Blake McShane joined the marketing faculty at the Kellogg School of Management in 2010 as a Donald P. Jacobs Scholar. He has developed and applied statistical methodology to topics ranging from optimizing internet ad-serving algorithms to forecasting home runs in baseball. His specific research interests include Bayesian hierarchical modeling, statistical learning, and generalized Markov models. More generally, he seeks to develop statistical methods to accommodate the rich and varied data structures encountered in business problems and to use these methods to glean insight about individual behavior so as to test and supplement existing theories. Blake earned his PhD and MA in Statistics, MA and BA in Mathematics, and BS in Economics from the University of Pennsylvania .

Bayesian hierarchical modeling; statistical learning; generalized Markov models; probability models for marketing; developing new methodology for unique data structures with application to business problems

• M.A., 2010, Statistics, Wharton School, University of Pennsylvania
• Ph.D., 2010, Statistics, Wharton School, University of Pennsylvania
• M.A., 2003, Mathematics, College of Arts and Sciences, University of Pennsylvania
• B.A., 2003, Mathematics, College of Arts and Sciences, University of Pennsylvania
• B.S., 2003, Economics, Wharton School, University of Pennsylvania

• Associate Professor, Marketing, Kellogg School of Management, Northwestern University, 2014-present
• Assistant Professor, Marketing, Kellogg School of Management, Northwestern University, 2011-2014
• Donald P. Jacobs Scholar, Marketing, Kellogg School of Management, Northwestern University, 2010-2011

• Marketing Science Institute Young Scholar
• Richard M. Clewett Professorship in Marketing, 2014-2016

• Editorial Board, Journal of Consumer Psychology, 2019
• Editorial Board, Advances in Methods and Practices in Psychological Science, 2017
• Editorial Board, Psychological Bulletin, 2017
• Consulting Editor, Perspectives on Psychological Science, 2015-Present
• Associate Editor, Journal of American Statistical Association, 2013-Present

Marketing Research; Data Analysis; Computation Statistical Methods; Probability Models for Marketing

Research (MKTG-590-0)
Independent investigation of selected problems pertaining to thesis or dissertation. May be repeated for credit.

Topics in Quantitative Marketing (MKTG-552-0)
This seminar required of 2nd-4th year students exposes students to working papers in current areas of active research. Students read, present, and discuss recent papers with the goal of improving their ability to evaluate a paper's academic contribution and managerial relevance and to further extend their knowledge of models and methods.

Quantitative Marketing: Statistical Modeling (MKTG-551-2)
This is a doctoral course on statistical models and topics alternate from year to year. Currently, in odd years the course is on Bayesian methods and computation while in even years the course is on applied and computational statistics. Marketing applications include but are not limited to conjoint analysis, choice models, data minimization, perceptual maps, etc.

Customer Analytics and AI (MKTG-482-0)
Marketing is evolving from an art to a science. Many firms have extensive information about consumers' choices and how they react to marketing campaigns, but few firms have the expertise to intelligently act on such information. In this course, students will learn the scientific approach to marketing with hands-on use of technologies such as databases, analytics, machine learning, and computing systems to collect, analyze, and act on customer information. While students will employ quantitative methods in the course, the goal is not to produce experts in statistics; rather, students will gain the competency to interact with and manage a marketing analytics and AI team. We will use the statistics program R in Customer Analytics and AI. R is harder to use that Stata but has become the industry standard (together with Python) and is extremely good for data management, visualization, and Machine Learning. Before you start the course, you will need to learn how to use R using tutorials and online course. There will be an assignment that is due at the beginning of the first class to make sure that you are sufficiently proficient in R before the course starts. Please do not take this class if you are not willing or able to make this investment. The course consists of lectures, in-class exercises, group work, and case discussions. You will use R throughout the class to work with individual-level customer data. The course has no final; instead, students are evaluated on their performance on weekly assignments. This course has no overlap with other existing analytics or AI courses at Kellogg. The course is an excellent companion to Retail Analytics.