Salespeople frequently face the dynamic screening decision of whether to persist in a conversation or abandon it to pursue the next lead. Yet, little is known about how these decisions are made, whether they are efficient, or how to improve them. We study these decisions in the context of high-volume outbound sales where leads are ample, but time is scarce and failure is common. We formalize the dynamic screening decision as an optimal stopping problem and develop a generative language model-based sequential decision agent - a stopping agent - that learns whether and when to quit conversations by imitating a retrospectively-inferred optimal stopping policy. Our approach handles high-dimensional textual states, scales to large language models, and works with both open-source and proprietary language models. When applied to calls from a large European telecommunications firm, our stopping agent reduces the time spent on failed calls by 54% while preserving nearly all sales; reallocating the time saved increases expected sales by up to 37%. Upon examining the linguistic cues that drive salespeople's quitting decisions, we find that they tend to overweight a few salient expressions of consumer disinterest and mispredict call failure risk, suggesting cognitive bounds on their ability to make real-time conversational decisions. Our findings highlight the potential of artificial intelligence algorithms to correct cognitively-bounded human decisions and improve salesforce efficiency.

When online sellers use AI learning algorithms to automatically compete on e-commerce platforms, there is concern that they will learn to coordinate on higher than competitive prices. However, this concern was primarily raised in single-dimension price competition. We investigate whether this prediction holds when sellers make pricing and advertising decisions together, i.e., two-dimensional decisions. We analyze competition in multi-agent reinforcement learning, and use a large-scale dataset from Amazon.com to provide empirical evidence. We show that when consumers have high search costs, learning algorithms can coordinate on prices lower than competitive prices, facilitating a win-win-win for consumers, sellers, and platforms. This occurs because algorithms learn to coordinate on lower advertising bids, which lower advertising costs, leading to lower prices and enlarging demand on the platform. We also show that our results generalize to any learning algorithm that uses exploration of price and advertising bids. Consistent with our predictions, an empirical analysis shows that price levels exhibit a negative interaction between estimated consumer search costs and algorithm usage index. We analyze the platform's strategic response and find that reserve price adjustments will not increase platform profits, but commission adjustments will, while maintaining the beneficial outcomes for both sellers and consumers.

Jim McHugh is the CEO of Mperativ, the revenue marketing platform that connects go-to-market strategy to the revenue operations engine. Today, there is greater demand than ever for better and more quantifiable connections between marketing budgets and how they drive business outcomes. Even with thousands of marketing technology products available, it has been difficult for them to clearly show the impact of marketing initiatives and spend on revenue outcomes, paving the way for new technologies to rise up to this challenge and create the future of marketing science. This year, we'll begin to see go-to-market teams establish a new realm of marketing science that brings together siloed data to quantify how marketing drives results, setting the stage for continuous, bitemporal data models that will unlock the true benefits of AI and ML for marketing. Connect marketing to revenue and predict business outcomes.

Jakaitis has proven success creating and scaling advanced analytics functions in insurance. Most recently, as Director of Business Intelligence for Carrot Fertility, she founded and scaled the company's data function, including overseeing full-lifecycle product management for data products, such as ROI models, engagement and utilization projections, and financial forecasting tools. Prior to that, she served as the Head of Marketing Science at Acrisure Technology Group and was a founding member of Altway Insurance, where she led a cross-functional team in go-to-market and marketing strategy for products in the insurance technology space. She also contributed industry and community thought leadership in the areas of marketing optimization, authenticity in marketing, applied AI, and algorithmic marketing. "Leah's skillsets in these areas are vital to the rapid growth of PCF as we continue to advance our use of and leverage our data and technology to inform strategy to benefit our partners and clients," said Rob Smith, President, Agency Operations.

CATONSVILLE, MD, September 23, 2019 - Chatbots, which use artificial intelligence to simulate human conversation through voice commands or text chats, incur almost zero marginal costs and can outsell some human employees by four times, so why aren't they used more often? According to new research, the main contributor is customer pushback. The machines don't have "bad days" and never get frustrated or tired like humans, and they can save money for consumers, but new research in the INFORMS journal Marketing Science says if customers know about the chatbot before purchasing, sales rates decline by more than 79.7%. The study authors, Xueming Luo and Siliang Tong (both of Temple University), Zheng Fang of Sichuan University, and Zhe Qu of Fudan University, targeted 6,000 customers from a financial services company. They were randomly assigned to either humans or chatbots, and disclosure of the bots varied from not telling the consumer at all, to telling them at the beginning of the conversation or after the conversation, or telling them after they'd purchased something.

Ironically, people are more likely to trust AI when they don't know it's being used. AI is proving itself superior to human intelligence in an expanding number of fields. That is, except when people know AI is being used. Yes, because in certain human-centric sectors, the performance of artificial intelligence starts to drop off if people are apprised of the involvement of an intelligent machine. In fact, human resistance would seem to be the achilles heel of artificial intelligence, since for all the recent advances of AI technology this resistance is preventing AI from doing its job in areas where human contact and interaction would normally play a central role. This message was brought home most recently by a study published in Marketing Science on September 20, titled "The Impact of Artificial Intelligence Chatbot Disclosure on Customer Purchases."

Ironically, people are more likely to trust AI when they don't know it's being used. AI is proving itself superior to human intelligence in an expanding number of fields. That is, except when people know AI is being used. Yes, because in certain human-centric sectors, the performance of artificial intelligence starts to drop off if people are apprised of the involvement of an intelligent machine. In fact, human resistance would seem to be the achilles heel of artificial intelligence, since for all the recent advances of AI technology this resistance is preventing AI from doing its job in areas where human contact and interaction would normally play a central role. This message was brought home most recently by a study published in Marketing Science on September 20, titled "The Impact of Artificial Intelligence Chatbot Disclosure on Customer Purchases."

Across industries, companies are faced with the challenge of how to sift through the massive amount of readily available user-generated content (UGC) to uncover valuable customer insights. A new study published in Marketing Science describes how to use machine learning to expedite the discovery of customer needs from user-generated content. The details of this approach are published in "Identifying Customer Needs from User-Generated Content" by Artem Timoshenko and John Hauser of the MIT Sloan School. This study is the culmination of the authors' work over the past several years. Applied Marketing Science (AMS) was instrumental in the development, testing, and refinement of the aforementioned algorithm, and has used the process various times to help clients identify consumer insights.

Gauging consumer needs is essential in marketing. Focus groups, interviews and surveys are currently the most common means of gathering this data. But the process can be time-consuming and expensive. The advent of machine learning technology and artificial intelligence (AI) has sparked interest in using the technology to yield valuable insights into consumer wants. Researchers at MIT devised a method of efficiently identifying customer needs from user-generated content (UCG) with machine learning, according to a study published in Marketing Science.

In marketing we are often confronted with a continuous stream of responses to marketing messages. Such streaming data provide invaluable information regarding message effectiveness and segmentation. However, streaming data are hard to analyze using conventional methods: their high volume and the fact that they are continuously augmented means that it takes considerable time to analyze them. We propose a method for estimating a finite mixture of logistic regression models which can be used to cluster customers based on a continuous stream of responses. This method, which we coin oFMLR, allows segments to be identified in data streams or extremely large static datasets. Contrary to black box algorithms, oFMLR provides model estimates that are directly interpretable. We first introduce oFMLR, explaining in passing general topics such as online estimation and the EM algorithm, making this paper a high level overview of possible methods of dealing with large data streams in marketing practice. Next, we discuss model convergence, identifiability, and relations to alternative, Bayesian, methods; we also identify more general issues that arise from dealing with continuously augmented data sets. Finally, we introduce the oFMLR [R] package and evaluate the method by numerical simulation and by analyzing a large customer clickstream dataset.