared by the retail would most likely have the below variables:

Customer Name, Age, Contact Number, Email Id, Postal Address, Goods Purchased, Sex, Total Spend. Likely to buy Laptop

• I would analyze the collected to find interesting hypothesis, and state them for future verification
• I would try to understand the data completely, and raise any questions to retailer, or make assumptions in case information is not available

Step 3→ Data preparation

• Once the consumers data is understood, the next step would be to check the data for any missing values. There might be cases where one or the other variable value might be blank, we would have to fill these missing fields with appropriate values .There could be three approaches to it:

1. Fill it with constants like 0 or “missing” for numeric and categorical values respectively
2. Replace missing values with Mean or mode
3. Replace missing values with random values like a nearby value or more precise value.

• Snippet of the data shared by retailer can be found below:

[pic 2]

I have cleaned the data for age by putting the mean value in snipet below