Studying User Behavior on OLX

OLX is one of the biggest online marketplaces in Pakistan, where people can advertise anything that they want to sell and other people can view these ads, contact the seller, and purchase the item. Recently, the website had also allowed real estate to be advertised and sold on their platform. Tens of thousands of people navigate the property section of the website each day. Looking through logs of event-streams, we identified and observed unique event pathways through exploratory data analysis taken by a user and to solve the issue of not knowing if the purchase went through or not, we devised a conversion metric for customers to daily buyers. We also used pattern mining techniques such as the apriori algorithm to see which events are most commonly with each other. Furthermore, we performed both click-stream and cluster analysis to identify similar behavioral patterns that users took on the web platform in an attempt to identify with high probability, users that are serious about purchasing property and which users are casually visiting the website without the intention of purchasing a property.