About

Peter Boatwright is the Allan D. Shocker Professor of Marketing and New Product Development at the Tepper School of Business. He serves as the Director of the Integrated Innovation Institute. His work focuses on marketing, new product development, and innovation, contributing to the strategic integration of business, technology, and analytics. As a faculty member at Carnegie Mellon University, he is involved in research and teaching that emphasizes experiential learning and practical application in the fields of marketing and innovation.

Research topics

• Artificial Intelligence
• Computer Science
• Human–computer interaction
• Data science
• Knowledge management

Selected publications

• FOCUS AND MODALITY: DEFINING A ROADMAP TO FUTURE AI-HUMAN TEAMING IN DESIGN

  Proceedings of the Design Society · 2023 · 20 citations

  • Computer Science
  • Artificial Intelligence
  • Computer Science

  Abstract The evolution of Artificial Intelligence (AI) and Machine Learning (ML) enables new ways to envision how computer tools will aid, work with, and even guide human teams. This paper explores this new paradigm of design by considering emerging variations of AI-Human collaboration: AI used as a design tool versus AI employed as a guide to human problem solvers, and AI agents which only react to their human counterparts versus AI agents which proactively identify and address needs. The different combinations can be mapped onto a 2×2 AI-Human Teaming Matrix which isolates and highlights these different AI capabilities in teaming. The paper introduces the matrix and its quadrants, illustrating these different AI agents and their application and impact, and then provides a road map to researching and developing effective AI team collaborators.

  PublisherOA PDFDOI

• Examining the Personalization-Privacy Tradeoff – an Empirical Investigation with Email Advertisements

  Figshare · 2018-06-30 · 8 citations

  articleOpen accessSenior author

  Dale Carnegie once said that the sound of one’s name is the sweetest for any person. Much internet personalization acts on this mantra by trying to create an online environment where customers are greeted by name and are recommended products based on their preferences. However, no clear empirical evidence exists as to whether consumers desire personalization or whether privacy concerns override the benefits of personalization. Using theories from psychology and consumer behavior, we address this dilemma by developing hypotheses related to how consumers respond to a firm’s collection and use of information for personalization. To test these hypotheses, we propose a multi-stage ordered probit model using a hierarchical Bayesian framework to account for consumer heterogeneity via individual level parameter estimates. The data for this research comes from a website which captured information on actual consumer responses to ten million email advertisements sent to 600,000 customers over a nine month period. We examine the impact of different types of personalization as well as measure consumer response at multiple levels. We also control for consumer and promotion specific characteristics in our model. Our results not only indicate the economic benefits of personalization but also highlight consumers’ privacy concerns. The main results are as follows: first, emails personalized only on the basis of consumers’ product preferences get a more favorable response from consumers than those with no personalization. Second, we show that more than 85% consumers react negatively to personalized greetings in an email, suggesting that consumers are likely to perceive a violation in privacy if they see their name in an email advertisement. Third, we show that consumer response is mixed if both personalized greetings and product-based personalization are used in an email. While most consumers react negatively if both personalized greetings and product-based personalization are used in an email, consumers who buy more often from a firm prefer emails where personalized greetings are accompanied by reliable product recommendations. This suggests that familiarity with a website mitigates customers’ privacy concerns. Overall, we show that customers differ significantly in their preference for different types of personalization and that ‘personalized’ personalization (where different customers receive different levels of personalization) is more effective than all-inclusive personalization (in which each advertisement is personalized at multiple levels).

  OA PDFDOI

• Note of correction: “Conjugate Analysis of the Conway-Maxwell-Poisson Distribution”

  Bayesian Analysis · 2018-07-30

  articleOpen accessSenior author
  PublisherDOI

• Uncovering the Coexistence of Assimilation and Contrast Effects in Hedonic Sequences

  Research Showcase @ Carnegie Mellon University (Carnegie Mellon University) · 2018-06-30

  articleOpen accessSenior author

  Most judgments consumers make are parts of sequences and hence unlikely to be free of context effects. Assimilation (contrast) refers to a positive (negative) relationship between the value people place on the context and the value they place on the target stimulus. A general presupposition for much of the work on assimilation and contrast is that one or the other, determined by various factors, occurs. We propose that assimilation and contrast can co-occur within a sequence of experiences and present a hierarchical Bayesian model separating these effects within a unique real-world data set. We find that while assimilation effects influence overall sequence means, contrast effects are simultaneously evident between adjacent items and after extremes within a sequence. This work is the first empirical demonstration of hedonic contrast using real-world data, and the only work thus far to identify and separate assimilation and contrast effects within the same sequence of evaluations.

  PublisherOA PDFDOI

• Adding Significance to the Implicit Association Test

  Figshare · 2018-06-30 · 5 citations

  articleOpen accessSenior author

  The Implicit Association Test has become one of the most widely used tools in psychology and related research areas. The IAT’s validity and reliability, however, are still debated. We argue that the IAT’s reliability, and thus its validity, strongly depends on the particular application (i.e., which attitudes are measured, which stimuli are used, and the sample). Thus, whether a given application for a given sample will achieve sufficient reliability cannot be answered a priori. Using extensive simulations, we demonstrate an easily calculated post-hoc method based on standard significance tests that enables researchers to test whether a given application reached sufficient reliability levels. Applying this straightforward method can thus enhance confidence in the results of a given IAT. In an empirical test, we manipulate the sources of error in a given IAT experimentally and show that our method is sensitive to otherwise unobservable sources of error.

  OA PDFDOI

• Unwrapping Packaging: Does It Pay, and “How”! The Role of Aesthetically Appealing Packaging in Product Valuation

  2018-01-01 · 2 citations

  articleOpen access

  There is little to no research specifically studying whether aesthetically appealing packaging (AAP) plays a role in the evaluation and experience of products. Do products truly benefit from AAP? Do all types of products benefit equally? In this paper, we firstly provide empirical evidence demonstrating the influence of AAP on product valuation and product attitude. We also propose and test a conceptual model of packaging. We find that AAP positively impacts product valuation and attitude for hedonic products but offers no such benefits for utilitarian products. We propose a dual cognitive-affective process of how AAP may positively impact product attitude and valuation, which produces differential effects for utilitarian and hedonic products. We further find that for familiar brands, affective reactions play a greater role than cognitive reactions in mediating the impact of packaging on product attitude, suggesting that the influence of AAP may be at a more nonconscious automatic level. We present this work as a significant first step toward a fuller understanding of the conceptual role of packaging in the entire product experience.

  PublisherOA PDFDOI

• Computing with the COM-Poisson distribution

  2018-06-30 · 56 citations

  articleOpen accessSenior author

  The Conway-Maxwell-Poisson (COM-Poisson) is a generalization of the Poisson distribution which can model both under-dispersed and over-dispersed data. However, the distribution, moments, and MLE cannot be computed in closed form. This paper describes computational schemes and handy approximations for the COM-Poisson.

  PublisherOA PDFDOI

• Using Computational and Mathematical Methods to Explore a New Distribution: The ν-Poisson

  Figshare · 2018-06-30

  articleOpen accessSenior author

  A new distribution (the v-Poisson) and its conjugate density are introduced and explored using computational and mathematical methods. The v-Poisson is a two-parameter extension of the Poisson distribution that generalizes some well-known discrete distributions (Poisson, Bernoulli, Geometric). It also leads to the generalization of distributions derived from these discrete distributions (viz. the Binomial and Negative Binomial). We use mathematics as far as we can and then employ computational and graphical methods to explore the distribution and its conjugate density further. Three methods are presented for estimating the v-Poisson parameters: The first is a fast simple weighted least squares method, which leads to estimates that are sufficiently accurate for practical purposes. The second method of maximum likelihood can be used to refine the initial estimates. This method requires iterations and is more computationally intensive. The third estimation method is Bayesian. Using the conjugate prior, the posterior density of the v-Poisson parameters is easily computed. We derive the necessary and sufficient condition for the conjugate family to be proper. The v-Poisson is a flexible distribution that can account for over/under dispersion commonly encountered in count data. We also explore an empirical application demonstrating this flexibility of the v-Poisson to fit count data which does not seem to follow the Poisson distribution.

  OA PDFDOI

• Stockouts and Restocking: Monitoring the Retailer from the Supplier’s Perspective

  Journal of Business and Economic Statistics · 2016-06-22 · 5 citations

  article

  Suppliers and retailers typically do not have identical incentives to avoid stockouts (lost sales due to the lack of product availability on the shelf). Thus, the supplier needs to monitor the retailer’s restocking efforts with the available data. We empirically assess stockout levels using only shipment and sales data that is readily available to the supplier. The model distinguishes between store stockouts (zero inventory in the store) and shelf stockouts (an empty shelf but some inventory in other parts of the store), thereby identifying the cause of the stockout to be either a supply chain or a restocking issue. We find that, as suspected by the supplier, the average stockout rate is much higher than published averages. In addition, stockout rates vary widely between stores. Moreover, almost all stockouts are shelf stockouts. The model identifies stores that may have restocking issues.

  PublisherDOI

• Incidental Prices and Their Effect on

  2016-01-01

  articleSenior author
  Publisher

Frequent coauthors

• Jonathan Cagan

  Carnegie Mellon University

  17 shared

• Joseph C. Nunes

  University of Southern California

  13 shared

• Joseph B. Kadane

  Carnegie Mellon University

  12 shared

• Tanuka Ghoshal

  Baruch College

  11 shared

• Sharad Borle

  Rice University

  11 shared

• Seth Orsborn

  Bucknell University

  9 shared

• Wagner A. Kamakura

  Rice University

  6 shared

• Peter E. Rossi

  5 shared