A quantitative approach for product portfolio optimization

Nowadays product proliferation is a very common issue for companies, uncontrolled product launches affect revenue, profit and service level, consequently there is a need to reduce the portfolio. In this project, we propose an optimization method for portfolio rationalization based on substitutabilit...

Full description

Autores:
Castellanos Villate, Diana Alejandra
Puentes Gantiva, Sergio Luis
Ramírez Osorio, Ana Valentina
Tipo de recurso:
Trabajo de grado de pregrado
Fecha de publicación:
2019
Institución:
Pontificia Universidad Javeriana
Repositorio:
Repositorio Universidad Javeriana
Idioma:
spa
OAI Identifier:
oai:repository.javeriana.edu.co:10554/45525
Acceso en línea:
http://hdl.handle.net/10554/45525
Palabra clave:
Modelo de optimización
Racionalización de SKUs
Sustituibilidad
Ganancia promedio
Cadenas de Markov
Optimization model
SKU rationalization
Substitutability
Average revenue
Markov chain
Ingeniería industrial - Tesis y disertaciones académicas
Productos nuevos
Ganancias
Cadena de valor
Rights
openAccess
License
Atribución-NoComercial-SinDerivadas 4.0 Internacional
Description
Summary:Nowadays product proliferation is a very common issue for companies, uncontrolled product launches affect revenue, profit and service level, consequently there is a need to reduce the portfolio. In this project, we propose an optimization method for portfolio rationalization based on substitutability and average revenue. In order to transform the substitutability into a quantitative criterion, a Markov chain approach was implemented. This approach describes the substitution behavior and allows to calculate the redistribution of customers in the remaining SKUs. For each possible portfolio, there is a Markov chain that must be evaluated to know the future revenue performance. So, the number of possible solutions and the complexity of the problem increase exponentially as the number of SKUs increases. A Tabu search metaheuristic was proposed to solve this combinatorial problem. Since all the companies do not have the same needs, requirements and expectations about the portfolio rationalization, two different contexts were defined. First context refers to companies that have no data input for the model because they have not done analysis about the rationalization. While second context refers to companies that have already defined a constraint for the reduction, the minimum percentage of SKUs to remove or the maximum revenue that the company is willing to lose. Aiming to evaluate the performance of the designed model, we simulated a case of study where a company is trying to reduce a portfolio of sixty products. Finally, from the analysis of the results we provided some insights about the way the model selects the products according to their revenue, preference and substitutability levels.