Ritual-based redesign of routines: an experimental approach in the restaurant industry

Purpose – This study aims to investigate the impact of the innovative ritual-based redesign of a routine in the challenging context of the dining-out sector, characterized by low employee commitment and high turnover. Design/methodology/approach – This study adopts a mixed methods experimental design. This study focuses on a ﬁ eld experiment in a real restaurant centered on the restaurant ’ s welcome entr (cid:1) ee routine. The routine is ﬁ rst observedasithappens,afterwhich itis redesigned as a ritual. Findings – The ritual-based redesign of the routine enhancesemployee sharing of thepurpose of the routine and reduces the variability of the execution time of the routine, which increases group cohesion among the restaurant staff.Besidesthepositiveimpactontheroutine ’ sparticipants,theritual-basedredesignhasabene ﬁ cialeffectonthe performanceoftheroutinebyincreasingtheenjoymentoftheend-consumersattherestaurant. Research limitations/implications – The ritual-based redesign of routines is a powerful managerial tool that bonds workers into a solidary community characterized by strong and shared values. This allows guidanceof thebehaviorof new andexisting employeesin a moreef ﬁ cientandless time-consuming way. Originality/value – Rituals have been traditionally analyzed from the customer perspective as marketing tools. This research investigates the employees ’ perspective, leveraging ritual-based redesign as a managerial tool forincreasing cohesionamongworkers.


Introduction
Routines are defined as the "specific actions enacted by specific people, at specific times, in specific places" (Feldman and Pentland, 2003, p. 94). Studies have found that the redesign of routines, defined as the introduction of an intentional change in the formal structure of routines (Bapuji et al., 2019), has conflicting outcomes. Some studies (Salvato, 2009) have shown that routines can be redesigned with beneficial results, whereas other studies (McGahan and Mitchell, 2003) have emphasized the disruptive consequences of routine redesign processes.
This paper presents the first attempt to investigate the outcomes of the innovative ritualbased redesign of a routine by combining literature on the redesign of routines with theories on rituals and testing the effects of the ritual-based redesign of a routine on all the constituent parts of the routine (i.e. ostensive, performative and performance).
As the adoption of ritual-based routine redesign is particularly salient for routines to which the participants tend to attribute symbolic meanings, we focused our study on the dining-out sector. Food is often identified as a way to express individual and collective meanings (Thomson and Hassenkamp, 2008). For example, coffee consumption is described as a "ritualistic pursuit of leisure" (Quintão et al., 2017, p.484); a birthday cake involves a ritual of steps that convey special meanings regarding positive expectations for the year ahead (Rossano, 2012), and specific foods are ritually procured and prepared for Thanksgiving (Wallendorf and Arnould, 1991). More specifically, food consumption experiences at restaurants are theorized as special instances where consumers have personal connections to the experience (Helkkula, 2011) as "they personally participate in the execution of consumption" (Kauppinen-Räisänen et al., 2013, p.674). Above all, those experiences are individual and subjective and, in turn, are imbued with emotional aspects that evoke feelings (Carù and Cova, 2008).
In tourism literature (Bargeman and Richards, 2020;Osman et al., 2014), rituals have been analyzed mainly as marketing tools directed toward customers to please them and enhance their consumption experience. This study shifts the focus away from customers to employees and their unaddressed perspective by leveraging rituals as managerial tools. The restaurant industry is traditionally associated with a low sense of employee belonging and high employee turnover (Madera et al., 2013), and rituals have been found to generate collective emotions and shared values that can bond workers into a community (Xiang et al., 2022). This study empirically tested whether the adoption of a routine redesigned as a ritual may lead to a more efficient and less time-consuming way of managing employees in the challenging context of the restaurant industry.

Theory and hypothesis
Ritualized interactions in the tourism and hospitality industry Ritualized guest-host interactions have come to the forefront during the coronavirus disease 2019  pandemic. Meaningful and predetermined sequences of actions (e.g. temperature checks, hygiene measures and social distancing) have produced emotional solidarity among employees and guests, facilitating the generation of group cohesion and future staff-guest interactions (Chan et al., 2021;Jeong et al., 2022;Xiang et al., 2022). Ritualized interactions have also been outlined in the context of robotic services in the hospitality field (Zhang et al., 2022), where consumers are becoming aware of predetermined sequences of actions that allow self-service ordering and autopayment systems (Guan et al., 2021;Wong et al., 2022). Bargeman and Richards (2020) explained the practice of going on a cruise as composed of ritual ingredients such as the bodily copresence of the participants (e.g. group travel bounded by the physical structure of the ship that creates barriers to outsiders), the common goal (e.g. having a good time on the cruise ship) and the shared mood (e.g. pleasure and happiness). All these ingredients during the cruise generate emotional energy in terms of IJCHM feelings and memories at the individual level, as well as group cohesion. Sterchele (2020) emphasized the ritual ingredients in terms of sacred experiences (i.e. stories posted on social networks, anecdotes and memorabilia) that extend the emotional power generated by the original experience, thus leading to revisiting the same location or attending the same event again. Leveraging the similarities between rituals and religions, Papastathopoulos et al. (2021) assessed the relevance of faith-based hotel attributes (e.g. halal food, prayer mat and distance from a mosque) from the consumer perspective.
Indeed, in the tourism and hospitality industry, rituals have been analyzed mainly from the demand side viewpoint (i.e. from the viewpoint of tourists as end-consumers), while the supply side perspective has been overlooked so far (i.e. the viewpoint of tourist service providers). Concerning the supply side, research has emphasized the relevant role of esthetic labor that tourism employees are expected to perform (e.g. body gestures, facial expression and voice adjustment) to please end-consumers (Tsaur and Tang, 2013). For example, when not serving customers, hotel staff "were required to stand at a 40-degree angle near the entrance to smilingly welcome customers. The employees were not allowed to stand with crossed arms [. . .]" (Tsaur and Tang, 2013, p. 20). Similarly, in the restaurant context, employee labor has been changed from primarily physical and mental to esthetic (Genc and Kozak, 2020). Restaurant patrons are increasingly valuing the way in which employees present themselves (e.g. how they speak to customers, their clothing style and their hairstyle) alongside the food quality and the overall environment. Given the high employee stress induced by this high degree of esthetic labor (Genc and Kozak, 2020), this study investigates the intuition that rituals may serve as strategic managerial tools for creating a sense of belonging and communitas (community) within an organization, which leads to a pleasant work environment.
Ritual-based redesign of a routine and its pillars Following Smith and Stewart (2011), we define organizational rituals as sets of activities that are invested with a special meaning that promotes the affirmation of shared beliefs and values, repetitive and highly invariable sets of sequential activities and acts not necessarily associated with the achievement of instrumental goals, since the focus is on the process rather than on the outcome (i.e. goal demotion). We now analyze in detail the three emerging distinctive traits of rituals and of the ritual-based redesign of routines before developing the hypothesis to be tested.
Special meanings and the ostensive aspect of routines The concept of special meanings is emerging as an essential characteristic of rituals. Although rituals involve explicit actions, the underlying meanings of those actions appear to be most important (Islam and Zyphur, 2009). Rituals enhance the affirmation of shared beliefs and values (Jong et al., 2015), the presence of which promotes group stability and in-group affiliation (Watson-Jones and Legare, 2016).
Considering the use of rituals to redesign routines, the fact that rituals, because of their special meaning, promote group cohesion appears to have a direct impact on subjective individual understandings of "why the routine is being performed or the purpose of the routine" (Feldman and Rafaeli, 2002, p. 321). According to Furlan et al. (2019), these understandings are the constituent elements of the routine's ostensive aspect. Feldman and Pentland (2003) suggested that this ostensive aspect should not be considered a single entity "because the ostensive incorporates the subjective understandings of diverse participants" (Feldman and Pentland, 2003, p. 101).
Although understandings are generated at the individual level, researchers state that they can be shared within members of an organization "supplying participants with mutually consistent interpretations and evaluations of information, as well as with reciprocal expectations concerning what actions are appropriate for the situation they face" (Dionysiou and Tsoukas, 2013, p. 193).
In this study, we hypothesize that the introduction of ritual-based redesign of routines increases the sharing of the individual understandings of the purpose of the routine compared to the status quo (i.e. the routine before the introduction of the ritual-based redesign).
Hence, the following hypothesis is proposed: H1. Ritual-based redesign of routines increases the sharing of the individual understandings of the purpose of the routine among the members of the organization who are performing the routine.
Invariable sequence of actions and the performative aspect of routines Rituals must be executed in the prescribed manner (Sosis and Alcorta, 2003), and this feature of rituals emphasizes their invariability (Hobson et al., 2018). Hence, rituals decrease variability in work processes, thereby strongly influencing the performative aspect of routines. Rituals introduce repetitive and invariable sets of sequential activities to create and recreate the underlying social meaning associated with their implementation. First, they freeze a sequence of actions that cannot change each time the routine is enacted. Second, they freeze the time it takes to perform each action in the standard sequence. Each action must be perfomed following strict protocols and standards both in terms of how the activity is performed and in terms of the time that it takes to perform the activity.
The standardization both of the sequence of activities and of each activity in the sequence should reduce the variability of the time that it takes to enact the whole routine each time. Through our field experiment in the dining-out context, we expect the introduction of the ritual-based redesign of routines to decrease the variability of the execution of routines compared to the status quo (i.e. the routine before the introduction of the ritual-based redesign).
Thus, we postulate the following hypothesis: H2. Ritual-based redesign of routines reduces the variability in the time of execution of routines.
Rituals and the effectiveness of routines Scholars have found empirical evidence that the presence of shared understandings contributes to the effectiveness of routines, defined as the extent to which the routine achieves the aims defined by the organization for the routine itself (Bapuji et al., 2019). Hence, the effectiveness of a routine is considered the routine's performance. For example, Mazmanian and Beckman (2018) analyzed budgeting as a ritual of quantification that can instill in middle managers a shared belief in the objective authority of numbers, thus making the budgeting process more efficient and less timeconsuming. The creation of shared understandings is also facilitated by the fact that actions in rituals are not necessarily associated with the achievement of specific instrumental goals (i.e. goal demotion). In fact, the focus is on the execution of the process rather than on its outcome (Legare and Souza, 2012). The emphasis on the proper execution of gestures substantiates the fact that when engaging in rituals, participants intensify their joint attention and coordinate their behavioral synchrony, thus generating perceptions of unity and cohesiveness (Hobson et al., 2018).

IJCHM
We expect that rituals, by facilitating the creation of shared understandings among routine participants and leveraging invariable sequences of actions, will provide the participants with a better guide to their actions, which will increase the effectiveness of the routine compared to the status quo (i.e. the routine before the introduction of the ritual-based redesign).
Hence, we hypothesize the following: H3. Ritual-based redesign of a routine enhances the effectiveness of the routine.

Research methodology Setting
This study used a mixed methods experimental design grounded on a field experiment in the dining-out sector. We focused on the dining-out sector because the connections between food and rituals have been frequently pointed out in literature (Ratcliffe et al., 2019). First, food has often been identified not only as a primary nutritional need but also as a way to express individual and collective meanings (Thomson and Hassenkamp, 2008). Second, in the dining-out context, people often consider the symbolic meanings derived from the food experience, which comprise the food served and how it is prepared and presented (Plester, 2015). Third, Fox (2003) showed that the order of the food consumed during a meal has ritualistic features, such as eating savory foods first and, at the end of the meal, sweet desserts.
Our experimental setting was an independent restaurant situated in the northeast of Italy. The ambiance of the 80-seat upscale restaurant is elegant, and its customers are generally established professionals and businessmen. The restaurant has a medium-to high-priced menu, with a complete meal, including the beverage, costing e50 on average. The menu changes twice a year to match the season (spring-summer versus autumnwinter). Thus, it was the same for the entire period of our experiment. The restaurant is managed by an entrepreneur who is the restaurant general manager, and his wife, who is in charge of the cash desk. They have equal shares in the restaurant. The staff are mostly employees with high seniority, with an average experience of 11 years. There are 16 full-time employees in all: 9 servers and 7 chefs.
We conducted our study on the welcome entr ee routine performed every day in the restaurant. Based on the restaurant's policy, guests are welcomed with an entr ee, also known as gift from the kitchen in English-speaking countries and amuse bouche in Frenchspeaking countries. We identified this welcome entr ee routine as suitable for our field experiment in this study for the following reasons. First, it comprises only a few actions, and this simplicity allowed us to overcome many of the methodological challenges related to the study of routines (Becker et al., 2005). Second, it is easy to manipulate in a field experiment, which enabled us to control the experiment and to create an unobtrusive setting for the study participants (McGrath, 1995). Third, this routine can be associated with many symbolic meanings (e.g. providing a warm welcome and immediately instilling a sense of staff friendliness). Fourth, it has a powerful effect in nurturing the priming effect (Davis and Herr, 2014) and, consequently, in influencing the entire consumption experience, as it is the first course served.
Familiarization stage Consistent with the research of Bapuji et al. (2019) on the dynamics of routines, we started our field experiment with the familiarization stage to develop a deep understanding of how the welcome entr ee routine is normally enacted. In this stage, we Ritual-based redesign conducted interviews with the entrepreneur and his wife as equal shareholders of the restaurant, observed the waiters at dinner time every day for one week (March 7-14, 2018) and conducted semistructured interviews with all the 16 full-time employees of the restaurant.
On the day of our arrival (March 6, 2018), the entrepreneur, after introducing us to the staff as external professionals invited by him, informed his employees of the duration of our study (21 consecutive weeks) and that he requested it because of an internal need for improvement. Hence, the staff were not informed of the real purpose of the study (Bapuji et al., 2019).
Next, the entrepreneur explained that the employees had to be actively involved in the study by participating in face-to-face interviews and that an intense training day would be scheduled on June 4, 2018, Monday, the restaurant's day-off, to introduce a change in the way of serving the welcome entr ee. Moreover, the entrepreneur revealed to all the staff that we would observe several dinners on some weeks. This assured the servers that their performance would not be monitored or evaluated in any way. In addition, the owner told the employees that for 21 consecutive weeks starting on March 20, 2018, the edible plate waste from the welcome entr ee and from the entire meal (i.e. the appetizers, first course, second course and dessert) would have to be put in two separate bowls, respectively, to be weighed instead of being thrown away directly.
From the familiarization stage, we wrote field notes focused on the actions that constituted the performative aspect of the welcome entr ee routine. The data gathered revealed that the welcome entr ee routine has five steps: (1) The servers welcome the guests and guide them to their seats.
(2) The servers gave the guests the menu and asked them which type of water they preferred.
(3) The servers served water and asked the guests what their food and beverage orders are besides water. (4) The servers served the beverages. (5) The servers served the welcome entr ee, saying, "This is an entr ee offered by our chef to welcome you to our restaurant." The entire sequence lasted 17 min on average. As with the entire menu, the welcome entr ee changes twice a year for the spring-summer season and for the autumn-winter season and is decided autonomously by the chef according to the seasonality and the menu. During the field experiment, the entr ee remained the same and consisted of two items: a cheese mousse seasoned with herbs and a creamy potato soup topped with two small slices of salmon (Plate 1).

Treatment
Our experiment called for redesigning parts of the welcome entr ee routine as a ritual. The entire field experiment ran for 21 consecutive weeks, from March 20, 2018, to August 12, 2018. After observing the welcome entr ee routine for 60 days (from March 20, 2018, to May 31, 2018, excluding Italian holidays and Mondays, as the weekly restaurant day-off), we conducted, together with the entrepreneur, an intense training day (from June 4, 2018, a Monday) to educate the employees on the redesigned welcome entr ee routine. After the training day, we observed the implementation of the redesigned welcome entr ee routine IJCHM for 60 days (June 5, 2018, to August 12, 2018), again excluding Italian holidays and Mondays.
During the training day, the presence of the entrepreneur was coherently justified by his role as the main proponent of the project, our participation supported the scientific basis of the proposed ritual-based redesign. The training day took place in the restaurant and was divided into two sessions (i.e. morning, afternoon). The nine servers attended both sessions, and the seven chefs joined the morning session only (as they were responsible for the physical preparation of the entr ee).
During the morning session, the entrepreneur first explained the aim of the training day, which was for the employees to learn a revised way of performing the welcome entr ee routine. Then, we described the detailed sequence of actions that the employees had to perform, by showing PowerPoint slides.
The changes mainly involved a predetermined sequence of actions and verbal instructions, fixed body gestures and symbolic meanings. First, we trained the servers to serve the welcome entr ee and to inform the guests to follow a specific order of consumption, tasting first the potato soup and second, the cheese mousse. The taste sequence for the two items was decided autonomously by the chef to balance the flavors. As a result of the redesign of the routine, the servers were asked to present the welcome entr ee as follows, after serving all the guests: This is an entr ee offered by our chef to welcome you to our restaurant. To enjoy the entr ee, mix the pieces of salmon with the soup for a few seconds, and then, taste the soup first. After finishing the soup, eat the cheese mousse.
We also emphasized the role of the voice volume and its cadence as elements of the ritualbased redesign of the welcome entr ee routine.
Second, we trained the servers to perform the following precise sequence of movements: after serving all the guests, ask them to consume the potato soup first, followed by the cheese mousse, while presenting the specific food items (e.g. the correct distance between the hand and the plate is 15 cm). Since specific movements are strongly associated with the definition of a ritual rather than random gestures (Vohs et al., 2013), we put emphasis on avoiding any form of random gesture.
Third, we communicated to the servers the symbolic meanings with which the ensuing ritual-based routine was strongly associated. Studies have shown that consumers who engage in ritualistic consumption behavior (e.g. a predetermined sequence of tasting: "the potato soup first, and the cheese mousse second") report better performance of a specific product than consumers who eat the same product with no associated ritual (Vohs et al., 2013). Moreover, according to Amati and Pestana (2015), a delay between a ritual and the moment of a product's consumption enhances the pleasure of consumption, so the detailed verbal instructions of the waiters on how to consume the entr ee enhanced the consumers' consumption satisfaction.
All in all, the three main features that constitute a ritual symbolic value (psychological), invariable sequence of actions (physical) and goal demotion were the foundations of the ritual-based redesign of the welcome entr ee routine. First, the symbolic meaning conveyed to the servers as the reason for their performance of the redesigned routine was the guests' greater enjoyment of their consumption experience at the restaurant. Second, the waiters had to adopt a prescribed and fixed temporal order in serving the welcome entr ee (i.e. taste first the potato soup, followed by the cheese mousse, with detailed verbal and non-verbal instructions on how to taste the food items). Third, when engaging in rituals, the focus must be on the correct execution of the entire sequence rather than on the outcome (Legare and Souza, 2012). During the training day, the proper execution of the gestures (i.e. the actions, voice volume and eye contact) for serving the welcome entr ee were strongly emphasized.
The day after the training day (June 5, 2018), the servers started to perform the ritualbased redesigned routine with the normal guests at the restaurant.

Rituals and the degree of sharing of the individual understandings
To test the first hypothesis, we analyzed the staff members' understandings of the purpose of the routine through semi-structured interviews. The interviews included the servers as well as the chefs, as both groups are involved in bringing the welcome entr ee routine to life. The interview process comprised two steps.
In the first step, we adopted a "why" interview technique (Erhardt et al., 2016). Sample questions are: Q1. Why do you serve a welcome entr ee to your customers?
Q2. What is the purpose of this activity?
Q3. What are the results that you expect from this activity?
The semistructured interviews ran from March 15 to March 17, 2018, and typically lasted between 15 and 25 min on average for each participant (Table 1). The field notes taken during the interviews were transcribed within a day of each interview.
We digitally transcribed the interviews to save the verbatim quotations. We adopted a thematic framework to analyze the outcome of the interviews ( Table 2). The process of coding the data included looking for themes, which DeSantis and Ugarriza (2000) defined as abstract entities that bring meaning and identity to a recurrent experience. We used an open coding procedure (Miles and Huberman, 1994) that included comparing quotations and grouping the similar quotations to form categories.
From the coding activity, seven categories of individual understandings emerged: (1) generate customer astonishment; (2) let customers feel at ease; (3) prepare customers for a better tasting experience of the whole meal; (4) instill an immediate atmosphere of familiarity; IJCHM (5) ease logistics; (6) increase customers' overall satisfaction; and (7) prepare the customers' palate for savoring the following course.
We assigned different codes to the abovementioned categories: ASTONISHMENT; COMFORT; EXPERIENCE; FAMILIARITY; LOGISTICS; SATISFACTION; and SAVOR.

Ritual-based redesign
Quotations from the interviews are reported in the Appendix.
In the second step, we developed a structured questionnaire based on the categories that emerged. Each of the servers and the chefs was asked to answer the same questionnaire twiceone at the end of the observation period and one at the end of the intervention period.
The respondents were asked to prioritize the seven categories related to the individual understandings of the purpose of the routine by ranking the categories in order of importance. The output was a personal ranking of individual understandings that ranged from the most important reason for performing the routine to the least important reason for the performance of the routine. We used the responses to assess the degree of consensus within and between groups (the chefs and servers) using the methodology of consensus mapping developed by Tarakci et al. (2014).
Mapping within-group consensus. To measure the degree of consensus within a group, Tarakci et al. (2014) used the vector model of unfolding (VMU). This approach is based on the transposed data matrix, which places the respondents in the columns as variables and the items (i.e. the understandings of why the routine is being performed) in the rows as cases.
H is the data matrix with m rows (i.e. understandings of why the routine is being performed) and n columns (respondents). It must be standardized so that all the individuals will have equal weight in the VMU. Consequently, according to Tarakci et al. (2014Tarakci et al. ( , p. 1057): The VMU in p dimensions is equivalent to minimizing the sum of the squared errors kEk 2 for H and the low dimensional representation XA 0 ; that is, where X is an m Â p matrix of the object scores for the m rows of the first p components and A is an n Â p matrix of the component loadings.
In the two-dimensional space, the results of the VMU can be represented using a biplot in which the rows of X (i.e. the object scores of the understandings of why the routine is being performed) are displayed as points, and the rows of A (i.e. the component loadings of the respondents) are depicted as vectors. Figures 1 and 2 show graphic features related to the resulting biplot representation of the VMU solution, developed through the software Wolfram Mathematica (Champaign, IL).
To correctly interpret the results, the following key points are fundamental. First, the cosine of the angle between two respondents is an approximation of their pairwise correlation (Linting et al., 2007). Respondents with small angles between their vectors had equivalent understandings of why the routine is being performed. Second, vectors displayed in a tight cluster show a high degree of within-group consensus, whereas a wide spreading of vectors indicates a low degree of within-group consensus. Third, the orthogonal projection of an item onto a respondent's vector depicts the respondent's ranking of that item. The more closely projected an item is onto the vector (the projections are shown with a dashed line), the more highly the respondent prioritizes the item. Moreover, because the first vector can be interpreted as the prototypical group member who represents better the whole group's opinion, the orthogonal projections of items onto the first axis unveil the group's overall understanding. Figure 1 shows the biplots of the servers in the pretreatment and posttreatment situations. Based on the projections of the items onto the first axis, comfort, familiarity and experience were the top three priorities of the servers in the pretreatment situation and logistics, satisfaction and savor were the least important. After the ritual-based redesign, the servers considered experience the most important item, followed by comfort and IJCHM astonishment, whereas the items that they valued the least did not change. The increase in priority of experience shows that after the treatment, the servers became more sensitive to the guests' evaluation of the tastiness of the whole meal. Moreover, the respondent vectors in the posttreatment situation are in a tighter cluster than those in the pretreatment situation, showing a high degree of within-group consensus. Figure 2 shows the biplots of the chefs in the pretreatment and posttreatment situations. Before the treatment, the chefs considered logistics, experience and astonishment the top three priorities and familiarity, comfort and savor, the last three. After the treatment, the chefs prioritized experience the most, followed by logistics and savor and astonishment, satisfaction and familiarity, the least. In addition, as the vectors are grouped in a tighter Ritual-based redesign cluster in the posttreatment situation, there was a higher degree of within-group consensus than in the pretreatment situation. The increases in priority of experience and savor after the treatment show that the chefs also leverage the power of the welcome entr ee in improving the taste of the whole meal. After the treatment, the prioritized understandings of the purpose of the routine appeared more common among both the servers and the chefs than before the treatment Mapping between-group consensus Following Tarakci et al. (2014), we assessed the between-group consensus. As the first axis represents the whole group's view, the correlation between the prototypical members of the two groups unveils the measure of their between-group consensus. Again following Tarakci et al. (2014Tarakci et al. ( , p. 1064, "the measure r(A, B) is operationalized as the correlation of the object scores of the items on the first axis for two groups, A and B." To obtain a graphic representation of the between-group consensus, we adopted classical multidimensional scaling (MDS). Figure 3 shows the MDS plot of the two groups in the pretreatment and posttreatment situations.
In our graphical representation, the chefs are at the center of the MDS plots. Following Tarakci et al. (2014), we used ten rings around the chefs that depict the correlational distance to the chefs, ranging from 0.9 to 0. The distance between the bubbles shows the degree of consensus between the groups. The sizes of the bubbles denote the degree of the within-group consensus in each group (a) and the rings that surround the bubbles depict the size of a bubble when there is perfect consensus within a group (a = 1).
The situation after the treatment showed a higher level of consensus on the prioritized understandings of the purpose of the routine. Indeed, after the treatment, the correlational distance progressively decreased, showing more aligned interpretations between the two groups. The common increase in priority of experience in both groups shows that a shared understanding of the welcome entr ee routine was emerging.

Rituals and variability in the time of their execution
To test the second hypothesis, we measured the variability of the time of execution of the welcome entr ee routine before and after the treatment. Between-group consensus between cooking staff and waiter staff IJCHM The data were collected over a total of 18 days during the observation period and another 18 days during the intervention period (on the 4th, 7th and 10th weeks in each period). The time was measured from the moment the customers entered the door of the restaurant to the moment the servers finished serving the welcome entr ee. The whole sequence of actions lasted approximately 17 min. A total of 21 tables are available in the restaurant.
One of the authors and a research assistant sat separately at two tables at the extreme side of the dining room. One was responsible for the first part of the dining room (for table numbers 1 to 10), and the other was responsible for the remaining area of the dining room (for table numbers 11-21).
Based on the restaurant's policy, each server is responsible for approximately three tables and is assigned specific table numbers. Since each observer focused on a specific area of the restaurant, they each measured the execution times of the same group of servers. Slight exceptions occurred mostly on weekends, when tables were often combined because of groups of clients.
The observation period yielded 488 observations among a total of 1,746 guests, whereas the intervention period generated 500 observations among a total of 1,718 guests.
The second hypothesis is that the ritual-based redesign reduces the variability in the time of execution of the routine by the employees who perform it. The results shown in Table 3 support this hypothesis. The variability of the time of execution was measured from the standard deviation of the time (in minutes) spent performing the welcome entr ee routine and it was greater before the treatment (SD = 2.58) than after the treatment (SD = 1.94). Hence, the time of execution appeared less variable and more stable during the intervention period than during the observation period.

Rituals and effectiveness of the routine
To test the third hypothesis, we measured the extent to which the ritual-based redesign increased the guests' enjoyment of their food consumption experience at the restaurant.
To assess the guests' food consumption experience, we weighed the edible food waste per person. This measure is widely used in restaurant management literature (Cozzio et al., 2021;Juvan et al., 2018).
The daily total food waste was divided by the daily number of customers to obtain the average amount of food waste from the welcome entr ee per person per day in grams. The whole period yielded 120 observations, equally distributed between the observation period and the intervention period. Table 4 shows that the average food waste per person per day from the welcome entr ee and from the whole meal were 4.07 and 16.16 g in the observation period and 3.49 and 15.17 g in the intervention period, respectively. The results of the oneway analysis of variance showed that the grams of food waste from the entr ee and from the whole meal were significantly different between the observation and intervention periods. This technique has been frequently used in experimental studies in the field of tourism (Shafer and Zhang, 2013). Notes: *p < 0.05; **p < 0.01; ***p < 0.001

IJCHM
In addition, the data showed the guests' average daily expense for food and beverage at the restaurant (obtained from the restaurant database) and the guest typology (i.e. adult women, adult men and children). On average, across the 120 observations, the restaurant's guests did not differ significantly in their average daily expense for food and beverage and in their typology.
We also ran two linear regression models to test the third hypothesis. We used the food waste in grams from the welcome entr ee and from the whole meal (Table 5) as dependent variables. For the explanatory variables in the two regression models, we used a dummy variable labeled treatment, which outlined the presence or absence of the ritualbased redesign of the routine, as well as the guests' average daily expense for food and beverage and the restaurant's daily numbers of female guests, male guests and child guests (baseline).
As shown in Table 5, the coefficient estimate for treatment_E indicates that the average food waste from the welcome entr ee after the treatment is 0.56 g lower than the corresponding average waste before the treatment. The t-test had a p value of 0.001, which shows a significant effect. Similarly, the coefficient estimate for treatment_M indicates that the average food waste from the whole meal after the treatment is 0.99 g lower than the corresponding average waste before the treatment. The t-test had a p-value smaller than 0.05, which shows a significant outcome.
The results support the hypothesis that the introduction of a ritual-based redesign of the routine will have a beneficial effect on the routine's performance by increasing the enjoyment of end-consumers at the restaurant.

Discussion and implications Conclusions
To complement our results, we conducted semistructured interviews with the study participants almost a month after the end of the field experiment. During the interviews, we asked for their opinions on the effects of the introduction of the ritual-based redesign of the routine. We took notes after the interviews. Notes: *p < 0.05; **p < 0.01; ***p < 0.001

Ritual-based redesign
The interviews revealed that a sense of collective solidarity and communitas (Islam and Zyphur, 2009) was achieved during the intervention period, which encouraged the emotional commitment of the participants. One server said: After the treatment I was strongly involved in the new way of serving the welcome entr ee and I perceived that the others were committed as well. We spoke intensively, almost every day, about the new way of serving the entr ee and I think that the mere fact of our speaking about the activity has reinforced our commitment to it.
Another server emphasized the transformative power of the ritual-based redesign to change disparate inputs into a collective output and to improve the consistency of the expected behavior of the other participants: I have been serving the welcome entr ee to our guests as it is a restaurant policy and I have never asked myself why I was performing that activity. The change in the way of serving the welcome entr ee helped me think about the meanings I derived from this activity. I think that we grew professionally as a group during this time. We also became closer to the cooking staff, sharing with them our feelings and thoughts about this introduction.
Another server reflected on his improved esteem and new awareness of his personal role: I have been feeling important in the role that I played in the restaurant and I clearly understood the relevance of the welcome entr ee routine as being our business card. I also perceived that our guests looked at me as an expert and they were more willing to ask me random questions about particular pairings during the meal.
Another study participant stated: Our role gained authority. The guests recognized it and the entrepreneur made it happen.
The intervention period significantly contributed to the rise in the importance of the activity itself: After the treatment, everybody is following a more systematic way of serving the welcome entr ee and, in turn, the importance of this routine in our tasks has increased. I realized that the activity of welcoming our guests should be the most relevant as it is the first and all the other activities depend on this one.

Theoretical implications
To the best of our knowledge, this study is the first attempt to empirically test the effects of a ritual-based redesign of all the constituent elements of a routine. All our hypotheses were supported, implying that the ritual-based redesign enhanced the degree of commonality of the individual understandings of the purpose of the routine, reduced the variability of the execution time of the routine and improved the effectiveness of the routine. Thus, this study represents a novelty in literature, as prior research merely analyzed the impact of redesign processes on the effectiveness of a routine (Bapuji et al., 2019;Mazmanian and Beckman, 2018). Second, this study adds to the discussion of customer experience management in the hospitality industry, which is meant to create unforgettable experiences for guests (Rahimian et al., 2021). Although consumers are increasingly leveraging contactless hospitality services that favor customer delight and trust, thus enhancing memorability (Hao and Chon, 2022;Wong et al., 2022;Zhang et al., 2022), this study revealed that the human touch still plays a vital role in the industry. In this study, rituals emerged as strong mechanisms that guided the participants' behaviors toward shared values and allowed coordinated lines of action. This, in turn, had beneficial effects on the enjoyment of endconsumers at the restaurant by increasing their pleasure with their experience (i.e. the routine's performance).
Third, the ritual-based redesign demonstrated both a transformational and stabilizing role in the organization, thus corroborating the "change within stability function of rituals" (Islam and Zyphur, 2009, p. 135). On the one hand, the ritual-based redesign preserved stability by instilling a sense of communitas (Islam and Zyphur, 2009) and spreading shared basic meanings in the group. On the other hand, the ritual-based redesign facilitated the transformation of the routine's participants by shaping both their cognition of, and their actions in, serving the welcome entr ee. The participants became more committed to the execution and the result of the routine and acquired more self-esteem. This theoretical advancement is particularly relevant in the specific context of the restaurant industry, where the employees' labor is mainly esthetic, as consumers are increasingly paying attention to how employees present themselves (Genc and Kozak, 2020;Tsaur and Tang, 2013). Against such background, rituals may act as powerful managerial tools that lower employees' high stress associated with the features of their labor, thus favoring a pleasant and cohesive mutual work environment grounded on collective learning pathways.
Fourth, as ritualized behaviors have been traditionally examined as marketing tools for pleasing end-consumers (Bargeman and Richards, 2020;Papastathopoulos et al., 2021;Sterchele, 2020), this study expands tourism literature by exploring the unaddressed perspective of employees. We have found that ritual-based redesign of a routine can be a powerful managerial tool by bonding workers into a solidary community.

Practical implications
Several studies in the dining-out context have examined rituals from a consumer perspective as marketing tools (Osman et al., 2014). Customer-based rituals linked to restaurant contexts may include singing happy birthday to a guest while presenting a cake and a lit candle, personalizing the check with a thank-you message from the servers or dining at the chef's table (Gardner and Wood, 1991). These examples of restaurant rituals are directed toward customers, to develop a relationship with them and to enhance their experience (Otnes and Lowrey, 2004).
Our research shifted the focus away from customers and adopted the employees' perspective. Prior research showed that rituals can be used to stimulate innovation (Jassawalla and Sashittal, 2002), shape identities (Coyne and Mathers, 2011), influence responses to environmental changes (Boyer and Linard, 2006) and challenge dominant values (Islam and Zyphur, 2009). Since the restaurant industry is traditionally associated with high employee stress, long hours, low employee commitment and high turnover (Madera et al., 2013), mechanisms through which core values can be shared and reinforced seem critical to investigate and implement in this field (Durand et al., 2007).
Following Erhardt et al. (2016, p. 39), "Individuals learn how to perform the ritual itself and, while performing the ritual, they also learn to recognize the underlying value of their performance of the ritual." Therefore, the ritual-based redesign of routines can serve as a mechanism for developing an organizational culture characterized by strong and shared values that guide new and existing employees' behaviors in a more efficient and less time-consuming way. Hence, practitioners may implement the redesign of routines as rituals to increase employees' sense of belonging and communitas within their organization, thus increasing cohesion among workers. This may create a competitive advantage for the tourism service provider, especially in the challenging setting of the restaurant industry.

Limitations and future research
This study had a few limitations. First, it was based on an industry and country where welcome entr ee routines are very common. Hence, caution should be taken about the applicability of our findings to foreign countries.
Second, as dining behaviors in fast-food restaurants differ from consumption experience in traditional restaurants, this study was limited in that it was based on a single field experiment conducted in an upscale restaurant. Future studies should investigate a wider range of restaurants (i.e. upscale, casual dining and self-service restaurants) to consider greater differences in guests' typology and in menu features.
Third, our study had a limitation in the measurement of the performance of the redesigned routine through the assessment of the edible food waste. Because of logistics reasons, the assessment of the edible food waste did not consider the amount of food served in the first place. This aspect is critical, as many cooks may not accurately weigh food when plating it. In addition, the measurement did not consider the amount of food that could have accidentally fallen off the plate or that could have been removed from the plate. Besides assessing guests' enjoyment at a restaurant by quantifying their food wastage, it would be promising to further scrutinize the effect of ritual-based redesign of routines on customer satisfaction by adopting other techniques (i.e. semi-structured interviews and surveys). Table A1. Individual understandings about the purpose of the routine IJCHM About the authors Claudia Cozzio is Assistant Professor of Management at the Department of Economics and Management at the Free University of Bolzano. She holds PhD in Management with a focus on the tourism industry. Her main areas of research include tourists' behavior, sustainable tourists' consumption, innovation in tourism and mental budgeting. Claudia Cozzio is the corresponding author and can be contacted at: claudia.cozzio@unibz.it Andrea Furlan is Full Professor of Management at the University of Padova, Department of Economics and Management. He graduated in Business Economics at the University of Padova, where he completed his PhD in Economics and Management with a doctoral dissertation on supply chain management. His main research interests focus on lean management and continuous improvement, operational and managerial problem solving, spinoffs, routines' replication and employees' proactivity.
For instructions on how to order reprints of this article, please visit our website: www.emeraldgrouppublishing.com/licensing/reprints.htm Or contact us for further details: permissions@emeraldinsight.com Category Quotations everybody outside is waiting for our food preparations. I feel more relaxed even if in an extremely stressful environment and I am more focused on what I need to prepare and on coordinating the staff without having too many external pressures and worries" (Giuseppe P., chef) "It allows to have a greater time management ability within the kitchen staff regarding the different food preparations" (Mauro, line cook) SATISFACTION "The needs of today's customers must be met immediately because in our society people are used to find the answer to what they are looking for with a click on the web. Restaurant clients are similar in terms of their impatience, so you need to offer a precise and timely service to be able to satisfy them and to increase their loyalty. The welcome entr ee service represents a little piece that can contribute to the guests' overall satisfaction" (Elisa, waiter) "I noticed that many clients are sensitive to the food presentation because, firstly, the clients "eat" with the eyes instead of with the mouth. Since we offer the welcome entr ee in an elegant way by serving two different little items in suitable dishes, I think that it can contribute to instill directly the perception of well-created food preparation and, in turn, it can increase the client's satisfaction experienced at our restaurant" (Muriel, waiter) SAVOR "The welcome entr ee aims at preparing the customer's palate for savoring the following course. Through the welcome entr ee, the guest starts to prepare himself/ herself to the savoring of the following course in terms of its mental imagination as well as in terms of the guest's dool toward the following course that his/she has ordered" (Giuseppe P., chef) "The welcome entr ee enhances the guest's desire to savor the following course. It is a way of opening the guest's stomach and preparing his/her taste papilla" (Giovanni, line cook) Table A1.
Ritual-based redesign