Marketing Dynamics 2022 Review

This week I loved the Marketing Dynamics conference 2022, which has evolved a lot since I started it as an invitation-only gathering back in 2004. For one, the full first day was devoted to PhD student presentations, with excellent discussants. For another, we saw 2 presentations with coauthors from Amazon Ads. The key constant throughout two decades though was the conference’s competitive advantage, expressed by a participant as ‘the only place where expert dynamic modelers actually come to your talk of 30-45 minutes and give you detailed feedback on how to improve your work’. Probably the best illustration was Prof’s Prasad Naik’s minutes-long feedback to presentation #10 below.

How to summarize such a multifaceted 3 days? Instead of digging deep in a few presentations, I am showing my top 11 presentations below, in order of appearance in the schedule. Please contact the authors directly if you are interest in their work:

1)    Pacing in Real Time Bidding for Display Ads demonstrates the importance of ‘water-level pacing’ your display ads by time of day, over even pacing (Uğurcan Dündar, WU Vienna et al.)

 2)    Real-Time Personalization in Dynamic Environments shows real-time sequential learning got 60% higher CTR than ‘learning, then earning’ (eg AB tests) in the context of news articles, where the value differs for each article and users are anonymous (Hong Deng, Erasmus University, et al., Best Paper Award)

3)    Customer Focused Strategy: A Dynamic Approach leverages deep learning and hidden markov models to show that switching a brand’s B2B strategy scope is associated with lower risks but also lower operating performance of the firms, thus establishing a trade-off between switching or staying put (Sonam Singh, UT San Antonio, et al., Best Paper Award)

4)    Motivating Sustainable Energy Consumption Within Organizations: The Role of Analytics and Optimized Communication quantifies the benefits of better AI forecasting of electricity demand peaks (80%) and effective communication of curtailment advice (20% of benefit) led to about $1M of cost savings per organization and less false alarms, halving the number of curtailment calls from 18 to 9 (Ceren Kolsarici, Queen’s University, et al.)

5)    An Artificial Intelligence Approach For Online Product Reviews and Review Synthesis shows that AI can write reviews to capture the key ideas of all reviews it was trained on, and that people can not distinguish AI-written wine reviews from those written by humans (Praveen Kopalle, Dartmouth, et al.)

6) Persistence of Consumer Lifestyle Choices: Evidence from Restaurant Delivery During COVID-19 estimates the long-term impact of COVID on consumption patterns, and findz that fast and large COVID gains for non-subscription products typically returned to baseline, but that subscription-based categories saw smaller but permanent gains (E. Shin Oblander, Columbia University, and Daniel McCarthy, Emory University)

7)    Coverage and Depth of Promotions revisits the question of price promotion profitability, for the first time combining coverage and depth. For high brand sales, any combination of coverage and depth is unprofitable, while price promotions can be a profit boom for small brands, consistent with my previous research. Moreover, if coverage increases in effectiveness, the brand should increase price promotion depth – and vice versa (Kay Peters, University of Hamburg, et al.)

8)    The Impact of Price Transparency on Retailer Pricing investigates the impact of government-required price transparency in gasoline stations, which does not decrease average price or price spread, but leads to more frequent price changes (Wiebke Keller, University of Tuebingen, et al.)

9)    How advertising and retail drivers grow brand metrics shows which Amazon Ad products and retail metrics (such as reviews, price and new products) drive brand awareness, consideration and sales on Amazon. The key finding is that these drivers have different effects, so advertisers should carefully chose depending on their goals (Vivian Qin, Amazon Ads, et al.)

10) The Role of Customer Mindset Metrics in Optimal Advertising Decisions shows how advertisers can make optimal ad spent decisions with knowledge of the ad responsiveness, sales conversion and potential of mindset metrics, such as awareness and liking (Vardan Avagyan, Erasmus University, et al.)

11) Curbing Unsolicited Advertising Mail: Helping The Environment But Hurting Business? shows that switching from opt-out to an opt-in regulation for unsolicited direct mail, drastically reduces waste without reducing consumer spending (Kristopher Keller, UNC, et al.) A true win-win!