Human vs. Automated Sales Agents: How and Why Customer Responses Shift Across Sales Stages

2023 | Information Systems Research | Citations: 0

Authors: Adam, Martin; Roethke, Konstantin; Benlian, Alexander

Abstract: Customers in sales processes increasingly encounter automated sales agents (ASAs ... Expand

Abstract: Customers in sales processes increasingly encounter automated sales agents (ASAs) that complement or replace human sales agents (HSAs). Yet, little is known about whether, how, and why customers respond to ASAs in contrast to HSAs across successive decision stages of the same sales process. Even less is known about customer responses to HSA-ASA combinations, where both agents assume distinct roles and focus on complementary tasks that are traditionally performed by only one single sales agent. Against this backdrop, this paper explores the influence of increasingly common sales representative (rep) types (i.e., ASA, HSA, and HSA-ASA) on customer decisions across sales stages. Drawing on information processing theory and the literature on hedonic-utilitarian decision making, we investigate customer responses to text-based ASAs from vendor companies in two common early stages of email sales processes (i.e., sales initiation stages) when customers successively decide whether to indicate their initial interest in an offer and then, whether to provide their contact information. Specifically, we conducted two complementary multi-decision experiments, namely (1) a randomized field experiment in a high-stakes sales initiation setting (n = 325) and (2) a subsequent randomized online experiment to complement the real-world insights (n = 408). Our core findings reveal reversing effect patterns of sales rep types across stages: although customers are more likely to indicate their initial interest to HSAs (versus ASAs) because of HSAs’ higher levels of social presence, they are less likely to provide contact information to HSAs because of HSAs’ lower levels of performance expectancy and effort expectancy. We also show that HSA-ASA combinations can be reasonable options for single ASAs, yet contextual features of the sales setting may affect differential customer responses to HSA-ASA combinations (versus ASAs) in each sales stage. Taken together, we uncover shifting effect patterns in customer responses to sales rep types across successive sales stages and shed light on the consecutive underlying mechanisms that explain these shifts. These findings have significant implications for vendor companies seeking to allocate HSAs and/or ASAs effectively across various decision stages in sales processes and beyond. History: Wonseok Oh, Senior Editor and Khim Yong Goh; Associate Editor. Funding: This work was supported by the Center for Responsible Digitality (ZEVEDI) and the German Research Foundation (DFG) [Grant 471168026]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2022.1171. Collapse

Topics: electronic mail social presence decision making recommender system information system use

Methods: experiment randomized field experiment regression analysis method experimental task field experiment

Theories: information processing theory

How synchronous participation affects the willingness to subscribe to social live streaming services: the role of co-Interactive behavior on Twitch

2023 | European Journal of Information Systems | Citations: 0

Authors: Bründl, Simon; Matt, Christian; Hess, Thomas; Engert, Simon

Abstract: Social computing has transformed the content industry. Content providers have ch ... Expand

Abstract: Social computing has transformed the content industry. Content providers have changed from providing merely content to offering social content experiences that allow users not only to passively consume content but also to actively participate. As the latest development, social live streaming services (SLSS) have emerged that enable so-called co-interactive behaviour, i.e., the live interplay between users while co-experiencing content. Yet, it remains unclear how co-interactive behaviour affects the monetisation of social live content. Based on a data set of 868,940 users from 59 gaming channels of the freemium-based SLSS Twitch, we empirically show that co-interactive behaviour has a stronger effect on users’ willingness to subscribe (WTS) than passive and active behaviour. Since we find that a user’s reciprocated interactions also directly increase other users’ WTS, our results call for dismantling the current dichotomy of user participation and social influences in IS research. We suggest a social-experiential perspective that integrates social influences as an inherent part of users’ participation experiences. This novel interactive second-person perspective contributes to research on user engagement and value capture on social content platforms, and it allows SLSS providers to obtain a more precise picture of user value to better develop or adapt their monetisation strategies. Collapse

Topics: participatory design social influence user behavior Twitch application programming interface

Methods: design methodology propensity score matching experimental group descriptive statistic logistic regression

Theories: general systems theory

Cultivating a Leadership Pipeline: Using a Real Options Lens to Understand Executives’ Strategic Staffing Decisions

2023 | Organization Science | Citations: 0

Authors: Campion, Michael C.; Campion, Emily D.; Campion, Michael A.; Bauer, Talya N.

Abstract: This paper adapts real options theory to explain how executives create and maint ... Expand

Abstract: This paper adapts real options theory to explain how executives create and maintain real options portfolios within leadership pipelines. Hypotheses flowing from our theorizing predict that executives often make seemingly risky staffing decisions for leaders who occupy stepping-stone positions. Focusing on their option (future potential) rather than project (current productivity) value, executives laterally transfer leaders in stepping-stone positions frequently, despite it resulting in lower short-term job performance, but often promote these leaders at lower levels of performance and sooner. Once leaders are promoted to destination positions where they may stay indefinitely, executives tend to transfer high-performing leaders more often but not when they are still improving the effectiveness of their current unit. We present evidence suggesting that executives make these decisions to improve other units and maintain a flexible system, possibly recapturing previous investments in developing those leaders. We provide empirical support for our hypotheses using eight years of data in a large retail organization (n = 25,004) where executives overseeing thousands of units made internal mobility decisions. These findings refine real options theory, show that it explains these phenomena better than existing theories, and provide important and immediately usable practical implications for executives. Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2022.1608. Collapse

Topics: task productivity real option IT career productivity human resource management

Methods: statistical hypothesis test descriptive statistic theory development logistic regression focus group

Theories: real options theory options theory

The Cost of Free: The Effects of “Wait-for-Free” Pricing Schemes on the Monetization of Serialized Digital Content

2023 | Management Information Systems Quarterly | Citations: 0

Authors: Choi, Angela; Rhee, Ki-Eun; Yoon, Chamna; Oh, Wonseok

Abstract: Leveraging a combination of analytical frameworks and empirical assessments, th ... Expand

Abstract: Leveraging a combination of analytical frameworks and empirical assessments, this study investigates the effects of wait-for-free (WFF) pricing schemes on the monetization of serialized, digital entertainment content, which has become increasingly pervasive on online platforms. WFF pricing is a strategy in which consumers are given the option to either wait a certain amount of time to acquire digital content at no cost or pay to consume it immediately. We evaluate the extent to which habit formation and presentbiased preferences driven by the consumption of addictive stock affect individual consumers' willingness to wait (or pay) for content, which, in turn, determines the efficacy of WFF pricing. We also examine the conditions under which consumers switch from waiting for free content to instantaneously purchasing content. Our findings indicate that WFF pricing increases the sales of serialized digital content, generating new demand from customers who would otherwise forgo participation in the market. In addition, the pricing design effectively generates sustained profits in the long run. We found that most consumers who initiate a purchase either upon initial market entry or upon switching continue to purchase as new episodes become available. Moreover, the results indicate that as a user accumulates free episodes of a specific series, given extended waiting periods, the likelihood of their conversion from a wait-for-free customer to an instant-purchase customer increases. In particular, WFF pricing effectively augments the willingness to pay of low-valuation consumers as habit formation builds up through time with the free consumption of serialized content. One free episode can elevate the likelihood of consumer purchase by up to 13%. However, as the number of free episodes consumed goes beyond a threshold, the likelihood of conversion decreases. We conclude with a discussion of managerial implications that can help content providers monetize their serialized digital content products. Collapse

Topics: price management consumer behavior advertising management digital good customer satisfaction

Methods: mathematical model robustness check econometric modeling proportional hazards model survival analysis

Theories: behavioral economics

Who’s Watching TV?

2023 | Information Systems Research | Citations: 0

Authors: Clark, Jessica; Paiement, Jean-François; Provost, Foster

Abstract: This work addresses the problem of “user disambiguation”—estimating the likeliho ... Expand

Abstract: This work addresses the problem of “user disambiguation”—estimating the likelihood of each member of a small group using a shared account or device. The main focus is on television set-top box (STB) viewership data in multiperson households, in which it is impossible to tell with certainty which household members watch what. The first main contribution is formulating user disambiguation as a predictive problem. The second contribution is a solution for estimating the likelihood that each individual in a multiperson household watches each TV segment. Kernel theories from the marketing, economics, and sociology literature inform the design of our method. This method learns priors for viewership in single-person households and then adapts them to the specifics of each multiperson household’s viewership history. Finally, we formalize two ad hoc heuristics that are currently used in industry (and research) for estimating audience composition of STB data and conduct a comparative analysis using simulated data, real large-scale viewership data, and a fully labeled panel-based data source. We find that our method has superior performance and practical value. The proposed solution has implications for advertisers, researchers who seek better understanding of TV viewership, and anyone using data generated by shared devices or accounts. A major TV provider has deployed this new method for use in its TV ad-targeting system. No personally identifiable information was gathered or used in conducting this study. To the extent any data were analyzed, it was anonymous and/or aggregated data, consistent with the carrier’s privacy policy. History: Eric Zheng, Senior Editor. Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2023.1204. Collapse

Topics: marketing management logistics management advertising management internet technology

Methods: machine learning experiment simulation longitudinal research descriptive statistic

Theories: social sciences theory economic theory

Using machine learning to create and capture value in the business models of small and medium-sized enterprises

2023 | International Journal of Information Management | Citations: 0

Authors: Costa-Climent, Ricardo; Haftor, Darek M.; Staniewski, Marcin W.

Abstract: Start-ups have revolutionised many economic ecosystems, becoming innovation pion ... Expand

Abstract: Start-ups have revolutionised many economic ecosystems, becoming innovation pioneers around the world. Most are based on data-driven business models, particularly relying on machine learning technologies. However, not all start-ups that use machine learning technologies manage to create and capture value. The existing literature on the use value enabled by information technologies does not take into account the unique capabilities of machine learning. The theory of data network effects offers a promising explanation of how to create value using machine learning. However, it does not explicitly describe how to capture value using machine learning. In contrast, business model theory explains how companies use technologies to create and capture value, but not specifically through the use of machine learning technology. Therefore, this study aims to improve the theo­ retical understanding of the key drivers of value creation and capture in start-ups with business models driven by this kind of technology. Statistical techniques are used in a sample of 122 start-ups to explore the theoretical relationships between these two theories. The analysis reveals the link between specific value creation and capture factors of the two theories, such as efficiency, novelty, and performance expectancy. The study also provides evidence of the need to adopt a co-evolutionary perspective of value creation and capture through the use of machine learning. Collapse

Topics: value creation Twitter social media database system website

Methods: descriptive statistic theory development regression analysis method general multivariate regression model linear regression analysis

Unravelling the Business Value of Artificial Intelligence

2023 | Americas Conference on Information Systems | Citations: 0

Authors: Du, Kui; Li, Gaoyong; Li, Liang; Zhang, Wei

Abstract: This study aims to investigate how firms can prepare for artificial intelligenc ... Expand

Abstract: This study aims to investigate how firms can prepare for artificial intelligence (AI) adoption and how AI creates business values for adopting firms. We propose that, first, leveraging AI for value creation depends on the presence of an enabling resource foundation, consisting of a stack of technological and managerial resources. According to the real options theory, the stronger such a resource foundation is, the more valuable the option it creates for a firm to adopt AI, and consequently the firm is more likely to exercise the option to adopt. Second, we propose that, after adoption, the depth of AI use can have negative impacts on employee productivity, while the breadth of AI use can have positive impacts. We empirically validate these hypotheses based on a sample of 6,854 manufacturing firms in a coastal province in China. Collapse

Topics: artificial intelligence individual productivity organizational value value creation productivity

Methods: survey experiment regression analysis method

Theories: real options theory resource based view of the firm options theory

Online Information Filtering: The Role of Contextual Cues in Electronic Networks of Practice

2023 | ACM SIGMIS Database | Citations: 0

Authors: Fadel, Kelly J.; Jensen, Matthew L.; Meservy, Thomas O.

Abstract: Online forums sponsored by Electronic Networks of Practice are commonly used to ... Expand

Abstract: Online forums sponsored by Electronic Networks of Practice are commonly used to exchange information related to problem solving, but the quality of solutions posted on forums can vary widely. To assist knowledge seekers, most forums offer contextual cues that signal solution quality, such as endorsements by the question poster, experts, or other community members. However, the influence of these cues on information-filtering decisions, particularly when presented simultaneously, remains poorly understood. This study draws on signaling theory to develop hypotheses about how four cue dimensions-source, valence, format, and volume-shape information filtering on an online forum. Results from three choice-based conjoint studies show that some sources are more influential than others and that positive cues weigh more heavily on filtering decisions than negative cues. Results also show that cues tend to be more influential when they employ a more granular format and reflect the input of a larger number of individuals. Collapse

Topics: decision making information storage and retrieval system logistics management website information exchange

Methods: conjoint analysis survey statistical hypothesis test literature study theory development

Theories: signaling theory

A New Take on the Categorical Imperative: Gatekeeping, Boundary Maintenance, and Evaluation Penalties in Science

2023 | Organization Science | Citations: 0

Authors: Fini, Riccardo; Jourdan, Julien; Perkmann, Markus; Toschi, Laura

Abstract: Extant theory suggests that candidates with an unfocused identity—those spanning ... Expand

Abstract: Extant theory suggests that candidates with an unfocused identity—those spanning different categories—suffer from a valuation penalty because evaluators are confused by their profile and concerned they lack the required skills. We argue that unfocused candidates may be penalized for another reason; they threaten established social boundaries. This happens in contexts where evaluators act as gatekeepers for social entities, such as professions. We test how the penalty applied to unfocused candidates varies in an academic accreditation process, a setting where evaluators decide on admitting candidates to an academic discipline and where candidates’ prior performance is observable. We find using data on the 2012 national scientific qualification in Italian academia that the valuation penalty applied to unfocused (multidisciplinary) candidates was most pronounced for the most high-performing candidates. High-performing yet ill-fitting candidates threaten the distinctiveness and knowledge domain of the discipline and are hence penalized by evaluators. High-performing multidisciplinary candidates suffered the greatest penalty in small and distinctive academic disciplines and when accreditors were highly typical members of their discipline. Our theory and findings suggest that the categorical imperative may be driven not only by cognitive or capability considerations as typically argued in the literature but also, by attempts to maintain social boundaries. Supplemental Material: The e-companion is available at https://doi.org/10.1287/orsc.2022.1610. Collapse

Topics: innovation management accounting entrepreneurship IT career productivity

Methods: multivariate statistics robustness check statistical hypothesis test descriptive statistic logarithmic transformation

Reject and Resubmit: A Formal Analysis of Gender Differences in Reapplication and Their Contribution to Women’s Presence in Talent Pipelines

2023 | Organization Science | Citations: 0

Authors: Fernandez-Mateo, Isabel; Rubineau, Brian; Kuppuswamy, Venkat

Abstract: A common explanation for women’s underrepresentation in many economic contexts i ... Expand

Abstract: A common explanation for women’s underrepresentation in many economic contexts is that women exit talent pipelines at higher rates than men. Recent empirical findings reveal that, in male-dominated selection contexts, women are less likely than men to reapply after being rejected for an opportunity. We examine the conditions under which this gender difference contributes to women’s underrepresentation in talent pipelines over time. We formally model and analyze the population dynamics of a generic selection context, which we then ground using three distinct empirical settings. We show that gender differences in reapplication are an important mechanism of gender segregation in some selection contexts but negligible in others. The extent to which gender differences in reapplication contribute to women’s underrepresentation is driven in part by the rejection rate. Higher rejection rates increase the stock of rejected applicants, which in turn enables gender differences in reapplication to disproportionally reduce women’s representation. The results demonstrate that interactions between individuals’ choices on the supply side and screeners’ behavior on the demand side may have consequences for gender inequality, even if we were able to fully eliminate demand-side biases. We discuss the theoretical and policy implications of our research for understanding women’s underrepresentation in talent pipelines. We also interrogate the effectiveness of common interventions focused on encouraging women to apply for opportunities in male-dominated domains. Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2022.1635. Collapse

Topics: crowdfunding crowdsourcing anonymity

Methods: theory development simulation descriptive statistic robustness check system dynamics