ABSTRACT:
The two papers opening this issue of the Journal of Management Information Systems present sophisticated studies of the effects of the products of information systems (IS). Both works rely on empirics, carried out on firm theoretical foundations. The first of the articles studies a lasting organizational effect of information stemming from digitalization. By contrast, the second deals with the effects of algorithmic advice on individuals.
Arun Rai, Xinlin Tang, Zhitao Yin, and Steven Du investigate how high- quality information obtained as a result of the digitalization of services in B2B logistics leads to higher customer loyalty. As we have learned acutely during the recent years of supply-chain stresses in the interrelated contexts of de-globalization, autarkic actions, the pandemic, and the move to online marketplaces, this is a highly important subject. We had overreached in aiming for supply-chain efficiencies at all costs, and the costs are now to be paid. Digitalization initiatives bearing on better information across supply chains have much to offer in responding to and compensating for the shocks experienced by the chain participants. The recently assimilated technologies, such as the sensorization assembled into the Internet of Things, provide technological enablers. The authors tease out the mechanisms through which the use value received by the business customers owing to the improved tracking information leads to a heightened exchange value received by the vendors in the form of increased customer loyalty. The researchers identify and test three specific factors that lead from the quality of tracking information to customer loyalty.
Advice rendered by algorithms is the type of advice that will continue to be increasingly utilized. As two major examples, both FinTech and HealthTech are changing with the incursion of the bot-offered advice, mutatis mutandis, of course. We need to learn much more about both the value and the perception of value of the algorithmic advice. Here, Sangseok You, Cathy Liu Yang, and Xitong Li present an empirically-based comparison of the effects of the algorithmic versus human advice on human judgment. Notably, the researchers also investigate how the transparency of the past performance affects that judgment. The theoretical frameworks the authors base themselves on combine the judge-advisor thinking with the cognitive load theory. The roles of trust and cognitive load are thrust to the foreground by the theoretical and the empirical basis of this work. Nuanced findings on the presentation of the past performance of algorithmic prediction encourage such presentation, as long as it does not cognitively overwhelm the user.
The platformization of online markets has called forth copious research in our discipline. The next two papers contribute worthily to this subject. In the first of the papers, Franck Soh and Varun Grover investigate the ecosystems comprising the developers of apps and the mobile platforms. The organizational app developers include a great number of companies eager to place their icon on your smartphone, and the competition is keen, with network effects as potent amplifiers. In this competition, the app developers leverage the innovations continually introduced by the platforms, whether they are authentication features or payment facilities. These platform boundary resources (PBR) are assimilated by the app developers, who compete within the platforms, but also—as the authors emphasize—cooperate among themselves in a process of distributed sensemaking. The developers cooperatively learn to innovate on the platform in the environment of technological uncertainty. Deploying a rich dataset of PBR-based innovations, the authors show the effects of the cooperation aspects of this coopetition. The study is generative: It will help in the future research to understand coopetition in other contexts.
Most markets are characterized by information asymmetry between the seller that knows the attributes of the offering more closely and the buyer that has limited access to this information. The phenomenon can inappropriately affect the selling price of all marketed products, and in the extreme lead to market failure, as the product quality is universally distrusted within the market, which is now assumed to be featuring inferior products. The problem has been established in economics as that of the market for lemons. Ingrid Bauer, José Parra-Moyano, Karl Schmedders, and Gerhard Schwabe address the issue with a blockchain-based resource relying on a multi-party certification of products’ attributes and history. Their principal research question is: “How does multi-party certification impact the market for lemons?” The authors design an experimental game in the context of a car marketplace. They show a positive impact, with a more efficient allocation of the resources on offer and with the blockchain as an appropriate vehicle for such signaling of product attributes.
With the pervasive use of information technology (IT), the negative aspects of its use have come to the fore of research in our discipline. Three subsequent papers take us to this darker side. W. Alec Cram, Martin Wiener, Monideepa Tarafdar, and Alexander Benlian empirically address the impact of algorithmic control on the technostress experienced by Uber drivers. Algorithmic control of workers transcends, of course, the boundaries of the gig economy. Its effects on the well-being, behavior, and productivity of workers need a continuing thorough investigation, redounding to a future regulatory environment. In what is termed sharing economy, we often recognize the effects of technostress in a starker light. The authors dichotomize algorithmic control into gatekeeping and guiding modes, corresponding to the two types of technostressors that represent challenge and threat. They also study the role of control transparency in the effect of the stressors on the growing body of gig workers. The results are sobering and important. The context is quite generalizable to the experience of the actors in the expanding gig economy and beyond.
Techno-invasion, that is, the “always-on“ environment enabled by IT that many of us experience, blurs the differences between work and personal life. Well accepted by some, it is resented by many. May this resenment lead to a deviant behavior of the affected employees? This is the question posed by Yang Chen, Xin Wang, Jose Benitez, Xin (Robert) Luo, and Dechao Li in the next paper. The authors empirically investigate whether techno-invasion impairs the employee‘s individual self-regulation, a psychological condition of being in emotional control of one’s actions. In a nuanced two-stage survey research project, the authors show that techno-invasion impairs the employees‘ self-regulatory strength and may indeed lead to a deviant behavior in the workplace. IT mindfulness is found to beneficially mitigate this effect. The authors surface the conditions under which such deviance can take place and thus alert us forcefully to the need of designing the working lives in harmony with the personal lives of the employees. This is particularly so as we have moved to hybrid workplaces occasioned by the pandemic.
Why do employees take IT-mediated shortcuts? For example, why does a data analyst consider only a subsegment of the data needed to arrive at a conclusion? Why does her colleague not drill down into detailed data before asking the system for a recommendation? Such shortcuts amount to skipping steps needed to properly complete the tasks facilitated by IT. In the next paper, Maryam Ghasemaghaei and Ofir Turel explain these noncompliant (deviant again?) behaviors by taking an ego-depletion perspective and provide an empirical valdiation of this explanation. In this view, the complexity of IT systems affects the employee self-regulation and leads to taking shortcuts. We can clearly see here how this research work expands the preceding results obtained by Chen et al. Here, the speed of IT-assisted decision facilitated by the shortcuts is shown to occur at the cost of performance. The work is another argument for tempering the complexity of the employee-facing IT.
As other scholars and I have reiterated at numerous occasions, our discipline has much to offer to make the world in transition a better place; in fact, to understand the nature of the possible transition trajectories and help define the desirable destinations. The computing world of the present runs on the clouds. The data centers supporting the public, private, and hybrid clouds are collectively significant energy consumers. Greening the clouds or, at any rate, moving them further away from the brown, is the objective of the research presented here by Chetan Kumar, Sean Marston, Ravi Sen, and Amar Narisetty. Within the context of a private, organizational cloud, the authors aim to design a resource allocation and pricing mechanism to balance the use of the cloud’s computing resources. In addition to this benefit, the mechanism allows for the use of a public cloud when private resources are utilized at capacity. Dynamic pricing is deployed as the means to regulate the users’ access to the cloud resources. The optimality of the design is shown formally and illustrated via simulation. The path to more complex, hybrid cloud settings is clear and will certainly be investigated by future researchers.
Since the epoch-making advances in machine learning during the two recent decades, predictive analytics (PA) based on these artificial intelligence techniques has taken its place in organizational IS. As are all components of organizational computing, PA is subject to attacks. One of the specific forms of invasion directed at PA are the attacks on a supervised machine learning (SML) system via adversarial inputs in order to direct the model to offer a wrong advice desired by the adversary. With predictive analytics increasingly deployed in mission-critical decision-making, a potential damage may be significant, and remedies are important. A credit card can go into the wrong hands or a wrong employee can be hired—and all of it at a grand scale. In the next paper, Weifeng Li and Yidong Chai use the design-science approach to present a mechanism for assessing and enhancing the robustness of a PA system grounded in SML vis-à-vis such attacks. The authors demonstrate the effectiveness of their approach and show the path forward toward broadening of its applicability to a wider class of PA systems.
Social media (SM) platforms innovate incessantly, as they compete for the users’ attention and engagement. The attentive traffic is the source of the high-margin advertising profits. All users are equal, but the influencers are more equal than others. It needs to be noted that SM influencing has become a mini-industry with notable income flows to the now numerous participants. Reaching the influencers with new features is becoming a primary target of SM innovations. Are the innovations reaching the targets? Reza Alibakhshi and Shirish C. Srivastava study empirically the effects of one of such notable innovations, a Story feature introduced by several SM platforms. The Story allows the users (and particularly the influencers, of course) to post temporary content, available for a short period (typically, 24 hours) and thus in no need of careful editing. The idea behind Story is for the influencers to draw their followers to the formerly private moments-in-time of the influencers, reported with great frequency and achieving the effect of spontaneity combined with a purported intimacy. The authors draw on a rich theoretical background, including social penetration theory, in their investigation of the effects off the Story feature. The findings are finely-shaded and do not point to uniform benefits garnered by the influencers, followers, and platforms. Moreover, unintended consequences rear their virtual heads. The study expands our theoretical and practical understanding of SM and will serve as an encouragement to platforms to investigate far beyond A/B testing in introducing potentially beneficial new features.