The effect of website events on lead purchase prediction
Loading...
URL
Journal Title
Journal ISSN
Volume Title
School of Business |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2022
Department
Major/Subject
Mcode
Degree programme
Information and Service Management (ISM)
Language
en
Pages
96+15
Series
Abstract
E-commerce and other digital mediums have gained popularity as a platform for purchasing products and communicating with companies and other consumers (Altunan et al., 2018). The rise of digital mediums can be recognized in the automobile industry as well, as online channels are increasingly present throughout the consumer decision-making journey of purchasing a car (CR Research, 2014; McKinsey&Company, 2013). Due to the rise of e-commerce and other digital channels, businesses are faced with an enormous amount of data created by consumers’ actions online (e.g. Alzubi et al., 2018; Cirqueira et al., 2020). Most online retailers convert only a small percentage of their customer visits into purchases, which is why online data offers attractive opportunities for companies to analyze their customers’ likelihood of purchase (Chaudhuri et al., 2021; Park & Park, 2016). Therefore, predicting website visitors’ purchase behavior through machine learning has gained popularity as it opens up new avenues for companies to drive e-commerce sales, for example, through product recommendations and content personalization (e.g. Altunan et al., 2018; Chaudhuri et al., 2021; Cirqueira et al., 2020). For instance, companies like Amazon and Target actively recommend commodities to visitors founded on knowledge from customers’ purchase decision processes (Chaudhuri et al., 2021). However, most current research focus on more low-involvement products, such as clothes and electronics (e.g. Chaudhuri et al., 2021; C. Chen, Xiao, et al., 2017; Hou et al., 2018). Consequently, the research around the prediction of high-involvement product purchases is relatively scarce. Therefore, this thesis aims to develop a model for predicting the purchase of a car for dealership’s leads based on website data. Six different machine learning models were fitted and analyzed for their ability to predict lead’s purchase behavior in the car dealership industry. The results indicated that logistic regression along with support vector machine was most suitable for this machine learning task. Moreover, the evidence supports that reserving a car and time spent on the website would predict purchases. One of the main contributions of this thesis is the demonstration of the possibility of purchase prediction in the car dealership industry. Moreover, the results support prior research findings, suggesting that digital mediums such as websites play a crucial role in the consumer decision-making journey in the automobile industry. Lastly, the study has taken the first steps in capturing the relationship between online interactions and offline economic events, as the purchase of cars often occur offline. Thus, the author calls for further research into the digital channel’s effect on the purchase process, especially in the high-involvement product markets.Description
Thesis advisor
Malo, PekkaPurontaus, Laura
Keywords
purchase prediction, machine learning, car dealership industry, high-involvement product, e-commerce