2024 Marketing Symposium

Discussing the past, present and future/potential of marketing science and models in the digital era. Where does marketing science come from and where is marketing science going?

Key Note Speakers & Panelists

Russell Winer (New York University)

Title:
The History of Marketing Science: Where We Have Been and Where We Are Going

Abstract:

The field of marketing science has existed roughly from the 1950s with articles in the journals Operations Research and Management Science. A report by the Ford Foundation in 1959 spurred the growth of more rigorous, quantitative approaches to handling business problems including marketing.  Key events followed including the founding of the Marketing Science Institute (MSI), the TIMS College of Marketing, and the establishment of the Journal of Marketing Research. Significant boosts to the field came with the first Marketing Science Conference held at Stanford in 1979 and the founding of the journal Marketing Science in 1983.  Since then, the field has exploded with more journals, more research, and significantly increased attendance at marketing science-related conferences.

The purpose of this talk is to describe the history of the field, to highlight past and current research areas, and to point to some directions that I feel that the field will take.  I will use word clouds to show the evolution of journal article keywords over the last 10 years which demonstrates a distinct trend towards the use of field experiments for better understanding cause-and-effect relationships.  In addition, I will show some examples from the current literature which highlight this trend as well as examples of possible future research in particular areas. Finally, I will make some concluding remarks about likely future topics in the marketing science literature.

Slides Link

Hotaka Katahira (Marunocuhi Brand Forum)

Title:
Not UGC but Structured User Narratives (SUN): A New Frontier for Future Marketing Science Research and Applications

Abstract:
Having been away from the mainstream of Marketing Science research for over a decade, I am not in a position to elaborate on the grand narrative of how Marketing Science has evolved in Japan nor predict its future trajectory.

Instead, I aim to shed light on a relatively unexplored area in Marketing Science that might interest some of emerging Marketing Scientists and Managers in the future.

The presentation is structured as follows:
1. Importance of First-Hand Information in Marketing: I will highlight the success stories of MUJI and P&G, demonstrating the critical role that first-hand information has played in their marketing.
2. Pros and Cons of UGC as a Data Source for Marketing Decisions: While User-Generated Content (UGC) is valuable as it reflects the authentic voices of users and thus enjoys emerging popularity in the discipline, it often lacks comprehensive user profiles and fails to represent the entire population. UGC participants tend to be a biased subset of total users.
3. Introduction to SUN and Its Advantages Over UGC: Structured User Narratives (SUN) are collected through computer surveys with a structured sample design, ensuring a robust set of respondent characteristics. For instance, we have successfully gathered information on people’s brand engagement and their reasons, yielding insightful results.
4. Synergy of SUN with Generative AI: I will explore how SUN can be effectively integrated with generative AI to enhance marketing insights.
5. Major Findings from SUN Applications: Key discoveries and insights obtained from applying SUN in various contexts will be shared.
6. Reviews by Marketing Executives Using SUN: Feedback and evaluations from marketing executives who have utilized SUN in their decision-making processes will be presented.

Finally, I will discuss the future directions and potential of SUN in advancing Marketing Science research and applications.

Slides Link

Makoto Abe (University of Tokyo)

Title:
Interaction between data and decision-making in the absence of statistical uncertainty: Probability-based A/B testing with Adaptive Minimax Regret (AMR) criterion for long-term metrics

Abstract:
Data and decision-making are the crux of Marketing Science. In early days, managerial decisions must be made with scarce and aggregated data. In contrast, in the era of big data, useful insight for decisions must be explored from abundance of disaggregated data. Now, the timing is ripe for data collection and decision-making interactions in order to optimize firms’ objectives.

As the importance of long-term customer metrics such as lifetime value and churn rate continues to grow, companies find themselves compelled to make prompt decisions before observing the full results.

Consider the following scenario: In an existing campaign, E, with a track record of 4 years, the average lifetime of acquired customers was 2.7 years. Now, to increase the lifetime, a new campaign, N, is being planned. If we were to implement N and compare its effectiveness with E, we need to wait for 4 years for the results to come out. If, however, we knew the probability distribution of the average lifetime of N from prior market research (e.g., 2 years with 0.4 and 4 years with 0.6), we would choose N with the higher expected value (3.2 years) over E (2.7 years). However, under "ambiguity," where this probability distribution is unknown but only its interval (support) is known, how should decisions be made?

We propose an approach called Adaptive Minimax Regret (AMR), applying the Minimax Regret criterion to the selection of campaigns E and N based on the results obtained at each point in time and updating it sequentially. The approach permits companies to respond before knowing the complete results of long-term metrics for swift marketing actions.

Facilitator

Masakazu Ishihara (New York University)