An increasing number of studies in service research focuses on human-robots interaction (Henken et al., 2020; Paluc et al., 2021). Within them a few studies offer preliminary insights on the impact of such an interaction on actors’ wellbeing (Odekerken et al., 2020). Much more investigation is needed to get light on actors’ physical, psychological and social well-being (Mele et al., 2021). The aim of this study is to understand how the use of emerging robotic solutions enhance service interactions and foster the patient and caregivers’s well-being, investigating the effects of their use on the patient’s quality of life and the impacts of caregivers' stress and quality of work.
The study adopts an action research approach to promote an experimental project, aimed at implementing the service robot Hiro, a Japanese minimal design service robot, in two research settings (a nursing home and a co-housing) settled in Italy. The research process follows two phases. In the first stage, the authors observe the interaction between HIRO and patients and how the robot can impact physical, psychological and social well-being. In the second stage, the research assesses caregivers' stress and quality of work. The change of practices for both actors is analyzed.
The results show that the introduction of service robots enables patients to reduce their cognitive and socio-emotional decline and mental illness, allowing them to keep the parameters of the state of agitation under constant control, and giving them the opportunity to interact with the robot in case of need (psychological well-being). It’s a first step to reduce stress and also depression and agitation, towards the emotional involvement of the patient in a reactive and conversational interaction (social well-being). The results will also show that the introduction of service robots can nudge patients to avoid a sedentary lifestyle and to carry out daily activities more frequently (physical well-being). Finally, the findings will also show that in reducing patient stress, there will also be a positive impact on the well-being of the caregivers.
A different service interaction arises whereas the robot mediates in the relationship between the patient and the caregiver. Hiro can offer help to the caregivers’ burden and at the same time engage patients in self-care activities. In this perspective the robot becomes a social boundary resource, i.e. the resource that serves as interface for the relationships between two or more actors. In sum, this study contributes to research on the human-service robots’ interactions and well-being implications in the healthcare contexts. The research offers a deeper understanding of the new dynamics that improve service interaction. Finally, this opens up new possibilities for researchers and managers who design service interaction involving technology.

Keywords: service robot, well-being, service interaction

Purpose and value (originality) - This study explores the innovative concept of smart retailing. Smart retail technologies and smart services, in general, offer a wide spectrum of promising resources in terms of service and overall customer experience enhancement.
However, despite the emergence of several research streams on « smart » and « technology retailing », we still notice gaps of various natures in the literature: 1- a fragmented view of the construct across disciplines (information systems, innovation management, customer experience, omnichannel retailing and service marketing); 2- a contrast between the expected benefits of technology and the lack of empirical studies that capture the smart and dynamic nature of emerging innovations in particular; 3- limits in current conceptual models confirming a dearth of scholarly studies leading to calls in various disciplines. In fact, a consensus among academics becomes apparent on the need for more conceptual papers in the marketing discipline especially those contributing new impactful theories and knowledge of computer-mediated environments.

Design/methodology/approach - Adopting a systematic approach seems to be particularly relevant in order to: 1- transform previous findings into a novel higher-order perspective especially in the case of a new and unstructured phenomenon; 2- build, via systematization, necessary coherence and commonalities especially in the case of a fragmented research topic across different disciplines; 3- unveil “big picture” common grounds and connections rather than specific causal mechanisms, and finally 4- contribute to theory development by identifying gaps in the extant research to help create new theory. Following Torraco (2005) methodological steps, after rigorous key word selection on a twelve years period of time and corpus cleaning, a final count (only top ranked journals Q1 Scimago) of over 79 articles has been analyzed using content analysis method.

Findings - This article provides a comprehensive review of over a decade of research around the construct of smart retailing. It contributes to 1- first clarifying this concept, 2- further proposing an integrative framework to better understand its antecedents, underlying process and consequences and 3- outlining the main future avenues for investigation.
More particularly, our study starts with a conceptualization of smartness beyond the dominant technological approach in the literature and its dimensions in regards with products and services. We then develop a multi-level framework for scholars detailing the mechanisms at work within the smart retailing process in terms of drivers, actors, relationships and consequences. finally, we offer a section on the main themes for future research in order to push the field further.

Research limitations/implications - Scholars can capitalize on the identified current and future themes opportunities to prioritize future empirical and theoretical studies. They can thus contribute to advances in service and marketing theory as well as enhance the external validity of research studies through empirical smart retail applications.

Practical implications - Managers can capitalize on the research findings to implement a smart retailing process and thus design more suitable smart service strategies for consumers (e.g. such as convenience, personalization and hedonism) to benefit from (e.g. operational efficiency and relationship value).

Artificial Intelligence in the form of service robots can completely redefine the service experience and especially the tourist experience. Prominent work has mainly been conducted at a conceptual level discussing service robots’ integration in service settings. Yet, recent empirical studies have tried to unveil the impact of service robots’ deployment leveraging online conversations suggesting a positive effect of service robots on customer evaluation of their stay. Nonetheless, little is empirically known about the factors moderating this relationship.
Therefore, with this study we aim to combine three research strands: human-robot interaction, electronic Word-Of-Mouth (eWOM) and big data analytics, to investigate the following research question: What are the moderating effects influencing the relationship between service robots and the customer evaluations of hotel services?
To this end, we leverage expectation-disconfirmation theory, the service robot acceptance model and extant eWOM research and theorizations to disentangle to what extent online review (OR) features conjointly moderate the effect of service robots on customer satisfaction, proxied by OR ratings.
Thus, our conceptual model includes four OR features, namely rapport, robotic discourse depth, submission device and reviewer experience which we expect to moderate the relationship between service robots and customer satisfaction. In particular, we hypothesize that the creation of rapport and the use of mobile submission devices have a positive moderating effect while robotic discourse depth and reviewer experience have a negative moderating effect.
We test our conceptual framework on the entire sample of ORs on TripAdvisor pertaining to 19 hotels - spread in three continents - leading the deployment of service robots in hotel’s operations. Novel OR features are extracted from each OR (e.g., robotic discourse depth) and combined with review (e.g., submission device) and reviewer (e.g., reviewer experience) attributes in the empirical models. Ordered logistic regression is developed to understand the significance and magnitude of each moderating factor.
The results show how the relationship between service robots and customer satisfaction is: positively moderated by the creation of rapport and the use of a mobile submission device; negatively moderated by robotic discourse depth. Reviewer experience is not found to be a significant moderator.
From a theoretical point of view, we enrich the debate at the intersection between online customer behaviors and perceptions about the use of service robots in hotels and their satisfaction with hospitality and tourism services assessing the impact of moderators for this relationship. Furthermore, we contribute to the eWOM and big data analytics literature by extracting an innovative metric for testing the reviewer effort towards service robots, namely robotic discourse depth. From a practical point of view, we provide implications for hotel managers, OR platforms managers, online consumers and tourism and hospitality practitioners dealing with service robots.
This is the first study assessing the conjoint impact of a rich set of moderators on the relationship between service robots and customer satisfaction. As such it makes a relevant contribution to the area at the intersection between human-robot interaction, eWOM and big data analytics.

Service encounters are increasingly dependent on technology (Paluch and Wirtz, 2020; Lariviere et al., 2017). Service firms constantly introduce new technologies to be on the cutting edge, and frontline employees (FLEs) are expected to embrace these technologies and engage more productively to offer better service (Huang and Rust, 2017). However, such infusions are not without challenges. Industry surveys by Reventure report that three quarters of workers are feeling stressed by technology. Academic research confirms these findings (Salonova, et al 2013).

We argue that while technology infusion may provide the means for FLEs to achieve superior service performance, it only serves as a potential facilitator in getting FLEs to engage in their work tasks better. In short, job engagement will be the mechanism which drives service performance. Our first aim is to establish this effect empirically.

Embracing new technologies infused into workplaces can consume significant amount of personal resources for FLEs (Yoo and Arnold 2016). Thus, the extent to which FLEs embrace the technologies and become more engaged may largely depend on their resource endowment to cope with potential resource loss occurring in the face of stressful conditions. Conservation of resource theory (COR) suggests that those better endowed with resources are less affected by resource drain (Hobfoll 2011). Thus, an FLE’s perception of the new technology as either compatible to their work or incompatible may influence how they embrace the technology to engage in their tasks. Underpinned by COR we examine task-technology fit (techno-fit), the degree to which a technology helps individuals to perform tasks (Goodhue and Thompson 1995); and technology stress (techno-stress), a negative psychological state associated with the use of new technologies (Ragu-Nathan, et al 2008); as resource capacities that act to buffer or expose the depleting effects of embracing the new technology. Our second aim is therefore to establish the boundary roles of techno-fit and techno-stress in the indirect relationship between technology infusion and FLE performance, through job engagement.

We conduct two studies; a scenario-based experiment and a field survey of FLEs who work alongside technology to offer customer service. In the first study, FLEs (n=175) were recruited through Prolific and exposed to a written scenario of a bank teller working alongside an interactive robot. Respondents were exposed to conditions of high and low levels of technology infusion. Their level of engagement and service performance were measured. Data confirmed that engagement mediated theeffect of technology infusion on service performance.

Next, a field survey was conducted in collaboration with a market research firm in Finland. Respondents (n=170) were FLEs from telecommunications and financial services industries and worked alongside CRM systems, and chatbots providing customer service solutions. Data confirmed that techno-fit had a significant reinforcing effect on the aforementioned mediating effect, whereas techno-stress had no significant moderating effect on the mediating mechanism.
Using COR theory, this study demonstrates the importance of technology infusion into service frontlines by examining the mechanism (why) and the conditions (when) under which technology infusion contributes to FLE performance.