Predict Store Traffic for a Top E-commerce Retailer

Project Description

Problem Description:

We are a top ten E-commerce retailer looking to create a working mathematical model or algorithm to predict customer traffic hours so that store associates can be fully available to connect with customers and engage in their journey.

Goal is to deliver a working algorithm to predict store traffic and pilot a few stores.

Technology stack: open to suggestions but current in house stack includes Java, Gurobi, Spark, hadoop, tibco suite - active spaces, BW, BE, EMS.

In your proposal please tell us your approach and suggest what factors should we look into to predict customer traffic - demographic, near by location, office or residential area, time of the year, upcoming events, one day sales etc. We can use online data as we do not have or own any of the external data. We will be glad to provide you any relevant data, if needed and if we own it.