Product portfolio management plays a key role in the performance of a company operating in a very competitive environment, as it is the case with Sonae MC. Being able to identify which products are Star and which factors make a product a Star is of paramount importance for the company. In this Master Final Assignment, a product classification model based on a combination of different customer segmentation is proposed to identify Sonae MC's Star products. A Logistic Regression and Decision Tree models are then used to verify main factors that affect the probability of a product being a star. Results show that the probability of a product being a star is influenced by the average frequency of sales, the percentage of sales when on sale, the price level and the percentage of customers in the following 3 segments: Quality, Family and Price/Promotion. Comparing Logistic Regression and Decision Tree models, the greater predictive capacity of the latter provides better classification metrics.
| Date of Award | 20 Oct 2022 |
|---|
| Original language | English |
|---|
| Awarding Institution | - Universidade Católica Portuguesa
|
|---|
| Supervisor | Leonardo Costa (Supervisor) |
|---|
- Star products
- Classification model
- Logistic regression
- Decision trees
- Product portfolio management
Product portfolio management: what are Sonae MC’s star products?
Carvalhaes, J. L. S. A. (Student). 20 Oct 2022
Student thesis: Master's Thesis