Class 9 Case study: Improve Marketing Efficiency for Tesco Using Supervised Learning

Author
Affiliation

Dr Wei Miao

UCL School of Management

Published

November 1, 2023

1 Business Objective

1.1 Background

Tesco is looking to promote its new private-label products to existing customers. The marketing analytics team decides to use the conventional mailing marketing strategy so that customer would receive color-printed leaflets via Royal Mails to their doorsteps.

1.2 Cost-Benefit Analyses

  • Cost: Each mail costs £1.5 to produce and another £0.5 to mail to the customers.
  • The cost is the marketing offer we send, cost_per_offer
# cost of sending an offer
cost_per_offer <- 1.5 + 0.5
cost_per_offer
[1] 2
  • Benefit: If customer responds to the offer, the management expects customers to spend £20 on trying the new products, where the COGS is 60%.

  • The benefit is the profit margin if a customer responds, profit_per_customer

# profit from a responding customer
COGS <- 0.6
profit_per_customer <- 20 * (1 - COGS)
profit_per_customer
[1] 8

1.3 Break-Even Analysis: Break-Even Response Rate

  • In order to break-even, we can calculate the break-even response rate from customers, which is the minimum response rate we need of a customer in order not to lose money from sending the marketing offer1
break_even_response <- cost_per_offer/profit_per_customer
break_even_response
[1] 0.25
  • Only if a customer responds to us with at least 25% response rate can we recover the costs of making an marketing offer.
  • If we send offers to customers whose expected response rate is lower than 25%, we make a loss by expectation.

2 Data Analytics

  • Data collection and cleaning

    • Split the data into a training set and a test set
  • Data analytics

    • Train predictive models on the training set

    • Predict customer response rate on the test set

  • Business recommendations

    • Target customers based on predicted response rate

    • Compute and compare ROIs for each targeting method: (1) blanket marketing; (2) decision tree; (3) random forest

Let’s work on the Quarto document together!

Footnotes

  1. The idea break-even is similar to the break-even quantity we learned in Week 1, the minimum incremental quantity we need to sell in order not to lose any money↩︎