Study Of Social Media Marketing And Purchase Intention Of Apparels Through Machine Learning Algorithms

Abstract

Aim/purpose – Consumers are showing interest in digital marketing tools that augment purchase intention (PI). Among several products, apparels are well-known for its own creativity and fashionable brands and the increasing rate of purchase also augments economic condition of the country. The present investigation aimed to predict PI as per dataset of Social Media Marketing Activities (SMMA) viz. Facebook (FB) add (advertisement), other social media (SM) add, short message service (SMS) add and online add among consumers who visited organized apparel sector in eastern India.

Design/methodology/approach – The forecast of PI through machine learning algorithm modelling plays an important role in recent days. In the present study, 599 datasets of consumers as per categorization of Likert scale in which 14 algorithms were selected by using WEKA tool.

Findings – The better performance accuracy predicted that three models followed by other eight models as per training and testing dataset. The PI has cumulative effect when these four SMMA viz. FB add, other SM add, SMS add and online add are functioning at a time.

Research implications/limitations – As per the study methods, a limitation of the research attributes viz. FB add, other platform, adv., SMS adv. and online adv. are focused on SMMA, without consideration of each platform adv. being used as Twitter, YouTube, Instagram, etc. Each platform along with FB may provide better data accuracy on machine learning algorithms. Other limitation observed is small number of respondents (599 nos.) and the study was carried out only in one region of India.

Originality/value/contribution – The findings showed that performance accuracy of PI of apparel products through the influence of SMMA was much better as per ML algorithms in the Eastern part of India.