Christina is a Senior Data Scientist in, the largest travel e-commerce company worldwide. She studied Electrical and Computer Engineering at Aristotle University of Thessaloniki, and later moved to the Netherlands to pursue her PhD in Computer Science at Delft University of Technology. For the last three years, she has been applying Machine Learning in to problems such as dynamic pricing and fraud detection. Last year, she formed a team that is currently driving the development of an experiment tool for scaling up experiments beyond A/B and causal inference.

Abstract has launched new verticals besides accommodations, such as attractions, striving to facilitate all aspects of our customers’ trip. But do these new products deepen the customers’ engagement? In other words, do our customers book more accommodations after having booked attractions with us? To answer this question, we could expose our users randomly to the new product with an A/B experiment, but we cannot effectively control the usage of the product. On the other hand, establishing causal relationships from observational data is extremely difficult due to spurious correlations. In such cases, we can use the method of instrumental variables. The talk will be about leveraging the power of A/B experiments to uncover causal relationships between indirectly controlled user behaviours and downstream business metrics. While promising in theory, we will highlight the challenges of implementing this method in the context of e-commerce, from selecting experiments that comply with the assumptions of valid instruments, to drawing valid causal conclusions from significant experiments with small effects.

Dr. Christina Katsimerou, The Netherlands