Advancing Behavioural Analytics at Scale: Machine Learning Frameworks for Predicting Customer Intent in Large Commerce Ecosystems
Year: 2024
Type: article
Abstract: This paper assesses machine learning algorithms in predicting purchase intentions in real-time in huge trade ecosystems. Findings indicate that these high-end models like Gradient Boosting and Neural Networks far surpass the performance of the logistic regression and the Gradient Boosting has the largest AUC value of 0.94. The most predictive intents are behavioral aspects such as product perceptions, visit time, and shopping cart activities. Live scoring results in higher accuracy in real-time sessions and it has reached 0.82 in the initial three minutes. Business experiments have proven to be greatly effective and such results encompass increased conversion rates, increased engagement and also an uplift in revenues when done by small and medium sellers. In general, the framework enables predicting intents at high values and at scale.
Author: Brahmnik Chachra
Language: English
Subfield: Marketing
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W7217995324