Identification of customer purchasing behaviour via clustering algorithm in the crushing spare parts business market

Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

School of Business | Master's thesis

Authors

Date

2024

Department

Tieto- ja palvelujohtamisen laitos

Major/Subject

Mcode

Degree programme

Business analytics

Language

en

Pages

101 + 7

Series

Abstract

The thesis investigates the application of the clustering algorithm to identify customer purchasing behaviour within the Crushing Spare Parts (CSP) business market. The study aims to address the complexities inherent in B2B customer segmentation by leveraging data-driven techniques to uncover distinct purchasing patterns and suggest strategic recommendations to improve current pricing strategies. Utilizing the CRISP-DM framework as the methodological foundation, the research integrates both quantitative and qualitative approaches. The k-means clustering algorithm is applied to transactional data from the case company, a Finnish industrial firm specializing in technology and services for aggregates, minerals processing, and metals refining. The analysis is guided by a theoretical framework that incorporates concepts from market segmentation literature, particularly focusing on the challenges and dynamics of industrial markets. The thesis identifies distinct customer clusters based on purchasing behaviours, including frequency, monetary value, and specific attributes like participation in Fixed Price Agreements (FPA) and the purchasing of proprietary parts. These clusters are further profiled to uncover strategic insights that can enhance the company's pricing practices and improve customer potential. Key findings include the identification of actionable customer segments that demonstrate potential for improving their purchasing patterns, possibly generating more revenue. Additionally, the study outlines several challenges encountered during the clustering process, such as data quality issues and the need for ongoing model refinement. Recommendations for future research emphasize the inclusion of additional variables, such as lead times and customer loyalty metrics, to further enhance the accuracy and applicability of the segmentation model. In conclusion, this thesis contributes to the existing literature on B2B customer segmentation by demonstrating the practical application of clustering algorithm in a complex industrial context. It offers valuable insights for both academics and practitioners seeking to implement data-driven segmentation strategies in similar business environments.

Description

Thesis advisor

Malo, Pekka

Keywords

B2B customer segmentation, RFM, k-means clustering, crushing spare parts

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202410206784

Collections

[gradu] Kauppakorkeakoulu / BIZ