Analysing the brand popularity of Sri Lankan banks on social media engagement predict the most effective post time

Abstract

" Being a part of social media is a current trend all over the world. Therefore, customers are demanding the services of banks via social media platforms. But still, Sri Lankan banks are new in adopting social media marketing and academic scholarship on customer engagement, while banks using social media commercially remain limited. Hence, their popularity is also comparatively limited to other banks worldwide. To enhance the performance of Sri Lankan banks in social media, banks should know what kind of content exactly need to posted and at what hour and day. Banks need to be agile in their social media approach and avoid getting tied to a single platform. Therefore the present study attempts to address this gap by constructing eight different types of Regression models. It compares and arrives to address what kind of post characteristics drive each consumer engagement on the Sri Lankan bank’s Facebook and Instagram profile pages. Predictor variables are selected, specialised for the Sri Lanka banking environment. The study results indicate that negative binomial regression is the best model for all the engagement on Facebook except the model of love and comment model on Instagram. Partial least squares regression is best for love reaction on Facebook, and multilevel regression is best for the model of like on Instagram. Further study indicates posting on Tuesday and during work increases the Brand popularity of Sri Lankan banks. It recommended as a model which is highly flexible and resistant to any future modifications. It extends current knowledge on the performance of Facebook and Instagram customer engagement by developing a framework that helps Sri Lankan banks determine their customer engagement strategies to increase their brand popularity."