TIPPING BEHAVIOR OF POLISH CONSUMERS – THE SEGMENTATION APPROACH

The main aim of the conducted research was to recognize the issue of tipping in restaurants by Polish consumers and to carry out consumer segmentation based on the reasons for tipping. The study was conducted in December 2017 on a sample of 1,000 people using the CAWI method. In order to isolate groups of reasons for tipping, the PCA analysis was used. Based on the separated groups of factors, with use of the k-average method, consumers’ segments were isolated and then subjected to profiling. The analysis of the results allowed to determine the frequency and amount of tips. It was found that satisfaction with the offer is the basic motivation for consumers to tip. As a result of segmentation, four groups of consumers, with different demographic, social, economic and behavioral characteristics, were distinguished. The obtained results lead to the conclusion that improving the quality of service, a nice atmosphere in the restaurant and meeting the expectations of customers in relation to the ordered dishes are the key factors that should be considered to encourage consumers to tip. #0#


Introduction
Tipping is an interesting and valuable phenomenon that influences the daily lives of thousands of hospitality professionals (Shih, Jai, Blum, 2016). In restaurants the issue of tipping is considered in the literature in four main aspects -economic, social, psychological and marketing. From the economic point of view, tips play a significant role in the salary of the staff, constituting a significant part of it, and they often do not undergo formal registration (Schwartz, 1997). In social terms, that customers who regularly visited eating establishments were more inclined to leave tips than occasional customers.
Consumers' propensity to giving tips was also analyzed with regards to the behavior and appearance of the waiter. As such, it was noted that the magnitude of the tip increased due to: good service (Azar, 2005;Lynn, 2004;Lynn, Sturman, 2010), direct behavior of the waiting staff (Seiter, Weger Jr, 2013), maintaining positive relations with the customer (Seiter, 2007;Seiter, Gass, 2002), authentic kindness of the waiting staff toward the customer (Azar, 2007;Bujisic, Choi, Parsa, Krawczyk, 2013) as well as an attractive appearance and outfit (Karagiorgakis, Malone, 2014;Lynn, McCall, 2000).
Additionally, research conducted so far has confirmed that the circumstances of the visit, such as: being on holidays (Greenberg, 2014), nice weather (Crusco, Wetzel, 1984;Rind, Strohmetz, 2006), size of the city where the visit was taking place (Garrity, Degelman, 1990;Lynn, Thomas-Haysbert, 2003;Rind, Strohmetz, 2006) and positive mood of the customer (Lynn, McCall, 2000) also influence both frequency of tipping and the magnitude of the tip. Moreover, a positive correlation between the size of the tip and the consumption of alcohol with the meal was observed (Conlin, Lynn, O'Donoghue, 2003;Sánchez, 2002).
As can be seen from the above-mentioned information, the consumer behavior of giving tips is relatively often the subject of research, but mainly those research concern consumers from North America and Africa. According to the authors' knowledge, there is no detailed recognition of the behavior of tipping consumers from European countries, including Central and Eastern Europe. Also in Poland, research on consumer behavior in terms of conditions and incentives to tip in gastronomic establishments has not been implemented so far. Therefore, it seemed reasonable to undertake research on these issues. Their results could have significant cognitive (understanding the scale and determinants of the phenomenon, comparison with the behavior of consumers from other countries) and application (modification of the offer and the manner of service in catering establishments) value.

Material and methods
The study was carried out in December 2017 on a sample of 1,000 people, with the use of the CAWI method. A preliminary study was conducted on a group of 48 people. To select the respondents the quota sampling was used -the structure of the study sample was parallel to the population of Poland in terms of gender, age, education and the size of the place of residence. The characteristics of the respondents are displayed in Table 1. All of the respondents were asked about the frequency of their visits to restaurants, the frequency of tipping and their economic, social and demographic features.
Respondents who give tips (with frequency of tipping from always to occasionally, N = 868) were ask about the magnitude of tips and reasons for tipping (the 5-point Likert scale was used, where: 5 -strongly agree, 4 -agree, 3 -neither agree nor disagree, 2 -disagree, 1 -strongly disagree). Cronbach Coefficient Alpha value for tipping reasons amounted to 0.836. Descriptive statistic covered calculations of frequency, means, statistical relevance of differences among the data with the use of the chi-square test and Spearman's rank correlation coefficient. Principal component analysis was used to isolate groups of reasons for tipping (Kaiser-Meyer-Olkin measure of sampling adequacy amounted to 0.811).
Cluster analysis was carried out based on the factors that had been isolated by the PCA analysis. The k-means method was used to isolate the clusters. The number of clusters was chosen on the grounds of the dendrogram and pseudo-F and pseudo-T 2 statistics. There were 4 accurately separated clusters. Variance analysis (ANOVA) was used for formal evaluation of the final clusters with the aim of exploring any significant differences among identified segments.
The results of the study were gathered using a Microsoft Excel sheet and analyzed with the use of statistical program SAS 9.4.

Results
The analysis of frequency of giving tips showed that 29.8% of the respondents tipped during each visit to a restaurant, 19.2% did it on average once every two visits, 7% of the questioned customers tipped on average once every 3-4 visits, 30.8% of the respondents were giving tips occasionally, whereas 13.2% persons declared that they never tip. The analysis of test results based on the independence test c 2 showed that the frequency of tipping increased with age, income, education level and frequency of visits to restaurants, but it does not depend on gender and place of residence.
The analysis of tips magnitude showed that over half of the tipping customers (50.69%) stated that the size of the tips left by them was within 5-10% of the bill, lower tips were given by 28.23% of the respondents, 13.48% left amounts ranging from 11 to 15%, and the remaining 7.61% tipped larger amounts. Similarly as in the case of the frequency of tipping, the relative amount of tips increased with age, income level, frequency of visits to gastronomic establishments, the frequency of tips and, to a lesser extent, the level of education, but it did not depend on gender and place of residence.
The most important reasons for tipping were satisfaction with the service (91.6% grades 4 and 5) and favorable rating of the remaining features of the restaurants offer, mainly: meals (82.9%) and atmosphere (82.1%). A less important reason for tipping was the good mood of the customer (69.2%) and the belief that giving a tip to the waiter is the right thing to do (70.1%). All the reasons connected to the restaurants' offer, as well as the recognition the tipping as an appropriate behavior, were positively correlated with the frequency of tipping and the magnitude of the tips. As a result of the PCA analysis, the three groups of reasons (customer's satisfaction with the offer, customer's mood and customer's beliefs) that had influenced the tipping behaviors of the respondents were isolated (tab. 2). As a result of the cluster analysis inclusive of isolated groups of reasons for tipping, four segments of respondents were chosen. They were different in terms of reasons for tipping, frequency of visits in restaurants, frequency of tipping, the magnitude of the tips, gender, education level, income and preferences toward features of the waiting staff (p<0.0001), but no statistically significant differences were found due to age (p-value 0.453) and place of residence (p-value 0.264). The identified segments were named: Positive, Convinced, Satisfied with the offer and Content.
Positive (45.6%) are the people whose main motivation to tip are personal beliefs and mood, although the appeal of the place is also important. They often visit restaurants (58.4% -a few times a month or more), and often give tips (53.7% every second time or always). The tips are relatively high: nearly 25% of this group members give higher tips than 10% of the bill. Respondents from this group have secondary or lower education (78.6%) and middle or low income (81.6%). This group's representatives considered all characteristics of the waiting staff to be less important by comparison with the other respondents.
Convinced (10%) to a large extent take into account beliefs when tipping, while the place's appeal and their mood is less important. They visit restaurants frequently (54.6% a few times a month or more). They tip occasionally and usually do not exceed 10% of the bill. Over half of them (51.3%) have primary and vocational education and almost 60% have middle income. Among the analyzed features of the waiting staff, they found appearance to be more important than the other segments.
Satisfied with the offer (34%) tip consider the restaurant's offer rather than being directed by mood and beliefs. They visit restaurants the most out of all groups (58% a few times a month or more). Similarly, they tip most often (55.4% always). The magnitude of these tips is relatively large: nearly 22% of those who were satisfied gave more than 10% of the bill. The majority (63.3%) has secondary or higher education, and their income is middle or high (56.1%). This group's representatives considered all characteristics of the waiting staff to be moderately important.
Content (13%) are led predominantly by their mood when deciding upon giving a tip. They moderately care about the restaurant's offer, and their beliefs play little part in their decision-making process. They visit restaurants rather seldom (55.2% once a month or less). They tip occasionally and give usually (86%) lower than 10% of the bill. The group consists of people with secondary and higher education (67.5%) but their income is rather low (57%). All the analyzed features of the staff, except the appearance, are more important to them than for the other respondents.

Discussion and conclusions
The results of the research allow to state that both frequency of tipping and the tips magnitude increased with age, income, education level and frequency of visits to restaurants -which was similar to the results of Lynn and Thomas-Haysbert (2003), Saayman and Saayman (2015) and Lynn and McCall (2000) -but were not dependent on the place of residence and gender -which was different then the research by Lynn (2006) and Saayman and Saayman (2015). The satisfaction of restaurants' offer (including service, taste of dishes, atmosphere, appearance and cleanliness) was the main factor motivating the respondents to reward staff members by tipping. The impact of service was also stated by Azar (2005) and Lynn (2004). Good mood of the customer was another important factor, which is confirmed by the research of Lynn and McCall (2000). Also the belief that tipping is an appropriate behavior had a positive impact on the respondents tipping behaviors. Similar result was obtained in the studies undertaken by Azar (2004) and Lynn (2006Lynn ( , 2015. Social pressure in a little degree influenced the tipping behaviors of the respondents, although Azar (2004Azar ( , 2005, Boyes, Mounts Jr. and Sowell (2004) and Saayman and Saayman (2015) proved the opposite. Consumption of alcohol was also a factor which motivated the respondents to tip, which confirms the results of the studies of Sánchez (2002) and Conlin, Lynn and O'Donoghue (2003). Nice weather and holidays also made customers more eager to spend money on tips, which confirms the observation of Rind and Strohmetz (2006) and Greenberg (2014).
The PCA analysis resulted in isolation of three groups of reasons for tipping -customers' satisfaction with the offer, customers' beliefs and customers' mood. All elements which component the satisfaction with the offer were found to be positively correlated with the frequency of tipping and the size of the tips, which reaffirms their influence on the tipping behaviors.
Cluster analysis isolated four segments of tipping respondents (Positive, Convinced, Satisfied with the offer, Content). The segment of "Satisfied with the offer" seems to be of particular interest of owners and staff of restaurants due to a number of reasons: firstly, it is comprised of consumers who frequently visit restaurants and often give tips, the value of which is quite high; secondly, the customers in this segment have middle and high incomes, which eliminates financial limitations of tipping. When it comes to the characteristics of the waiting staff, this group members valued speed of service and kindness.
It can be concluded that improvements to service quality, a kindly attitude towards the customer, a nice atmosphere in the eating establishment, as well as recognition and fulfillment of the customers' needs with regards to the menu, are the key factors that need to be taken into consideration when encouraging consumers to tip.