Ismail Parsa is Head of Data Science at HepsiBurada, the leading e-commerce company in EMEA.
He architected and deployed data science applications in all dimensions of online retail – including Customer, Seller, Product and Operations – that resulted in double digit growth in engagement, growth and lifetime value. He is currently working on platform solutions related to real-time personalization and interaction management (RTIM).
Prior to HepsiBurada, Ismail worked at Amazon Worldwide Marketing as lead data scientist where he defined the strategy and led machine learning-based development of Amazon’s product cross-sell/upsell programs, and the acquisition, reactivation and retention applications to support their loyalty programs (prime, mom and student). Through automated interaction management, he successfully managed a $4B portfolio of direct and email campaigns. He improved the open rates 50%, the click rates 25% and the conversion 100% while reducing customer opt-outs 50%. Prior to Amazon, he worked at Microsoft as a Principal Applied Researcher in the areas of audience intelligence, behavioral targeting and display advertising; at Citigroup as VP of Analytics; and at Epsilon as Director of Analytics.
Ismail also served the international Knowledge Discovery and Data Mining (KDD) community in various roles including General Chair (KDD-2000), Treasurer, Industrial Track Chair, Exhibits Chair and KDD-Cup Chair. He is a Ph.D. candidate in Economics at George Mason University where he was awarded graduate fellowship from the Center for Study of Public Choice for his assistance to Dr. James M. Buchanan, a Nobel Laureate in Economics.
The four key analytic dimensions of online retail are: customer, seller, product and operations. All areas have their respective analytic applications. However, in a customer-first operation, customer satisfaction and loyalty can only be maximized when all is solved based on customer need and want. Hence, predictive data science, and in particular customer segmentation and targeting, channel affinity, lifetime value and related optimization work is at the epicenter of retail analytics. We show how and where it needs to be deployed for double digit growth in conversion, revenue and profitability. Of course a proper data science process supported by a scalable, open source platform, and an ability to experiment quickly are essential for success.