Gaining insight to B2B relationships through new segmentation approaches: Not all relationships are equal

Matthew O'Brien, Ying Liu, Hongyu Chen, Robert Lusch

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

B2B market segmentation has both structure complexity and computation complexity. The existing market segmentation methods can not directly address these two challenges simultaneously and provide a comprehensive view of the whole problem. Nor are they able to provide guidance on the selection of the most suitable solution among candidates. This study formulates the B2B segmentation as a multi-dimensional optimization problem that integrates both customer behavior and marketing effectiveness. It applies an integrated segmentation method that unifies two market segmentation approaches: the Embedded Exchange Approach (a descriptive model) and the Predictive Satisfaction Approach (a predictive model). It proposes the use of an evolutionary based, multi-objective segmentation method to solve the structural and computational challenges. The method generates a set of Pareto optimal solutions which not only gives a holistic view of possible solutions in the Pareto optimal space but also allows marketers to use solution selection algorithm based on the properties of Pareto optimal sets. The study develops a solution selection algorithm that represents a good tradeoff of two objectives based on the geometric shape of the Pareto optimal solution front.

Original languageEnglish (US)
Article number113767
JournalExpert Systems With Applications
Volume161
DOIs
StatePublished - Dec 15 2020
Externally publishedYes

Keywords

  • B2B market
  • Market segmentation
  • Multi-objective market segmentation
  • Pareto optimal solution

ASJC Scopus subject areas

  • Engineering(all)
  • Computer Science Applications
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Gaining insight to B2B relationships through new segmentation approaches: Not all relationships are equal'. Together they form a unique fingerprint.

Cite this