Telecom AT&T Marketing Division has a budget constraint for promoting a new consumer offer. They've requested the Customer Value Management (CVM) team to strategically drive the promotion.
The CVM team tested the offer on a random sample of 2,240 customers.
As the Machine Learning Manager, develop a machine learning model using the experiment results to identify the most receptive customers from the full customer database for the new offer. This will enable the team to maximize campaign effectiveness by targeting the most likely customers to adopt the offer and optimize marketing spend by focusing on a high-potential segment.
Id :
Customer’s unique identifierYear_birth:
Customer’s birth yearConsumer_seg:
Market segment fo CustomerRegion:
Customer’s geographical locationCltv:
Customer lifetime ValueActivation_dt:
Date customer joined the Telco OperatorRecency:
Number of days since last purchase.Device_type :
Smartphone or Non Smartphone flagData_rev_6mths:
Data Revenue in last 6 months.Voice_rev_6mths:
Voice Revenue in last 6 monthsTot_rev_6mths:
Total Revenue in last 6 months .Total_purchases:
Total purchase count of all products in the last 6 monthsTotal_engagements:
Count of engagements across all channels whether offer was suuscessfully purchased on notDigital_channel_purchases:
Count of purchases done via digital channelsUssd_channel_purchases:
Count of purchases done via USSD ChannelOther_channels_purchases:
Count of purchases done via Other ChannelsPrevResponseCnt :
Count of responses for previous promotional campaign.Complaint_cnt:
Count of previous complaintsResponse(Target Label) :
Purchase or Not based on the new offer.
I choose Random Forest because of it a versatile and powerful algorithm that is commonly used for classification tasks due to its ability to handle various data types, robustness to overfitting, and interpretability through feature importance analysis.