Abstract
Service robots continue to permeate and automate the hospitality sector. In doing so, these technological innovations pose to radically change current service production and delivery practices and, consequently, service management and marketing strategies. This study explores the various impacts of robotization in the sector by offering one of the first empirical accounts on the current state-of-the-art of service robotics as deployed in hospitality service encounters. The results suggest that service robots either support or substitute employees in service encounters. They also offer hospitality businesses a novel point of differentiation, but only if properly integrated as part of wider marketing efforts. Finally, the automation of tasks, processes, and, ultimately, jobs has serious socioeconomic implications both at the microlevel and macrolevel. Consequently, hospitality executives need to consider where and how to apply robotization to strike a balance between operational efficiency and customer expectations. Displaying ethical leadership is key to reaping the benefits of the robot revolution.
Introduction
We live in an era of rapid change whereby the dynamic, highly competitive business environment, along with ever-changing customer preferences, and the constant emergence of new technologies force organizations to continuously reorganize and reinvent themselves. For instance, innovations in information and communication technology (ICT) have already changed the way we look after ourselves (Combs et al., 2016), trade (Gomber et al., 2017), wage war (Weinberger, 2013), and spend our leisure time (Buhalis & O’Connor, 2005). Most recently, a particular technological innovation, service robotics, has hit the headlines, promising to automate much of the work around us (Harari, 2017). Research by McKinsey Global Institute (Manyika et al., 2017) estimates that 375 million workers (14%) worldwide will need to be retrained for new roles as the automation of labor progresses in the coming decades. However, experts believe that not all sectors will be affected the same way. Industries that rely heavily on repetitive, manual labor are expected to be among the first to feel the impacts of impending automation (Huang & Rust, 2018).
The service sector provides many examples of labor-intensive tasks ripe for automation; call center agents, retail salespersons, receptionists, and taxi drivers are just some examples of occupations that rely largely on systematic, unskilled labor (Huang & Rust, 2018). Particularly people-dependent is the hospitality industry (Melissen et al., 2014). Restaurants, cafés, bars, pubs, and hotels of all types depend on an armada of human laborers. Be it waiters, baristas, maître d’s, chefs, kitchen porters, bellboys, or housekeepers, the global hospitality industry would not exist as it is today without people. Accordingly, Noone and Coulter (2012) argue that this dependence on human labor makes hospitality an increasingly appealing sector for applying emerging technological innovation.
However, little is known about the theoretical or practical impact of service robotics on hospitality management and marketing. Studies have begun to conceptualize and predict the impact of robotics. For example, Ivanov and Webster (2019b) alongside Li et al. (2019) have studied service robotics in relation to employment. In addition, Tung and Au (2018) as well as Lu et al. (2019) have researched the impact of service robotics on customer experience. However, there are few empirical studies to be found (Ivanov et al., 2019). As such, this study examines the use of current state-of-the-art service robotics in the hospitality industry and aims to better understand how this technology can transform service operations. It focuses particularly on the role of service robots in relation to service production and delivery. This study seeks to answer the following research questions:
The findings of this study advance academic discourse on how service robots are used in hospitality to produce and deliver customer services. In doing so, this study provides much-needed empirical evidence in this field. It will allow hospitality researchers and practitioners to better understand how service robots are transforming service encounters. The results reveal the management and marketing strategies used for innovative, automated service offerings. In addition, they provide an up-to-date conceptualization of the different roles robotics technology plays in hospitality service encounters.
Service Encounters in Hospitality
Services research traditionally falls within two main paradigms: service marketing and service management (D. Bowen, 2016). A key interest in both paradigms is the way that services are produced and delivered to customers. Collectively, these processes are referred to as service encounters (Lin & Mattila, 2010; Voorhees et al., 2017). Bell (1973) viewed service encounters as a “game between persons.” On the other hand, Surprenant and Solomon (1987) define service encounters as the “dyadic interaction between a customer and a service provider” (p. 87). Extending these views, Voorhees et al. (2017) note the chronological nature of service by dividing it into precore service, core service, and postcore service encounters. Lillicrap and Cousins (2010) illustrate what this may mean in the practical context of hospitality. For example, in à la carte restaurants, the production and delivery of services can be broken down into sequential encounters: taking bookings, greeting and seating, taking orders, serving, billing, taking payment, and clearing. Similarly, in hotels, Ball et al. (2011) note that the sequence of service encounters generally includes placing reservations, checking in, consuming auxiliary services, staying overnight, eating breakfast, and checking out.
For decades, research into service encounters has focused on the social interactions between people. However, recent research suggests that as service organizations increasingly turn to technological innovations, the way we consider and consume services is changing (Larivière et al., 2017; Ostrom et al., 2015). Service encounters are increasingly enhanced by and delivered using technology (Ostrom et al., 2015). As such, the significance of technology-mediated customer contact is growing (Froehle & Roth, 2004). For example, the National Restaurant Association (2016) has reported that tableside self-service technologies are becoming commonplace in restaurants across the United States. In addition, a study by UKHospitality (2017) found that restaurant-goers in the United Kingdom have increasingly turned to mobile applications when looking for somewhere to eat. In the context of accommodation, Buhalis and Leung (2018) noted an increased tendency to book and pay online, check in using self-service technology, and order room service via mobile applications.
Hospitality Service Robotics
In the context of hospitality service encounters, service robotics is one of the most transformative technological innovations to date (Ivanov & Webster, 2019a). Fueled by advances in electric and mechanical engineering and computer science (e.g., increases in raw computing power, agglomeration of unprecedented amounts of data, novel techniques, and processes such as machine learning or deep neural networks), robots have moved from the confines of factories to dynamic human environments (Ivanov et al., 2019; Wirtz et al., 2018). In particular, recent years have witnessed accelerated development in service robots for the hospitality industry (J. Bowen & Whalen, 2017; Murphy et al., 2017). These developments include robots that cook complex meals and robots that serve customers (J. Bowen & Morosan, 2018). In the United States, California-based creator has developed a burger robot that can fulfill up to 120 orders an hour (Troitino, 2018), while Café X has created robot baristas that can produce up to three beverages in 40 s (Canales, 2018). In Japan, several hotels have replaced many frontline service staff with interactive robots (Osawa et al., 2017). In the United Kingdom, the food technology sector, most notably restaurant robotics, is seeing an increasing amount of interest and investment (Dobberstein, 2019).
Ivanov, Webster and Berezina (2017) and Ivanov and Webster (2019a, 2019b) attribute the recent increase in hospitality service robotics to the following reasons: increased cost-effectiveness, better resource utilization, more accurate demand prediction, better quality control, improved process management, and the removal of human error. According to J. Bowen and Morosan (2018), however, the primary reason for this increase across most markets is the shortage of labor. For example, in Japan, the increased proportion of elderly citizens, falling birth rates, strict immigration policies, and a significant predicted growth in service demand have forced hospitality operators to utilize emerging technologies (Schneider et al., 2018). Frey et al. (2016) observe a similar trend in most other developed nations. They suggest that, in the near future, leveraging service robots will play a key role in ensuring steady productivity and growth in gross domestic product (GDP; International Federation of Robotics, 2018).
Robotics in Service Encounters
Wirtz et al. (2018) define service robots as “system-based autonomous and adaptable interfaces that interact, communicate and deliver service to an organization’s customer” (p. 909). This definition encapsulates what sets service robots apart from other technologies in hospitality service production and delivery. For example, unlike self-service kiosks or preprogramed tablets, service robots can react and adapt to their environments more flexibly (Ivanov & Webster, 2019b). Often, they can gather input data using sensors, analyze these data instantly, formulate a plan, and immediately execute decisions using physical actuators (Ivanov & Webster, 2019a). In addition, more complex systems can subsequently learn from previous interactions, adapt, and optimize their future behavior accordingly (Belanche et al., 2020). For example, a service robot that serves food and drink must continuously analyze and react to its environment to avoid obstacles. While doing so, it must acknowledge various social factors (e.g., customers and employees) in the near vicinity. This results in human–technology interactions previously unseen in hospitality service contexts (De Keyser et al., 2019).
Larivière et al. (2017) argue that, in general, technology has played two key roles in physical service encounters. First, it has supported service employees by providing them with more efficient data processing and analysis capabilities. This enables them to understand customer requirements better, thus improving job and customer satisfaction (Marinova et al., 2017). These advancements have alleviated employee workloads by performing repetitive and monotonous tasks such as dealing with routine orders or transactions. This frees employees to focus on more complex tasks that require problem-solving or emotional intelligence (Huang & Rust, 2018). Second, technology has automated service encounters and replaced employees in a sequence of tasks or substituted them completely (Mathath & Fernando, 2015). According to Rosenbaum and Wong (2015), the self-service systems noted previously, such as check-in kiosks at hotels or airports, are examples of this. Although previous research has extensively discussed the use of technology in services, Wirtz et al. (2018) suggest that current academic literature on the use of robotics in service encounters is still in its infancy.
Due to recent advances in both hardware and software technologies, the robotization of tasks that were previously considered impossible to automate are now a reality. This has fundamental implications on hospitality operations, management, and marketing (Ivanov & Webster, 2019a; Murphy et al., 2019). In addition, leading thinkers and technologists predict that this trend will continue to accelerate (Bughin et al., 2019). As such, greater scholarly attention should be paid to the ways in which service robotics will transform service production and delivery in hospitality service encounters.
Method
Despite increased interest in hospitality service robotics from researchers and practitioners alike (Murphy et al., 2017), applications of robotics in actual hospitality service settings remain relatively few and far between (Ivanov & Webster, 2019a). Due to this, an exploratory qualitative approach was deemed suitable for this study. Observations and semi-structured interviews were adopted as the method of inquiry. This was due to their ability to produce rich data from a limited number of cases and participants (Brewer, 2000). Data collection was carried out from July to December 2018. After extensive research, Japan and the United States were identified as the most appropriate locations to carry out this study due to their leading positions in deploying hospitality service robotics (International Federation of Robotics, 2017). A list of key organizations was collated, including newly founded companies and incumbent multinationals, to represent front-of-house and back-of-house robots in various hospitality contexts, including hotels, restaurants, coffee shops, and bars. A total of 14 organizations were contacted to arrange site visits. These visits consisted of on-site observations and interviews with senior executives. As some businesses had several venues, observation access for 28 sites was granted (14 in Japan and 14 in the United States) as shown in Table 1. However, only six of the 14 organizations contacted were able to arrange a formal interview, quoting issues of scheduling with key personnel. This was because a purposive sampling strategy targeting what Aguinis and Solarino (2019) call “elite informants” was adopted. In selecting interview participants (Table 2), the key criteria were that informants were up to date with current state-of-the-art service robotics and had a comprehensive understanding of how and why the technology was used in their organization. When studying emergent phenomena, Bogner and Menz (2009) stress the importance of targeting experts for their relevant interpretive knowledge, referred to as “know-why,” and their procedural knowledge or “know-how.” Senior executives of robotized hospitality businesses, founders of hospitality robotics companies, hospitality technology investors, and change management executives were considered experts as the agents designing and/or overseeing the implementation of service robotics in hospitality.
Characteristics of Observation Locations.
Demographic Characteristics of Interview Participants.
To mitigate the lack of access to elite informants in Japan and the United States, a second round of interviews (N = 7) was carried out in the United Kingdom. As per this study’s purposive sampling strategy, these targeted experts fell into two groups: companies that develop robotics for hospitality services in Japan and the United States and further afield or companies that operate hospitality businesses in Japan and the United States and are actively seeking to implement service robotics in their operations. After the additional seven interviews, no new themes emerged. As such, saturation was deemed to have been reached, and data collection was halted (Aguinis & Solarino, 2019).
To gain a better understanding of the service robotics currently in use, as well as the potential benefits and/or challenges of robotizing hospitality service encounters, data collection began with the observational phase. Observations were semistructured and followed an observation guide but allowed deviation from the script and additional comments to facilitate thick description (Denzin, 2001). Due to the theoretical focus of this research, an observation guide was developed using Lillicrap and Cousins’ (2010) service sequence model. This model divides the delivery of hospitality services into distinct encounters. The service sequence model was chosen for its broad applicability to a myriad of contexts including food and accommodation services. The observations focused on establishing patterns of behavior within five key areas of service production and delivery: (a) meet and greet, for example, what happens when customers enter the establishment and how they are seated or welcomed, (b) ordering/check-in, for example, who takes the order and who deals with check-in requests, (c) eating, clearing, and room service, for example, how and by whom the food is served, what happens if there is an issue with the food or customers wish to order something else, and how room service is ordered and delivered, (d) paying/check-out, for example, how payments are taken and gratuity policies, and (e) prearrival of guests, for example, what happens after guests leaves, whether there is a queue and, if so, how this is managed. In addition to these five areas, several contextual factors including the position of robots within the servicescape as well as employee and customer appearance/demeanor were also noted.
On average, the observations lasted for four hours. This was to capture a wide range of customer–robot encounters over a single service period (breakfast, lunch, or dinner) or peak service time (check-out, check-in) if possible. To minimize potential bias caused by the observer, also known as the Hawthorne effect (Jones, 1992), a covert approach of a complete observer (Kawulich, 2005) was adopted to ensure customer interactions with service robots were not influenced. In businesses focused on food service, the observer was seated incognito among those being observed. In accommodation businesses, observations were made from the lobby or the lobby bar. As suggested by Lincoln and Guba (1985), a systematic approach to member checking was followed at the end of each observation session. Here, the observer debriefed a ranking operations team member to discuss fieldnotes and seek clarification on or confirmation of instances the observer was unsure about. A subsample (15%) of locations was visited twice on different days at different times to establish consistency through data triangulation (Creswell, 2007).
The observational data were used as the basis for the next data collection phase: semistructured interviews. These went in-depth into the current applications and implications of using service robots to produce and deliver service in conjunction with (or instead of) human employees in hospitality. Questions were created using the observations made while also considering current gaps in research (e.g., Ivanov & Webster, 2019a; Murphy et al., 2017). The questions were designed to explore the value of integrating service robots into service processes by identifying the most suitable tasks for automation in hospitality (e.g., friction points, repetitive or manual tasks). They also highlighted the impact automation may have on hospitality employment (e.g., training, retention, retaliation) as well as the impact on management and marketing (e.g., operations, service design, brand positioning, profitability). On average, the interviews lasted for 41 minutes. All 13 interviews were conducted in English, although one interview was partly mediated by an external translator (English to Japanese). The translator occasionally clarified interview questions to one participant. All interviews were recorded, transcribed by hand, and anonymized.
A thematic approach that built on a priori themes or categories, as used in previous literature, was adopted for data analysis (Creswell, 2007). First, all interview transcripts and field notes (approximately 50,000 words of observation data and 65,000 words of interview data) were printed. These data were then fully read, relevant or interesting sections were marked, and notes were made in the transcript margins (Huberman & Miles, 1997). Afterwards, formal coding was conducted. Following suggestions by Strauss and Corbin (1990), the data were coded in two distinct stages: open coding and axial coding. In open coding, disparate themes were identified and labeled. In total, 59 unique codes were extracted. During this process, some of the codes began to merge together and show a hierarchical relationship (Creswell, 2007). In axial coding, these were organized thematically through building on the a priori categories put forward by Larivière et al. (2017). These categories assume that technology can primarily play either a supportive (Support) or a substitutive (Substitute) role in physical service encounters. However, the coding process was not “prefigured” as the authors remained open to additional emergent categories (Crabtree & Miller, 1992). Indeed, in addition to Support and Substitute, two further primary roles were observed. The first was differentiating service encounters (Differentiate), while the second was improving tangible and intangible service offerings (Improve).
In line with Lincoln and Guba’s (1985) approach to member checking, the resultant four major themes or code families (Creswell, 2007) and their descriptions were sent via email to 25% of interview participants for confirmation and comments. Following Lincoln and Guba (1985), the authors of this study were interested in whether the analysis represented and reflected participant views or if any key themes were absent. Based on suggestions from two key informants (P7 and P10), Improve was divided into Improve and Upskill.
To test analytical consistency across the five established themes, an intercoder reliability check was carried out. According to Mayring (2014), calculating intercoder reliability is considered good practice to account for subjectivity and minimize bias, especially when qualitatively analyzing underresearched or emerging phenomena. At this stage, the peer review (Creswell, 2007) was carried out using two independent coders. A random sample (18 cases across all five themes) of interview and field data were sent to two researchers from different backgrounds and demographics (Coder 1: Female, 30 years old, Computer Science; Coder 2: Male, 27 years old, Geography). Both independent coders had limited knowledge of the research phenomenon and the current study. Percent agreement and Cohen’s kappa were chosen as measures of intercoder reliability. As discussed by Mahmud (2012) and Roaché (2018), these measurements are two of the most commonly used methods of establishing intercoder reliability in exploratory qualitative research. As illustrated in Table 3, good (>0.61) or very good (>0.81) agreement was established across all major themes against both measures with both independent coders (Landis & Koch, 1977).
Intercoder Reliability Check Results.
Roles of Service Robots in Hospitality
The following section illustrates the use of current state-of-the-art service robotics in hospitality service production and delivery. In accordance with previous research on the use of technology in service encounters (D. Bowen, 2016; Larivière et al., 2017), two principal roles of technology were observed: supportive automation (Support) and substitutive (Substitute) automation. In addition, three new technology roles specific to service robotics were also found: automation for novelty (Differentiate), automation for better products (Improve), and automation for better jobs (Upskill). Quotes from in-depth interviews were used to illustrate the roles of automation in service operations. Figure 1 shows a breakdown of the five themes and Table 4 presents where these types of automation were observed.
Roles of Service Robotics in Service Encounters.
Roles of Service Robots in Observed Locations.
Support
When technology is used in tandem with human capabilities, it can effectively enhance service encounters (D. Bowen, 2016). This supportive automation was found in 16 (57%) of the locations observed. Service robots worked particularly well when used to perform relatively simple, well-defined customer-facing tasks. These included taking orders, dealing with payments, providing more information about products, managing restaurant queues, and performing hotel customer check-ins. They also performed well when completing repetitive operational back-end tasks that require precision. These included slicing vegetables, spreading sauce, seasoning and grinding meat, stretching dough, frothing milk, and heating ingredients to a specific temperature. In general, the technology seemed to work in harmony with employees; both added unique value to the service encounter. For example, the technology performed repetitive tasks with great precision while employees focused on dynamic tasks that required problem-solving skills or emotional intelligence.
However, not all observations were positive. At times, the robots seemed to hinder employees. For example, one table-clearing robot roamed around a restaurant collecting empty plates. However, it had not been programed to deposit these anywhere and so continually carried the same plates. Employees had to chase the robot to empty its tray, and sometimes, the robot would not stop at all. Once, this resulted in an employee’s toes being run over. Similarly, several instances where customers ignored a robot maître d’ and entered without approval were also observed. In these cases, a human employee had to step in and explain the service process step by step. They often had to take the customer back to the robot to complete the check-in procedures. One participant noted the following: For the most part, robots work well for what they are intended. However, sometimes they require additional assistance. For example, we have a cleaner robot that cleans the lobby, but at times, the floor might still be dirty even after it has finished, especially in the corners and near the edges. So, even though the robot helps, these kind of areas need to be rechecked by humans. (P2: Manager, Japan)
Substitute
As Larivière et al. (2017) suggest, technology may also replace employees altogether in service encounters. This substitutive automation was observed to varying degrees in 12 (43%) of the sites. In these cases, service robots were used to carry out an entire service experience (i.e., the full sequences of service encounters). Examples included an autonomous bar manned by a virtual bartender and a coffee shop manned by a robot barista where ordering, serving, and taking payments were managed without any human involvement. In addition, there was a robotized hotel where customers could check-in and out, store their luggage, have their luggage taken to their room, order room service or taxis, and control the room through interacting with robots.
Although most service encounters were observed to be successful, it was evident that the more automated elements the service process included, the more chances there were of technical hiccups. For example, the payment system malfunctioned several times at the autonomous bar and the coffee shop. This halted the service process, and employees had to resolve the problem. In the autonomous bar, customers had to push a button to contact an employee. At the coffee shop, an employee was specifically appointed to monitor the floor using surveillance cameras and resolve any problems or service failures. As elucidated by one participant, For us, the technology is there to do all the heavy lifting. It allows us to deliver consistent service. But that alone is not enough—it’s important to have employees on duty to detect and resolve any issues that arise. This is non-negotiable. (P5: Founder, United States)
Differentiate
Service robots are still a relatively novel sight in service settings. As such, they provide an opportunity for businesses to stand out (Mest, 2017; Murphy et al., 2019). There was strong evidence of this both in Japan and the United States. Automation used specifically for novelty was observed in 18 (64%) locations. Interestingly, this was done both intentionally and unintentionally. One interviewee stated the following: “People talk about their unusual experiences more than ordinary ones and that generates added interest which leads to business growth” (P1: Manager, Japan). Another interviewee stated the following:
We never tried to position ourselves as a super trendy, high-tech restaurant. We simply focused on making the best product possible as affordable as possible. Robotics was an obvious choice. The publicity just happened, people started talking and taking pictures. (P3: Founder, United States).
The desire to capitalize on the novelty factor of these technologies was especially evident from where they were placed: Robots were, without exception, given the most visible location and would often be the first thing customers see when entering. Naturally, this attracted public interest. In many instances, people would enter the establishment just to take a photo with a robot. In addition, robots were often explicitly featured in promotional materials (e.g., posters, signs) and embellished with hats, aprons, name tags, and other accessories to make them appear more human (and perhaps more picture-worthy). In the name of novelty, one restaurant had gone as far as to install a robot personal assistant on every table. Customers could interact and have simple conversations with the robot while waiting for their meals. The integration of robots as part of the servicescape (Bitner, 1992) bears testament to the role of emerging technologies as points of differentiation (Liu & Mattila, 2019).
However, adopting robots simply for their novelty may not be sustainable in the long run. As one participant noted, “The [autonomous] café had great impact and was well received when it first opened, but the interest quickly died down. There were no repeat customers, which made it difficult to sustain business” (P2: Manager, Japan). Similar narratives were noted across businesses with multiple sites, strong brand identities, and well-established customer bases. In these instances, simply implementing a novel technology did not always have a lasting impact. Another participant stated the following: We’ve tried initiatives like that before, but with limited success. Like ordering your food on iPads. We spent a fortune on that. But people didn’t really go for it at the top end, they wanted human interaction. So to be accepted, service robots need to mimic that. And do it very well. It needs to be consistent, not just something you do for the buzz. (P10: CEO, United Kingdom)
Improve
Automating service processes may improve process management, quality control, demand prediction, and create cost savings (Ivanov et al., 2017; Ivanov & Webster, 2019a; Noone & Coulter, 2012a). As such, the utilization of service robots was, in many cases, observed to create consistent, affordable, hospitable, and healthier service offerings. This type of automation was observed in seven (25%) locations. For example, delegating certain tasks to robots (e.g., clearing tables and delivering used plates to the kitchen) allowed employees to spend more time with customers. One participant remarked, “I think we’ve actually increased our hospitality by using tech” (P1: Manager, Japan).
In addition to creating a more attentive service, service robots were used to produce and serve higher quality food at lower prices. This benefit was noted by the following participants: “We wanted to make nourishing, healthy food affordable. So, we decided to use robotics to do just that” (P3: Founder, United States), “While [the] fast casual [sector] spends on average 20% on ingredients, for us it’s more like 40%” (P4: Founder, United States), and “Cross-contamination is a huge problem in commercial kitchens. Our technology helps businesses alleviate that” (P9: Developer, United Kingdom). A further participant noted that We saw an opportunity to use robotics in restaurants and the hospitality industry principally to do two things: to improve the quality of the product offered to the consumer and to reduce food waste by having much smarter [predictive] ordering and management systems. (P8: Developer, United Kingdom)
Upskill
As discussed by D. Bowen (2016), the increasing use of technology in services may change the role of employees in service encounters. As well as improving service offerings, automation technologies were observed to change what it means to be an employee engaged in hospitality service encounters. For example, in several businesses (29% of those observed), waiters, receptionists, baristas, or cooks adopted new labels for their service roles. These included product specialist, concierge, burger consultant, guide, garde manger, and chef technician. One participant stated, “The skillsets of specialists are fundamentally different than the skillsets of traditional workers” (P2: Manager, Japan), while another expressed a similar view: Our view is very much: use humans to do human specific jobs, and let’s try and automate the mundane tasks. That creates an environment where you have more interesting jobs for the people in the restaurants, and you’re creating another layer of employment for people in maintenance, design, and operations of the equipment. So effectively we’re up-skilling the required labor in restaurants. (P8: Developer, United Kingdom)
The increased operational efficiency gained from service robots allowed businesses to allocate more time and resources to improving individual employee competencies through training, development, and internal promotion. As one participant remarked, Similar to how Google lets its employees use 10 percent of their time to pursue personal projects, we let our staff spend around 5 percent of the time they’re paid just to study. We even have plans for a book budget. And as opportunities arise, staff are offered chances to move onto more demanding tasks, like repairing the machines. (P4: Founder, United States)
Implications of Service Robots on Hospitality
The five roles of service robots discussed (Support, Substitute, Differentiate, Improve, and Upskill) illustrate how technological innovation can transform operations, management, and marketing for hospitality service offerings. In addition, evidence suggests that this has profound implications for people management practices (see Figure 2). The following section discusses the impact and implications of using robots for service production and delivery in hospitality service encounters from both practical and theoretical viewpoints.
Impacts of Service Robotics on Hospitality.
Management of Operations
Hospitality executives should carefully consider the level of automation they require and where this should be implemented in their service process (Ivanov et al., 2017; Ivanov & Webster, 2019b). Although some flexibility in the use of this technology was observed, the division between supportive and substitutive automation was clear. Some businesses had opted to automate as much of their service processes as possible, whereas others used automation to modify a specific part of their service production or delivery. The degree in which service robots were used to automate service operations seemed dependent on the desired business model. Operators aiming to provide a more affordable alternative to service offerings from their competitors cut costs through reducing the need for human labor. While this allowed for greater control over the service process, it also led to highly standardized, streamlined service encounters and scripted customer experiences. This form of service primarily added value for business travelers seeking convenience or customers buying takeaway drinks. Operators aiming to appeal to a broader market used service robots predominantly to support employees by alleviating their workload and addressing common pressure points (e.g., flagging a waiter to order more or to pay the bill). Ultimately, this method benefited all stakeholders.
The contradictions of introducing technology can be illustrated by the high-tech/high-touch dichotomy (Brochado et al., 2016). On one hand, service robots may support or substitute employees by serving customers tirelessly in multiple languages or preparing food with consistent precision. On the other hand, service robots do not currently cope well with uncertain or dynamic conditions (Tung & Au, 2018). As soon as customers deviate from the prescribed customer journey, the systems fail. In other words, the more touchpoints automated, the more possibilities there are for things to go awry. Therefore, it is imperative that operators develop recovery strategies specific to service robots (Tung & Au, 2018; Zhu et al., 2013). Based on the findings discussed, service failures were usually handled on an ad hoc basis with no clear strategic direction or oversight. While this may be sufficient in the early adoption stage of service robots, a more systematic approach is required as automation technologies expand.
Marketing
It is imperative that service robotics fit well with the desired brand image of a business (Kuo et al., 2017) and should enhance customer perceptions of the company (Liu & Mattila, 2019). As such, the marketing strategy used should emphasize the newness and innovativeness of this technology. Moreover, this should be consistent across all marketing materials and channels (Arruda, 2016; Liu & Mattila, 2016). In the United States and Japan, the leveraging of service robots in hospitality marketing was somewhat mixed. There were many examples of (relatively young and small) companies that built their entire value proposition around automation. As such, these companies had a well-defined and consistent presence across multiple marketing channels, including social media and mobile applications. However, there were also companies (often comparatively old, large, and entrenched) that seemed to adopt service robots as a quick fix or to appear trendy and innovative. Unfortunately, this often resulted in poorly aligned marketing communications and created a mismatch between old and new values as well as service expectations.
Targeting the right segment is equally important to align marketing communications. For example, millennials are often seen as tech-savvy and on the look-out for new experiences (van den Bergh & Behrer, 2016). In general, patrons consuming robotized service offerings represented two key segments: young professionals visiting alone or in small groups and young families accompanied by small children. Both groups had several things in common: They wore trendy clothes (often donning symbolism influenced by science or science fiction such as NASA and Star Wars), had technological gadgets with them (e.g., smart watches, gaming consoles, electric scooters or skateboards, go-pro cameras), and often paid using contactless or mobile pay. In many cases, it seemed that visiting a robotized service provider was a logical extension of who these people were as individuals.
While integrating service robots into operations may generate added interest in the short and medium term, the lack of a formal approach to service robot integration will likely limit the marketing potential of this technology. However, newly founded service companies can effectively map service robots directly into their marketing strategies. For companies with well-established brand identities, it may be better to create a new branch of offerings, perhaps aimed at the next generation of consumers, instead of trying to force emerging technologies into existing models.
People Management and Social Responsibility
Delegating the most routine tasks to service robots allows businesses to increase their operational efficiency through more consistent, standardized service offerings. However, according to Huang and Rust (2018), it may also offer executives an opportunity to allow employees to focus on more complex tasks in service production and delivery, particularly tasks that require creativity, problem-solving, or empathy. The hospitality organizations observed indicated a definitive trend in this direction. For example, in terms of creativity, line cooks working alongside robots tended to focus more on the presentation (i.e., plating) of food rather than the more arduous, mechanical tasks of preparing and cooking ingredients. Similarly, waiters primarily oversaw robotic systems carrying out routine customer service tasks and only stepped in when unpredicted behavior (e.g., customers wanting to place a special order or issues with payment systems) occurred. Finally, where the service process retained a human presence, the service robots increased the depth of attention each customer received. Employees often went above and beyond to serve customers and to educate them on the product/service, the local area, or the technology.
The increasingly sophisticated level of service automation available to businesses presents hospitality executives with an ethical dilemma: should they aim for higher profits by substituting employees with robots? Or, should they use technology to support existing employees to improve working conditions, and ultimately, the service offering? Corporate social responsibility (CSR) can be defined as a business’s responsibility to integrate economic, social, and environmental concerns into their strategies (Melis et al., 2009). Modern CSR efforts have focused on tackling climate change and addressing human rights issues (KPMG, 2017). With the rise of automation, another aspect of CSR may need to be emphasized: business’s responsibility to customers and employees. The findings discussed here suggest a subtle move toward this. Due to the potential for service robots to displace employees, hospitality operators are placing greater emphasis on career progression through internal promotion and by advocating employee development through various learning schemes (e.g., specific budgets allocated for personal learning materials).
Nurturing the professional growth of employees is a classic concept in human resource management. For example, the social exchange theory (see Blau, 1964; Cropanzano & Mitchell, 2005) has long since asserted that investing in employees through training, development, or career progression can lead to increased performance and higher retention levels (Nerstad et al., 2018). However, while this works well in theory, this may not be the case in practice as the service sector, particularly hospitality and tourism, has an extremely high employee turnover (People 1st, 2015). As an increasingly large number of jobs are at risk of computerization (Frey & Osborne, 2017), the scale of reskilling required is unprecedented. Hospitality executives must carefully and proactively consider how to position themselves within this debate. It could be suggested that applying innovative technologies should come with an understanding of how important it is to futureproof this sector. For example, the United Nations’ (2018) Sustainable Development Goals stress the importance of achieving sustainable economic growth through the provision of “decent work.” However, what this might mean in practice for the hospitality sector in light of robotization is an ongoing debate. In any case, failure to take action may prove costly as skepticism and fear of automation has already induced strikes and protests around the world (Hernandez, 2018; Porter, 2018).
Theoretical Implications
Apart from the changes posed to hospitality operations management, marketing, and human resources, the advent of service robotics in hospitality service encounters has fundamental theoretical implications. The present study extends previous conceptualizations of the role of technology in service encounters (D. Bowen, 2016; Larivière et al., 2017) by considering the specific roles of service robots in hospitality service encounters. As argued throughout the article, while these seem to overlap, they go beyond those postulated in extant literature. This is because the novel capabilities offered by emerging frontline service technology (i.e., service robotics) have implications that go beyond the actual service interaction (De Keyser et al., 2019; Marinova et al., 2017). As established by previous literature, and as in line with previous frontline service technology, service robots can support or substitute employees in service encounters. However, due to the nature of this particular technology, the way this is done differs fundamentally from other static or preprogrammed technologies such as self-service kiosks or tablet computers (Wirtz et al., 2018).
First, as illustrated herewith, service robots can be mobile, allowing for greater visibility within the servicescape along with more complex, dynamic service interactions (Osawa et al., 2017). Furthermore, unlike previous frontline technologies, service robots may include a social dimension (Fong & Nourbakhsh, 2003). This can be due to an anthropomorphic design (e.g., shape, expression, external visual cues such as name tags) or the nature of the interaction itself (e.g., placing orders through natural language vs. scripted options, use of nonverbal communication cues such as gestures; Murphy et al., 2019). These unique features allow service robots to permeate deeper into the very core of producing and delivering offerings in hospitality service encounters (Ivanov & Webster, 2019b) and, in doing so, differentiate the human–robot service encounter from previous human frontline technology service encounters (Belanche et al., 2020).
Second, unlike previous frontline service technologies, service robots are characterized by their ability to sense and to make sense of their surrounding environment, as well as to take immediate actions that in some tangible way manipulate the physical world around them (Ivanov & Webster, 2019b). This allows for an unprecedented way to capture new types of data from service interactions (e.g., behavioral as opposed to transactional), as well as provides a novel means to act on insights gained to improve service encounters. Due to the largely passive nature of previous frontline technologies, frontline employees have played a key role in identifying areas of improvement in service offerings and practices due to their unique position in service delivery (D. Bowen, 2016). However, the advent of service robotics starts to disrupt this dynamic by facilitating new ways of collecting customer insight. As discussed herewith, service robots are already improving the service production (e.g., by eliminating issues of cross-contamination) and service delivery (e.g., by removing nonvalue-adding processes such as carrying plates back and forth from the kitchen) in hospitality service encounters. However, service robots’ ability to collect data from service encounters opens up a myriad of other, unforeseen ways to improve service encounters over time, making the service innovation process much more dynamic and potentially less dependent on human employees (Buhalis & Sinarta, 2019).
Third, as service robots are able to take on tasks hitherto done solely by humans, the role of human employees in service encounters is posed to change (Tuomi et al., 2019). This could mean new job titles (e.g., burger consultant, chef technician), new tasks (e.g., from operations to supervising robots), new skills (e.g., robot maintenance), as well as new approaches to people management (e.g., paid personnel development schemes). Based on discussions herewith, it seems service robots, due to their unique characteristics, facilitate this transition in a way previous frontline technologies have not. In other words, service robots seem to affect the sociotechnical system of hospitality on a level that is firmly rooted in service encounters, but has unprecedented implications for the wider service work ecosystem as well (Subramony et al., 2018).
Conclusion, Limitations, and Future Research
Automation using service robots has spread from the confines of factories to dynamic human environments. Service businesses across multiple domains, including hospitality, are being disrupted by these technological innovations. First, service robotics offer hospitality businesses effective means to increase efficiency and cut costs. However, the degree in which service production and delivery can or should be automated varies greatly in different contexts. As such, business executives should carefully consider how and where emerging technology should be applied. A clear strategy for dealing with inevitable technological hiccups is essential. Second, service robots offer marketing managers an attractive point of differentiation if they fit well with existing branding strategies. It may be best to focus on aligning marketing communications across digital channels and, in terms of demographics, target the young professional workforce. Third, service robots are likely to impact conventional people management practices. As such, frontline hospitality employees may see a shift from traditional waiting duties such as order-taking and payment processing to more specialized roles including burger consultants, product ambassadors, and experience guides. Simultaneously, back-end employees may experience similar changes. For example, chef duties may shift from repetitive tasks such as chopping vegetables or flipping burgers to more creative tasks including plating food or researching and developing new recipes.
This study contributes to existing literature on technological innovations in hospitality by analyzing transformations in the management and marketing of services due to the adoption of service robots. The findings of this study allow a better understanding of the strategic implications of automating parts of the service process, or all service processes. They reveal changes in service operations (internal, operational) and how customer expectations and satisfaction (external) can be managed. Based on these findings, the following recommendations can be suggested for hospitality professionals:
Setting a strategic service vision. Adopting innovative service automation through service robots should be based on competitive strategies to obtain the right customers in the marketplace. This vision should manifest itself in clear requirements for service quality (e.g., precise vs. flexible outcomes, error-free vs. bespoke experiences) and be integrated into service process designs that cover all touchpoints in the customer journey. This includes distributing tasks between machine and human labor in cases where human–robot collaboration is needed. The varying degrees of automation will require careful consideration of potential points of service failure. This should include customer interactions with technology (e.g., faulty robots, customers lacking the knowledge to use robots) and the corresponding service recovery strategies (i.e., service quality by design).
Communicating brand technology alignment. Customer acceptance is key to the successful adoption of service robots. In addition to having a clear service vision, communicating how service robots fit the brand and how the brand fits the desired characteristics of target customers (e.g., tech-savvy, efficient, forward-thinking) will assist in managing service relationships. Furthermore, it will create a barrier to entry into the marketplace, especially for service line pioneers.
Participating in futureproofing the hospitality industry. There is growing concern that automation will displace human labor, at least to some degree. Despite this, the advent of service robotics may also lead to a new era of people management in the hospitality sector. Operators may be encouraged to invest in the development of their staff and consider service offerings that are good for their customers and the planet. How this will happen remains to be seen, but the role played by the hospitality industry in facilitating lifelong learning for employees is likely to increase. Furthermore, to ensure socially beneficial adoption of service robots, greater regulation on the use of automation technology may be needed to nudge businesses in the right direction. As of 2019, only a handful of ethical guidelines for robotics development and deployment exist, regulations even less so (Palmerini et al., 2016; Boden et al., 2017; International Organization for Standardization, 2019).
In terms of the limitations of this study, service robots are still a relatively new phenomenon in hospitality service contexts. As such, the practical applications readily available to study are limited. To mitigate this issue, the research presented here collected observational data across two countries and 28 sites. However, in doing so, the time spent on each site was limited to an average of 4 hr. Although steps were taken to ensure sufficient research depth, more time spent in each location could have led to more specific insights. Furthermore, observations were only carried out by one of the authors. Although steps were taken to mitigate observer bias, using a team of researchers on each site could have increased consistency through investigator triangulation (Creswell, 2007). Finally, interviews were only conducted with elite informants (Aguinis & Solarino, 2019), otherwise known as expert agents with extensive knowledge of current service robotics development and deployment in hospitality. Conducting interviews with customers could have offered a broader view of the current effectiveness of service robots and revealed customer motivations for visiting establishments that make use of these robots.
As more practitioners continue to adopt robotics technology in hospitality, a quantitative approach that builds on the service robot acceptance model (sRAM; Wirtz et al., 2018), the artificially intelligent device use acceptance (AIDUA) model (Gursoy et al., 2019), or the service robot integration willingness (SRIW) scale (Lu et al., 2019) could provide further assessments for automation technology applications. Second, this research primarily adopted a managerial view. However, it is of equal importance to consider the short- and long-term impacts of automation on the employees delivering services and the customers receiving them and on the service ecosystem (Subramony et al., 2018). Although some research addresses human–robot interactions in relation to hospitality and tourism customers (e.g., Ivanov et al., 2018; Tussyadiah & Park, 2017), more research in different service contexts is needed. For example, employee attitudes toward service robots (e.g., acceptance or potential rejection), business models for leveraging automation (e.g., own or lease), and how to integrate service robots as part of people management practices warrant further research.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, or publication of this article.
