Coefficient SE t p
Constant 4.633 0.14 32.27 0.0000
Robotic Service Acceptance (X) c1 0.715 0.05 13.99 0.0000
Type of Value (M) c2 0.036 0.09 0.40 0.6910
Int_1 (XM) c4 -0.163 0.10 -1.70 0.0912
Reason for Stay (W) c3 -0.045 0.09 -0.51 0.6127
Int_2 (XW) c5 -0.049 0.10 -0.51 0.6137
Female 0.146 0.10 1.53 0.1267
Age -0.044 0.00 -1.23 0.2212
Business 0.341 0.20 1.67 0.0956
Prior -0.175 0.10 -1.74 0.0837
Number -0.008 0.12 0.06 0.9501
R2 = 0.53, p <0.001
F(10, 220) = 24.73
Table 10: Results PROCESS Model 2
Figure 4: PROCESS Model 4
Indirect effect (X on M = a1). The results in Table 11 show that the regression coefficient for a1 was .269 and was statistically significant, t = 2.277, p = .0237, with a 95%
confidence interval ranging from .036 to .502. This means that Robotic Service Acceptance is .269 higher when the customer has a hedonic value.
Indirect effect (X on M = b1). The effect b1 = .715 indicates that two people who experience the same Type of Value, but that differ by one unit in their level of Robotic Service Acceptance are estimated to differ by b1 = .715 units in Customer Experience. The sign of b1 is positive, meaning that those relatively higher in Robotic Service Acceptance are estimated to be higher in their Customer Experience. This effect is statistically significant, t = 14.190, p = .000, with a 95% confidence interval ranging from .616 to .814.
Total indirect effect (a1b1). Overall, there is a significant indirect effect (a1b1 = .192), and it is statistically different from zero. This is shown by a bootstrap confidence interval that is entirely above zero (95% CI: .019 to .369).
Direct effect (X on Y = c1’). The results in Table 11 show that there is no direct effect of Type of Value on Customer Experience (c1’ = .034, p = .7070). Moreover, the direct effect is not statistically different from zero, t = .376 and the 95% confidence interval ranges from -.413 to .211.
X Type of value (hedonic vs utilitarian)
M Robotic Service
Acceptance
Y
Customer Experience
Total effect (c1 = c1’ + a1b1). Finally, the results show that there was no total effect, as c1 was .226 and was not statistically significant, t = 1.847, p = .0661. The 95% confidence interval ranged from -.015 to .468.
Control variables. In the case of the indirect effect (X on M = a1), the control variables Gender (Female) (p = .0055) and Age (p = .0001) appear to have a significant effect. Which means that females and older people will score lower on Robotic Service Acceptance.
Table 11: Results PROCESS Model 4
Consequent
Robotic Service
Acceptance (M) Customer
Experience (Y)
Variable B SE p B SE p
Type of Value (X) a1 .269 .118 0.0237 c1' .034 .090 0.7070
Robotic Service Acceptance (M) --- --- --- b1 .715 .050 0.0000
Female -.349 .125 0.0055 .139 .095 0.1466
Age -.019 .005 0.0001 -.004 .004 0.2759
Business .338 .270 0.2117 .333 .204 0.1040
Prior Experience .059 .134 0.6604 -.181 .101 0.0734
Number of stays .015 .163 0.9251 -.009 .122 0.9448
Reason for Stay -.127 .118 0.2806 -.048 .089 0.5887
Constant i1 5.489 .198 0.0000 i2 1.188 .314 0.0002
R2 = .121 R2 = .089
F(7, 223 ) = 4.390 F(8, 222 ) =
30.319
p< .001 p< .001
Effect SE p LLCI ULCI
Direct effect c1' .034 .090 0.7070 -.143 .211
Total effect c1 .226 .123 0.0661 -.015 .468
Boot SE Boot LLCI Boot ULCI
Indirect effect a1b1 .192 .089 .019 .369
PROCESS Model 7
To check whether the influence of Type of Value (hedonic vs utilitarian) on Customer Experience has a moderated mediation effect, PROCESS Model 7 was used for the analysis (see Figure 5).
Figure 5: PROCESS Model 7
Moderated mediation. The findings from the statistical analysis (see Table 12) reveal that Type of Value has a significant effect on Robotic Service Acceptance (a1 = .269, p < .05).
This means that the Robotic Service Acceptance is .269 higher when the customer has a hedonic value. However, there is no moderated mediation effect, as the interaction effect is not statistically significant (p > .05). This means that Reason for Stay (business vs leisure) does not moderate the relationship between Type of Value (hedonic vs utilitarian) and Robotic Service Acceptance. The b-path shows a significant effect of Robotic Service Acceptance (M) on Customer Experience (Y) (b1 = .034, p < .001). However the total indirect effect shows a negative and not significant effect (a1b1 = -.029, 95% CI: -.355 to .308). The direct effect of Type of Value (X) on Customer Experience (Y) is also not significant (95% CI: -.143 to .210).
This means that the Customer Experience is not different for a hedonic or utilitarian Type of X
Type of Value (hedonic vs utilitarian)
Y Customer experience M
Robotic Service Acceptance W
Reason for stay (business vs leisure)
Value. Again, in the case of the indirect effect (X on M = a1), the control variables Gender (Female) (p = .0055) and Age (p = .0001) appear to have a significant effect, which means that females and older people will score lower on Robotic Service Acceptance. All these results confirm the findings of PROCESS Model 4, and complement the findings by showing that there is no moderated mediation effect.
Table 12: Results PROCESS Model
Consequent
Robotic Service
Acceptance (M)
Customer
Experience (Y)
Variable B SE p B SE p
Type of Value (X) a1 .269 .185 0.0239 c1' .034 .090 0.7095
Robotic Service
Acceptance (M) --- --- --- b1 .717 .050 0.0000
Reason for Stay (W) a2 -.127 .118 0.2825 --- --- ---
Int_1 -.040 .236 0.8650 --- --- ---
Female -.351 .125 0.0055 .138 .095 0.1487
Age -.019 .005 0.0001 -.004 .004 0.2560
Business .340 .271 0.2103 .337 .204 0.0993
Prior Experience .060 .134 0.6569 -.183 .100 0.0698
Number of stays .017 .163 0.9192 -.005 .122 0.9671
Constant i1 5.560 .185 0.0000 i2 1.176 .312 0.0002
R2 = .121 R2 = .522
F(8, 222 ) = 3.829 F(7, 223 ) = 34.719
p< .001 p< .001
Effect SE p LLCI ULCI
Direct effect c1’ .034 .090 0.7095 -.143 .210
Boot SE Boot LLCI Boot
ULCI
Indirect effect a1b1 -.029 .169 -.355 .308
6 Discussion
The use of AI and robotics is becoming more and more prominent in the hospitality industry and is rapidly and radically changing the nature of service encounters, and customers’
service experience (Ostrom et al., 2015; Rust & Huang, 2014). In this initial attempt to investigate the influence of Robotic Service Acceptance on Customer Experience, with the moderating roles of Type of Value and Reason for Stay, an online survey experiment was conducted.
The results of this study showed that Robotic Service Acceptance had a significant influence on the overall Customer Experience of the guests. Thus, H1 was accepted. Murphy et al. (2019) highlighted that AI could enable robots to know customers better, and create relationships that could increase customer experience, customer loyalty, and engagement with a service provider. Besides, this result corresponds to the finding of van Doorn et al. (2017) who stated that guests had a higher overall Customer Experience when the service delivery was provided by a service robot. With these results, the study at hand contributes to the field of customer experience by combining these findings with technology acceptance theories.
Furthermore, the present study investigated whether the relationship between Robotic Service Acceptance and Customer Experience was moderated by Type of Value (hedonic vs utilitarian). This study revealed that the relationship is not moderated by Type of Value and therefore H2 needed to be rejected. However, it can be said that H2 is partially accepted. Further investigation of Type of Value led to the result that people with hedonic values show a higher Robotic Service Acceptance, and therefore also show a higher Customer Experience. This corresponds to the results of Roy (2018) and Dedeoglu et al. (2018), which show that hedonic values are seen as extra elements that influence the satisfaction and the customer experience in a psychological way. Moreover, this research supports the general idea that experiences play a bigger role in a hedonic valued service than in a utilitarian service (Jones et al., 2006).
In addition, this study also investigated a second moderating effect, namely Reason for Stay (business vs leisure). Other than expected, the influence of Robotic Service Acceptance on Customer Experience is not moderated by Reason for Stay (business vs leisure). Hence, also H3 needed to be rejected. Victorino et al. (2005) found that leisure travellers value the hotel’s innovativeness more than business travellers do. However, this research does not match these findings, since no differences between groups were found in terms of customer experience. An explanation for this insignificant result might be that most of the respondents involved in this research indicated that they travel for leisure purposes. According to Dahl and Hoeffler (2004), the formation of images or scenarios regarding an inexperienced occurrence has no effect on evaluations of new products or services, such as service robots. This can be clarified mainly because of consumers’ difficulty in imagining those inexperienced experiences (Dahl &
Hoeffler, 2004). Therefore, it might be difficult for the concerned respondents to imagine the experience for another travel purpose, and evaluate their customer experience based on this imagination.
Furthermore, results showed that Gender, and especially Females have a negative impact on Robotic Service Acceptance. This means that females are less likely to accept service robots than non-females. This confirms the findings of de Graaf and Allouch (2013) which showed that women are more hesitant to interact with robots, have a negative perception of them and are less inclined to use these service robots.
This study also revealed that Age has a negative impact on Robotic Service Acceptance and Customer Experience, which means that older people are also less likely to accept service robots, and have less customer experience when the service is provided by a service robot.
Several studies confirm this finding, Blut et al. (2021) state that older people are more distrustful of technology and have a negative attitude towards robots, hence they are less likely to accept and use them.
Finally, a further investigation of the PROCESS Models showed that the Robotic Service Acceptance and Customer Experience is higher when the customer has hedonic values.
This means that there is an indirect relationship between Type of Value and Customer Experience, which is mediated by Robotic Service Acceptance. The further investigation shows that people with hedonic values show higher Robotic Service Acceptance and also score higher on Customer Experience. The fact that the hedonic Type of Value leads to a higher Robotic Service Acceptance, is in line with previous research of Klamer et al. (2010). The study of Klamer et al. (2010) shows that hedonic and pleasurable factors play an important role in the acceptance and use of service robots. They say that when users feel emotions such as playfulness and enjoyment while interacting with service robots, it has an influence on the intention to use. Furthermore, the study of Park and Kwon (2016) found that perceived enjoyment had a significant impact on the intention to use the technology. Besides, the relationship between Type of Value and Robotic Service Acceptance is not moderated by Reason for Stay.
In summary, despite the rapid development of new technologies in the hospitality industry, including robots and artificial intelligence, research on AI-enabled customer experiences is still limited. This work facilitates a better understanding of how the use of robotic services in the hospitality industry can impact the customer experience of guests. The results led to suggestions for hospitality companies in terms of successfully managing their communication strategies to ensure that customers start and continue to use service robots during their stay at a hotel.