Interview Query
~3 hrs
AnalyticsMachine LearningStatisticsMarketing

Starbucks Promotion Strategy Analysis

A data science take-home/portfolio exercise about using customer features V1-V7 to decide who should receive a promotion, optimizing IRR and NIR. Submission includes explicit requirements, train/test CSVs, a notebook analysis and strategy implementation, a written report summarizing results, and a scoring script for evaluation.

StarbucksStarbucks
Data AnalystData ScientistMarketing Analyst
Updated 13 days agoReviewed byIQIQ Team

Data

The data for this exercise consists of about 120,000 data points split in a 2:1 ratio among training and test files.

In the experiment simulated by the data, an advertising promotion was tested to see if it would bring more customers to purchase a specific product priced at $10. Since it costs the company 0.15 to send out each promotion, it would be best to limit that promotion only to those that are most receptive to the promotion.

Each data point includes:

  • one column indicating whether or not an individual was sent a promotion for the product
  • one column indicating whether or not that individual eventually purchased that product
  • seven additional features associated with the individual, provided abstractly as V1-V7

The provided files are:

  • training.csv, which includes the promotion and purchase columns above so you can learn from historical outcomes
  • Test.csv, which includes only V1-V7 for each individual — these are the new customers you must decide whether to target

Task

Use the training data to understand what patterns in V1-V7 indicate that a promotion should be provided to a user.

Specifically, your goal is to maximize the following two metrics:

  • Incremental Response Rate (IRR), which is the difference between a ratio of the number of purchasers in the treatment group and a ratio of the number of purchasers in the control group.
  • Net Incremental Revenue (NIR), which is the revenue made or lost by sending out the promotion.

Metric definitions

Incremental Response Rate (IRR)

IRR depicts how many more customers purchased the product with the promotion, as compared to if they did not receive the promotion.

Mathematically:

IRR = purch_treat / cust_treat - purch_ctrl / cust_ctrl

where:

  • purch_treat = number of purchasers in the promotion group
  • cust_treat = total number of customers in the promotion group
  • purch_ctrl = number of purchasers in the non-promotion group
  • cust_ctrl = total number of customers in the non-promotion group

Net Incremental Revenue (NIR)

NIR depicts how much is made or lost by sending out the promotion.

Mathematically:

NIR = (10 * purch_treat - 0.15 * cust_treat) - 10 * purch_ctrl

Deliverable

Submit exactly one PDF file that contains:

  • your analysis
  • your promotion-targeting strategy
  • enough explanation for a reviewer to understand how you decided who should receive the promotion
  • any charts or tables you use
  • the code used for your solution, embedded in the PDF

Any code included in the PDF must be executable and sufficient to reproduce your analysis and strategy from the provided input files.

Download project files