Scholarlink

<p>In this work, we introduce a computational framework for marketing analytics. To demonstrate the advantages of our proposal, we conducted a statistical assessment in Milagro, Ecuador, with a case study specifically focused on brand experience, examining corporate communication and key factors that shape consumer behavior during the purchasing process. The data model presented in this manuscript primarily integrates the HJ-Biplot, hierarchical clustering, and the disjoint principal component analysis method, where the strength of our model lies in their integration, with the disjoint components providing a distinctive methodological advantage and enhanced interpretability. Our methodology is designed to support the marketing dimensions that companies wish to incorporate into their studies, providing insights into consumer behavior and helping firms optimize their strategies. A tailored survey was administered via a <Emphasis FontCategory="NonProportional">Google</Emphasis> form to 712 participants, generating the real data matrix for the computational experiments. Moreover, we carried out two complementary types of simulations to illustrate the versatility of our proposal: (i) simulations based on latent dimensions, which evaluate the adaptability of the methodology across different structural components of marketing, and (ii) simulations based on empirical sampling, including row-wise and column-wise generation schemes that preserve or diversify observed variable combinations. Finally, we provide a <Emphasis FontCategory="NonProportional">Python</Emphasis> software library that enables researchers to implement and extend our computational approach in diverse marketing contexts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A statistical software framework for marketing analytics: methodology, simulations, case study, and a python library

  • 期刊论文

  • 发表日期：

    2026年3月19日

查看全文
  • Maria Fernanda Contreras-Peña,
  • Juan-Salvador Victoria-Mas,
  • Carlos Martin-Barreiro,
  • Xavier Cabezas

摘要

In this work, we introduce a computational framework for marketing analytics. To demonstrate the advantages of our proposal, we conducted a statistical assessment in Milagro, Ecuador, with a case study specifically focused on brand experience, examining corporate communication and key factors that shape consumer behavior during the purchasing process. The data model presented in this manuscript primarily integrates the HJ-Biplot, hierarchical clustering, and the disjoint principal component analysis method, where the strength of our model lies in their integration, with the disjoint components providing a distinctive methodological advantage and enhanced interpretability. Our methodology is designed to support the marketing dimensions that companies wish to incorporate into their studies, providing insights into consumer behavior and helping firms optimize their strategies. A tailored survey was administered via a Google form to 712 participants, generating the real data matrix for the computational experiments. Moreover, we carried out two complementary types of simulations to illustrate the versatility of our proposal: (i) simulations based on latent dimensions, which evaluate the adaptability of the methodology across different structural components of marketing, and (ii) simulations based on empirical sampling, including row-wise and column-wise generation schemes that preserve or diversify observed variable combinations. Finally, we provide a Python software library that enables researchers to implement and extend our computational approach in diverse marketing contexts.