Assessment of service quality in a la carte restaurants within the mega city of Istanbul: A mystery shoppers’ study

Although the definition of quality varies across scientific fields and sectors, there are a few prominent definitions. Crosby (1979) defined quality as suitability for requirements, while Juran (1992) defined it as suitability for use. On the other hand, Deming defined good quality as a predictable degree of consistency and reliability, with a quality standard appropriate to the customer (Petersen, 1999). Quality has also been defined as the degree to which performance meets expectations (Chandrupatla, 2009). The American Society for Quality (2020) defines it as the degree to which goods and services meet customers’ needs and satisfy them.

Quality means different things to different people and is related to processes or results (Harvey & Green, 2006). As a concept, quality can be characterized as a philosophy of life; a management style; a process that aims to meet the needs of its citizens to make them happy; a way to achieve efficiency with an understanding that increases competitiveness within the sector and prevents waste; a system that unites the parties; and a continuous improvement process (Akgül, 1999).

Quality in the service sector has a very broad meaning (Mendocilla et al., 2021). Service quality significantly enhances customer satisfaction, yet its quantifiable benefits are challenging to measure due to inadequate resource allocation (Djofack & Camacho, 2017). Modern customers are sophisticated, affluent, educated, and time-poor, requiring restaurants to focus beyond just price, products, service offerings, for practical benefits.

For many restaurant customers, dining out is more than just eating outside the home (Ryu & Han, 2011). When restaurant customers decide to dine out, they are affected by the physical environment of the restaurant as much as the food and service (Hyun & Kang, 2014). Restaurants that have developed in today’s conditions can show differences in their physical attributes as well as the service they offer (Yurtseven & Yıldırım, 2014).

Restauranteurs and other stakeholders argue that atmosphere is important in creating a successful restaurant concept (Ryu, 2005). The concept of restaurant atmosphere was first defined by Kotler (1973) as the design of the space created by appealing to certain emotions that increase the likelihood of purchase in the consumer. The atmosphere of a business significantly influences consumer decisions, making it crucial to create a memorable dining experience for customers (Ryu & Han, 2011).

Restaurants are places where people experience excitement, pleasure, and well-being while also appeasing their hunger (Othman & Goodarzirad, 2013). How people feel in a restaurant is as important as good food (Quinn, 1981), as they seek a dining experience different from that in their homes (Ryu, 2005). Though out-of-home food consumption has increased, especially in urban settings, restaurant businesses have not been sufficiently researched. Customers prioritise not only the taste and quality of their food, but also the facilities and service features offered by the restaurant. This study focusses on the facilities and features offered by the restaurants and the quality values that restaurant customers give them.

Research has previously been conducted on ALCR (Axnestasya et al., 2023; Maretha & Juliana, 2023; Putri et al., 2024) and restaurant quality evaluations (Jun et al., 2017; Pezenka & Weismayer, 2020). However, there is a significant gap in the literature regarding quality assessments of ALCR businesses. In addition, although the SS method has been used to evaluate businesses in many service sectors, there have been no such studies in the tourism sector.

Through an extensive literature review, a model for ALCR was developed, based on a research model Tuncer et al. (2021) originally developed to measure FDR service quality (Figure 1).

Figure 1:

Map of the northern region of Portugal.

The proposed multi-dimensional and hierarchical service quality model was adapted for ALCR enterprises by adding the “Manager Quality” sub-dimension to Tuncer’s model.

This study aims to evaluate the product and service quality offered by businesses in the restaurant sector — a service sector aiming to provide high-level service — through observation by tourism professionals and academicians using the SS method. Since the characteristics of businesses in the service sector differ, service measurement models also vary (Tuncer et al., 2021). ALCR businesses are expected to be of high quality, and for this reason, ALCR businesses serving in mega-city Istanbul were chosen as this study’s sample group.

Although studies have been conducted to evaluate various dimensions of restaurant service quality, the first scale developed by Stevens et al. (1995) is DINE-SERV. The DINESERV model is proposed as a reliable, relatively simple tool for determining how consumers view a restaurant’s quality — but it is suitable for consumer use only. This study instead takes the opinions of professionals and academics in the field of tourism, rather than those of consumers, as the basis of its research. For these reasons, the DINESERV model was not used in the study.

This research was conducted in the Turkish tourism sector and utilized the purposeful random sampling method to target ALCR enterprises for in-depth analysis over a long period. This sampling method is used to reduce bias and increase credibility and reliability (Baltacı, 2018).

Mega-city Istanbul is one of the most important centres for tourism in Turkey, with the number of tourists increasing every year. There was an increase in the number of foreign tourists coming to Istanbul in 2024 increased by 14.01% compared to the previous month and 11.68% compared to the previous year (T.C. Kültür ve Turizm Bakanlığı, 2024). Therefore, the diversity of tourism in the city of Istanbul encompasses cultural, health, gastronomical tourism, religious and shopping tourism, to name just a few areas. Businesses serve the tourism sector, offering accommodation, food and beverage services, travel services, entertainment services, and more. All types of restaurant businesses providing food and beverage — local restaurants, traditional restaurants, fast food restaurants, fine-dining restaurants, ethnic restaurants — serve local and foreign tourists coming into the city.

The sample size was determined based on the depth and breadth of data sought from the individuals. According to Shenton (2004), qualitative research based on observation and interviews does not require large sample sizes, because the data obtained repeats itself after a certain stage.

In light of this information, restaurant establishments in Istanbul constitute the sample population of the study. According to the statistics obtained, 587 ALCR establishments serve under the supervision of the Ministry of Tourism throughout the city of Istanbul (T.C. Kültür ve Turizm Bakanlığı, 2024). Among these, a total of 311 establishments were deemed sufficient to form the research sample. This number shows that 52.9% of ALCR establishments serving in Istanbul province have been reached. In the study, the limited population (n < 10,000) formula, one of the formulas developed for qualitative research, was used to calculate the number of samples. The formula used is given below ( Ural & Kılıç, 2005): n=N.P.Q.Z α2(N-1).H2+Z α2.P.Q {\rm{n}} = {{N.P.Q.Z_\alpha ^2 } \over {\left( {N - 1} \right).H^2 + Z_\alpha ^2 .P.Q}} n

Sample size

The observation technique — one of the qualitative research methods — was used in the study. An observation form, consisting of four dimensions and 13 sub-dimensions measuring these dimensions, was prepared as a result of a comprehensive literature review, scales used in previous similar studies, propositions, and expert opinions. The observation form was re-evaluated and revised by 10 senior managers of 10 tourism enterprises, 10 tourism professionals, and 10 tourism academicians. After finalising the product and service quality criteria that customers expect from restaurant businesses, the researchers piloted the form in 30 restaurants. As a result of the pilot application, it was determined that the Cronbach’s alpha values of the product and service quality criteria were high (> 0.600). High Cronbach’s alpha values indicate reliability in the product and service quality criteria (Kilic, 2016).

In the secret shopper assessment (SSA), as in previous studies, each SS was informed and trained about the background, objectives, data collection, and study results before the observation. According to the quality criteria, which were tested for validity and reliability, the restaurant establishments were observed using the SSA method between February 1, 2022, and February 1, 2023. The evaluation periods were the same for all restaurant establishments and took place between 18:00–23:00 on Fridays, Saturdays, and Sundays, which are considered the busiest periods for restaurants.

The process of observing and inspecting each restaurant establishment with SSA was repeated once every four months. Each inspection was conducted by a different tourism professional or academic. After the evaluation, the SSA report was given to the business, and the arrangements made in the next SSA process were evaluated.

Data analysis was performed using the SPSS 15.0 statistical package. Data were presented as frequency, percentage, mean, and standard deviation. Normal distribution was tested using Kolmogorov-Smirnov and Shapiro-Wilk tests. The significance levels of Kolmogorov-Smirnov and Shapiro-Wilk tests were found to be p < 0.05, and it was concluded that there was no normal distribution. Skewness and kurtosis values were found not to be between +1.5 and −1.5, and it was thus concluded that the data were not normally distributed (Tabachnick et al., 2013). Non-parametric tests were used in the statistical evaluation of the data. The Kruskal Walls H test was used to measure quality improvement as a result of secret shopper evaluations, and the Friedman test was used for comparisons following improvements (Büyüköztürk, 2004).

Table 1 displays the characteristics of the restaurants included in the sample.

Out of the 311 establishments, 112 (36.0%) were categorized as TR, 106 (34.1%) were classified as LR, and 93 (29.9%) were FDR. Most of the participating restaurants (52.4%) were chain restaurants managed by department heads and business managers (42.8%). Additionally, 35.0% of the businesses had between two and 10 branches, 35.4% had been in the sector for for 10 to 14 years, and 40.8% employed between 10 and 49 staff.

Table 2 presents the operational information of the sampled restaurant establishments. The majority of the restaurants used the ALC service method only for dinner (36.3%). 49.8% of local guests preferred the establishments; 69.8% of local guests preferred Turkish cuisine, compared with 66.6% of the foreign guests.

Considering that the measurement tools were originally developed in English and later translated into Turkish, exploratory factor analysis (EFA) is required. The model proposed in the research was evaluated using AMOS. The factor loadings of the items according to the EFA results are given in Table 3.

According to the EFA results (Table 3), items with insufficient factor loadings and items that loaded onto multiple factors were removed and excluded from the final measurement model. After the EFA, the surviving items were sent to CFA for a more rigorous psychometric evaluation to empirically determine whether the research variables were differentiated or not. After CFA, items with insufficient factor loadings were removed. The results indicate that the proposed main four-factor structure, and the 13 substructures associated with these factors, provided very good outcomes (x² = 575.221, df = 192, x²/df = 2.996, root mean square error of approximation = 0.066, comparative fit index = 0.981, Tucker-Lewis index = 0.988, and incremental fit index = 0.981).

Table 3 presents the mean and standard deviation values of the IQ of the sampled restaurant establishments related to H₁, and Table 4 presents a statistical evaluation of the IQ of the sampled restaurant establishments. H₁ argues that the perceived quality of different establishment types affects restaurant customers’ perception of IQ.

It was determined that the FDR IQ values were lower (2.47 ± 1.25), but in the last control, the FDR value remained high (4.57 ± 1.30) while decreasing in the other restaurants (Table 4). The results show that establishment type affects customer perception of IQ. Thus, H1 is supported.

Based on the Kruskal Walls H test, it was found that the IQ value increased for all restaurants during the secret shopper evaluation process (p < 0.001). Based on the Friedman test, it was found that quality persistence was achieved (p < 0.001) (Table 4).

Table 5 presents the mean and standard deviation values, as well as a statistical evaluation, of the physical environmental quality of the sampled restaurant establishments, in relation to H₂.

H₂ argues that the perceived quality level of establishment types affects restaurant customers’ perception of physical environmental quality. FDR physical environmental quality values were found to be higher (3.74 ± 1.23), and the FDR value remained high (4.30 ± 0.96) while it decreased in other restaurants in the last control. These results show that the type of establishment affects customer perception of physical environmental quality. Thus, H₂ is supported.

It was found that the PEQ value increased for all restaurant establishments during the secret shopper evaluation process according to the Kruskal Walls-H test (p < 0.001), and the Friedman test showed that quality retention was achieved (p < 0.001) (Table 5).

Table 6 presents the mean and standard deviation values of the MQ of the sampled restaurants and the statistical evaluation of the MQ of the sampled restaurants in relation to H₃.

H₃ argues that the quality level of a given type of establishment affects the restaurant customers’ perceptions of management quality. TR MQ values were higher (3.74 ± 1.14), and the FDR value remained high (4.27 ± 0.97) while it decreased in other restaurants in the last control. The results show that customer perception of MQ is affected by business types. Thus, H₃ is supported.

Based on the Kruskal Walls H test, it was found that the MQ value increased for all restaurant establishments during the secret shopping process (p < 0.001), and quality retention was achieved (p < 0.001) according to the Friedman test (Table 6).

Table 7 presents the mean and standard deviation values of the OQ of the sampled restaurants, as well as the statistical evaluation of the OQ of the sampled restaurants, in relation to H₄. H₄ argues that the perceived level of quality of different types of establishments affects the restaurant customer’s perception of OQ.

It was determined that the FDR OQ values were higher (4.44 ± 0.82), and the FDR value remained high (4.69 ± 0.82) while it decreased in other restaurants at the last control. These results show that customer perception of OQ is affected by the type of business. Thus, H4 is supported. The Kruskal Walls-H test revealed that the OQ value increased for all restaurant establishments during the secret shopping process (p < 0.001), and the Friedman test demonstrated that quality persistence was achieved (p < 0.001).

As in many studies on the restaurant sector, this study has various limitations.

The biggest limitation of the study is that it was conducted in Istanbul. Due to constraints such as time, economic, and labour costs, this study cannot be representative of all ALCR restaurant establishments. This goal can be pursued in future research by including ALCR businesses operating in other provinces of the country with a larger team of researchers.

This study — conducted only in restaurants such as FDR, TR, and LR that provide ALS service — cannot represent the restaurant sector. Future studies could be conducted in establishments providing different services (such as fast food, buffet, table d’hote), and in specific types of restaurants such as ethnic and concept restaurants. This will better reflect the overall population of restaurant establishments.

Although the study reached a significant portion of the FDR businesses offering ALS operating in Istanbul, the rate of businesses reached remained at 29.9%. Considering the higher number of FDRs in holiday-tourism regions of Turkey, such as the Aegean and Mediterranean, if future studies are conducted in these provinces, the proportion of FDR businesses in the study can be increased.

This study was also applied to the chain and non-chain restaurant establishments and disregarded their national and/or international characteristics. Specific results may emerge with the inclusion of these different characteristics in future studies. For example, being national or international may make a difference in the quality characteristics of restaurant businesses. Management and franchising characteristics may also affect the quality criteria.

Within the scope of the study, four quality criteria were measured for ALSR in the model developed based on previous studies (Wu & Mohi, 2015; Tuncer et al., 2021). Future studies can be planned more comprehensively, possibly with an increased number of measured quality criteria.

Another limitation of this study is that only the data collected by tourism professionals through the SSA method was used to determine ALSR quality values. In future studies, the data diversity can be enriched by including the data obtained from restaurant customers.