MS773 S25: Beacon Electronics Final Project
Customer Targeting Analysis
This assignment will help you build practical skills in the analytics of customer profitability and predictive modeling. You will analyze realistic customer data from the technology retailer Beacon Electronics, and make a recommendation about which customers to target for a promotional campaign.
Submission Instructions
√ Download your group’s Excel workbook and data on Blackboard
√ Complete this assignment together with your assigned group
. It is a violation of the Academic Code of Conduct to discuss this assignment or collaborate with other groups
. Each group’s dataset is different
√ This assignment can be completed entirely in Excel and Enginius. You are welcome to use Python or other tools if you wish
√ Write your answers to the questions below in the Answers Tab of your Excel workbook
. Ensure answers are easy to read in Excel — do not run them off the screen (points will be deducted for poor formatting)
. Submit your workbook on Blackboard, including all of your calculations
√ Late penalties:
. Up to 3 hours late: - ½ letter grade (e.g., from A- to B+)
. 3 to 8 hours late: - 1 letter grade (e.g., from A- to B-)
. Over 8 hours late: not accepted
Assignment
Beacon Electronics, a popular retailer specializing in computers, electronics, and tech gear, has been a fixture of the New England technology scene for decades. Customers love Beacon for its deep product selection, knowledgeable staff, and vibrant in-store experiences. The store is known for a high-touch customer experience; many of the commission-based salespeople are tech geeks themselves, and are always eager to help.
The company has been comparatively less successful building its online business. In an effort to increase online spending, Beacon has developed a novel promotion: selected in-store customers will receive a two-part promotion called Beacon Bucks. Customers chosen for the promo are presented with a surprise at the checkout counter: they instantly get $10 off of today’s purchase. They also receive a code to get 10% off of their next online purchase, as long as they come to the site within the next month. The goal is to drive brick-and-mortar customers online, in hopes that they will become more frequent customers with higher Customer Lifetime Value.
To test the idea, Beacon Electronics ran an experiment of this promotion with 4,000 customers selected at random. Of the customers who received the Beacon Bucks test promotion, some number subsequently made online purchases. Beacon hopes to learn more about the customers who came online. If they can understand the characteristics of customers who responded, then they can target who gets the offer in the future. For example, if in the experiment, Beacon spent $40,000 (4,000 x $10) to drive 500 customers online, that would be $80 per customer. Your job is to Beacon spend its promotional dollars more effectively.
Part 1: Develop a prediction model and targeting recommendation.
The database has been split into two parts. The training (calibration) database includes 1,800 customers, of whom 400 went on to make online purchases. The validation database (out-of-sample data) includes the remaining 2,200 customers, of whom some number made online purchases. Use the training database to calibrate a prediction model. Then test this model against the validation database to develop your targeting recommendation.
In the Excel workbook provided, answer the following questions:
1. Show and interpret your model, including the parameters and model statistics
. Briefly interpret the model coefficients and explain how it aligns with your expectations
. Which model did you use and why (for example, did you use RFM?)
. How confident are you in your model’s predictive performance and why?
2. Use your model to develop and recommend a targeting recommendation.
. Show your targeting recommendation
. Justify your recommendation. Be sure to quantify the benefit of using your recommendation
3. If you could change one thing about the way Beacon conducted the promotion
experiment, what would you change and why would it result in better targeting?
Part 2: Use Customer Lifetime Value to refine the promotion.
Based on past analysis, Beacon Electronics has found that its online customers have an average retention rate of 55% per year. Early insight from the promotion test, however, shows that customers recruited through the promotion are less likely to churn. At last week’s Marketing meeting, your colleague, a recent Masters in Management graduate, said that Beacon Electronics should increase the $10 promotional offer to something higher.
If customers recruited through the promotion have a 60 or 65% retention rate, how much more should Beacon Electronics be willing to spend on the promotional offer?
In the Excel workbook provided, answer the following questions:
4. What is the CLV of Beacon’s customers?
. What is the CLV of the average customer?
. What is the CLV of customers who are targeted for the promotion?
. You may assume:
— Beacon’s average gross margin is 30% (you can ignore the discount that is part of the promotional offer)
— Beacon Electronics uses a 12% discount rate for internal calculations
5. How much should Beacon increase the promotional offer?
. What is the right amount to give away (instead of today’s $10 offer)?
. Justify your recommendation by quantifying the expected cost and benefit