MASTER THESIS
MSC BUSINESS ADMINISTRATION: MARKETING
The short- and long-term effects of hotel price promotions:
a tradeoff between increased purchase intentions and decreased quality
perceptions and its effect on the customer’s willingness to pay.
Amsterdam, January 27th, 2016
Student Kristian Kabbedijk Student nr. 11207620
Supervisor Adriana Krawczyk
Keywords: hotel price promotions; purchase intention; perceived quality; willingness to pay
Statement of originality
This document is written by Kristian Kabbedijk, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
2
TABLE OF CONTENT
ABSTRACT ... 4 1. INTRODUCTION ... 5 2. LITERATURE REVIEW ... 8 2.1 Price promotions ... 8 2.2 Purchase intention... 9 2.3 Perceived quality ... 10 2.4 Customer’s willingness-to-pay ... 12 2.5 Conceptual Model... 14 3. METHOD ... 16 3.1 Research Design ... 16 3.2 Sampling ... 17 3.3 Measurements ... 18 3.4 Data analysis ... 20 4. RESULTS ... 23 4.1 Sample Demographics ... 23 4.2 Descriptive statistics ... 24 4.3 Correlation analysis ... 24 4.4 Statistical testing ... 255. DISCUSSION AND CONCLUSIONS ... 31
5.1 General discussion ... 31
5.2 Managerial implications ... 33
6. LIMITATIONS AND FUTURE RESEARCH ... 36
6.1 Study Limitations ... 36
6.2 Future Research Suggestions ... 38
7. BIBLIOGRAPHY ... 40
3
LIST OF TABLES AND FIGURES
Table 2.5 Hypotheses overview
Figure 2.5 Conceptual Model
Table 3.4 Scale analysis for internal consistency
Table 3.5 Outlier analysis
Table 3.6 Distribution analysis for normality
Table 4.1 Descriptive statistics of demographics
Table 4.2 Descriptive statistics of mean, standard deviation and sample size
Table 4.3 Means, standard deviations, correlations and reliabilities
Table 4.4 Output PROCESS: model summary for change in variance
Table 4.5 Output PROCESS: testing for mediation, the direct effects
Table 4.6 Output PROCESS: testing for mediation, the indirect effects
4
ABSTRACT
One of the crucial challenges for hotel marketers is the perishability issue of the services industry. Hotels try to cope with unsold capacity by introducing remarkable high discounts through both their own online channels and those of intermediaries. However, little guidance is given in academic literature about the actual effects of hotel price promotions. By focusing on the popular use of comparative price advertising, this study therefore complements the extant research by exploring the short-term tradeoff effects of hotel price promotions on both purchase intention (positive) and perceived quality (negative), and contributes to previous studies by hypothesizing a subsequent negative effect on customer’s willingness-to-pay on the long run. Data was collected from 104 participants in an experimental vignette study with a between-subjects design, controlling for the presence of a price promotion. Results suggest that a price promotion strongly relates with the increase of purchase intentions. The analysis however did not find a relationship with perceived quality. This conclusion sheds new light on the previous findings about the customer’s use of quality cues in discount framing, and provides interesting suggestions for further research. Furthermore, the results show that the customer’s willingness to pay is significantly lower as a result of the consumer being previously exposed to the hotel’s price promotion. The study investigated that this strong negative relationship is mediated by the level of purchase intention and so that the higher the purchase intention, the less detrimental the effect is on the customer’s willingness-to-pay. However, the paper concludes that purchase intention alone is not sufficient to compensate for this negative relationship.
5
1. INTRODUCTION
Given the positive influence of price promotions with regards to monetary savings, consumers exhibit more favorable attitudes towards this promotion (Chandon et al., 2000) and will be more likely to afford this particular product or service with a price discount (Yang et al., 2016). Consumers not only consider the product price, but also the price of alternative products offered by the competition (Oh, 2000). Chiang & Jang (2006) show that hoteliers can easily improve short-term value through offering discounts since perceived price is a good predictor of purchase intention. They argue that the more favorable the perceived price, the higher the perceived value that leads to purchase intention. These findings might indicate why globally almost half of all upscale and luxurious hotel properties have participated in online flash deals to stimulate customer acquisition (Piccoli and Dev, 2012). Due to product perishability, unsold rooms are a direct loss and therefore offering price promotions seems to be an attractive way for an immediate increase in purchase intention and consequently an increase in sales, especially during low-season periods (Steed & Gu, 2004).
However, other papers provide reasons to believe that these positive effects of price promotions do not compensate for the negative results. Montaner and Pina (2008) argue that monetary promotions generally change the initial perceptions of the product. Previous studies have demonstrated that perceived quality is related to price promotions because price is considered to be a general quality cue (Ye et al., 2014; Chang, 2009; Chiang & Jang, 2006; Oh, 2003; Suri et al., 2002; Grewal et al., 1998). Suri et al. (2002) explain this by stating that fixed pricing elicit stronger positive affect than a discounted price format and therefore consumers are less likely to thoroughly process price information. As a result, consumers perceive the product quality as higher in the case of fixed pricing formats and lower for discounted price formats. Also, previous studies have shown that price discounts reduces the customer’s internal reference price, which refers to the price the customer expects to pay prior
6
to the purchase occasion (Delvecchio et al., 2007; Chandrashekaran & Grewal, 2006; Kalwani and Yim, 1992; Grewal et al., 1989). Johnson & Cui (2013) most recently confirmed this by demonstrating that discounted prices create downward pressure on the internal reference price that leads to lower post-promotion price expectations. Ariely et al. (2006) explain how the price level of a product or service may be prevailed due to collective anchoring which was triggered by previous ‘accidents’ and manipulations. When the internal reference price decreases, consumers expect the price to be lower and therefore perceive the value to be less (Chiang & Jang, 2006). Furthermore, Johnson & Cui (2013) and Chandrashekaran & Grewal (2006) state that the internal reference price is a common reference point for consumers to assess the amount that they are willing to pay. To understand the implications of each pricing strategy, many academics dedicated their research to the importance of customer’s willingness-to-pay (e.g. Wardell et al., 2008; Hinterhuber 2008; Monroe, 2003; Wertenbroch & Skiera, 2002). They all conclude that customer’s willingness-to-pay is a key measurement tool in forecasting market response and assessing financial impact on your business. However, no clear directions are given in the case of hotel price promotions and their effect on willingness-to-pay on the long run.
In conclusion, the findings of previous studies would imply that price promotions can enlarge a hotel’s customer base while at the same time these customers might perceive the quality differently and expect lower prices after previously being exposed to a price promotion. As a result, the once effective strategy with short-term benefits would hurt the pricing strategy on the long term and eventually affect the hotel’s profit margins due to decreased numbers in customer’s willingness-to-pay.
Price promotions clearly have an impact on consumer decision-making, but no consensus is reached amongst academics about the net benefits of price promotions when compensating for the negative results too. However, because of the practical significance of
7
price promotions, both academics and practitioners require answers for this field of pricing effects. The fact that academics have not reached consensus yet for the short-term tradeoff effects between purchase intention and perceived quality is disregarded by hoteliers since discount pricing strategies are often applied. Also, there is too little guidance on how this might damage the profit margins for the long-term when price expectations decrease. This paper therefore aims to discover the short-term tradeoff effects and define their relevancy for the hospitality industry in specific by means of an experimental vignette study. In addition, this research paper also tries to define the long-term implications of price promotions in terms of willingness-to-pay and looks into the mediating influence of purchase intention and perceived quality on this long-term effect. The following research question applies: What is the effect of online hotel price promotions on the customer’s willingness to pay and what is the role of purchase intention and perceived quality for this effect?
In the next section, the different constructs are being discussed in more detail and, if applicable, presented in the context of the hotel (services) industry. Afterwards, the research study is introduced and the research strategy is explained. Following, the results of the study are presented and implications for both academic fields and business practices are discussed. Lastly, potential shortcomings of this study are elaborated on and suggestions for future research are given.
8
2. LITERATURE REVIEW
This section provides a theoretical framework of the concepts that are closely related to the research topic. Firstly, price promotions will be introduced within the context of the hotel industry. Afterwards, the relevant theory that leads to each hypothesis will be discussed. 2.1 Price promotions
The study of Oh (2000) examined the involvement of price in customers’ value judgments. It states that the hotel industry’s fundamental pricing strategy is merely based on the demand volume. This initially means that when demand exceeds supply, prices increase and the other way around (Steed & Gu, 2004). The most commonly applied room pricing model to cope with this demand-based pricing is called yield or revenue management. This can be defined as the process of finding the optimal combination of the right price and customer profile that would lead to the highest occupancy and maximized revenue (Kasavana & Brooks, 2001).
However, a well-known concept within this services industry is perishability which often is experienced as a challenge while aiming for the highest possible revenue. Unused capacity cannot be stored for later sale and directly becomes a missed opportunity to optimize the financial outcome in terms of Average Daily Rate (ADR): the ratio that considers the total availability of rooms in order to calculate the net financial performance (Steed & Gu, 2004). Also, the practice of demand-oriented pricing ignores the impact that the customer’s psychological evaluation has on the decision whether to buy the product or to switch to competitor products. To prevent rooms from being empty overnight, lots of hoteliers introduce price promotions to stimulate customer acquisition on the short-term, especially online (Piccoli and Dev, 2012, Yang et al., 2016). These promotions could be non-monetary, such as free gifts, but the traditional price promotion is still the most common form used by marketing strategists (Montaner et al., 2011). As shown in the study of Piccoli and Dev (2012), discounts that are most commonly used in terms of online flash sales (both through
9
intermediaries and own channels) lie between 45 and 55 percent for which one could argue that this is a substantial amount of money. This high percentage of discount might be explained by the introduction of flash sale websites that offer a variety of other products and/or services too (e.g. Groupon). However, besides the fact that these intermediaries demand higher discount rates, they also charge commissions (Piccoli and Dev, 2012). In comparison to third parties, Kang et al., (2007) have shown that the hotel’s own website channel is perceived as the most effective in terms of profitability and survivability. Regardless of the channel that is used, the aim of implementing price promotions is the short-term result of attaining (new) customers that would not have been triggered when the promotion would have been absent. A commonly applied presentation format of price promotions is comparative price advertising (McKechnie et al., 2012) where both the original selling price and the discounted price are communicated. This way of framing is known for effectively influencing the frame of reference by comparing the lower selling price with the original value, so that the monetary savings potentially create higher perceived value. (McKechnie et al., 2012; Grewal et al., 1989).
2.2 Purchase intention
Chandon et al. (2000) suggest that price promotions positively influence consumers’ purchase intentions due to beneficial monetary savings. Chiang and Jang (2006) confirm this theory within the online hotel booking context. They conclude that customer’s interest is likely higher when consumers perceive that the room price is more affordable than their internal reference price, which can be defined as a stable cognitive reference point (Oh, 2000). Because of that, the perceived value increases (Chiang & Jang, 2006) and consumers tend to express higher levels of purchase intentions and lower intention to search for alternatives (Oh, 2000). Evaluation of the offer is made by comparing the discounted price to the internal reference price (IRP), in which case the purchase intentions become higher when the
10
discounted price is actually lower (Chandrashekaran & Grewal, 2006). This implies that for a price promotion to be really effective in terms of purchase intention, it needs to be lower than what the consumer would expect. Moreover, previous researchers demonstrate that price promotions lead to higher purchase intentions because it lowers the risk for product trial (Kim et al., 2014) and it decreases the required efforts in the purchase decision-making (Chandon et al., 2000). The decreased sacrifice due to discounts also complies to the theory of planned behavior. This theory explains that behavioral intentions are dependent on three aspects: attitude, subjective norm and perceived behavioral control (Armitage & Christian, 2003). Those that were not able to purchase the product without the promotion, might become able to buy it against the discounted price. As a result, they perceive higher behavioral control. When holding attitude and subjective norm constant, the increase in perceived behavioral control positively influence behavioral (purchase) intentions (Armitage & Christian, 2003).
Based on previous discussed literature, this study aims to confirm that the implementation of a price promotion relates positively to purchase intention. The posited hypothesis however, applies to the context of the hotel industry so that the presence of a discounted room offer will result in higher intentions to book this room.
Hypothesis 1: There is a positive relationship between a hotel price promotion and the level of purchase intentions.
2.3 Perceived quality
Perceived quality refers to the perception of the overall quality or superiority of a product or service (Keller, 2003). As price increases within a buyer’s acceptable price range, the perceived quality of the product increases (Dodds et al., 1991). This suggests that a price increase automatically indicates the improvement of quality. The price-quality relationship remained a popular topic in the following price-related studies. Grewal et al., (1998) argued in
11
line with the attribution theory and stated that consumers often explain the applied discount as a result of poorer quality. Though the directionality was consistent, they found no significant proof for this relationship. Building on this paper, Suri et al., (2002) researched the differences between fixed and discounted price formats and found significant differences for perceived quality (fixed: M = 7.33, discount: M = 6.56, t(32) = 2.52, p < .05.) which they explain by the evaluation of the discounted price offer being associated with less positive affect. As a result, price information is more thoroughly processed and the offer evaluation is focused on the product’s attributes that leads to lower quality perceptions than in the case of heuristic processing (Suri et al., 2002). Interestingly, Oh (2003) demonstrated that when consumers pay less than what they perceive to be a fair price or market price, they perceive the selling price as less expensive, while at the same time it elicits higher quality perceptions. In contrast, a number of later studies (Chang, 2009; Chiang & Jang, 2006) again found that perceived price is negatively related to perceived quality. The authors exhibited that when consumers perceive a price to be more affordable, they would expect a lower quality of service and subsequently the purchase becomes less valuable. Most recently, Ye et al. (2014) confirmed the price-quality relationship for the hotel industry and the e-commerce environment in specific, demonstrating again that price positively influences perceptions of (service) quality.
Building on the majority of the findings above, this study presumes that lowering prices by introducing price promotions would lower (affect) the perception of quality.
Hypothesis 2: There is a negative relationship between the presence of a hotel price promotion and the perceived quality of the hotel services.
12
2.4 Customer’s willingness-to-pay
In terms of pricing strategies, it is important to consider what the customer would be willing to pay for a product and/or service (Wardell et al., 2008). As cited by Wertenbroch & Skiera (2002): “knowledge of consumers’ WTP is crucial in estimating demand and designing optimal pricing schedules”. In business practice, the willingness to pay can be defined as the amount of money the consumer would spend for a product or service (Hofstetter & Miller, 2009). In theory, using this value as a selling price would be the most profitable. However, it is not as simple as that. Many hotels introduce discounted prices on a regular basis as a response to perishability (Steed & Gu, 2004) that might lower the revenue due to lower selling prices.
A common way of introducing a (monetary) price promotion is through comparative pricing strategies. This involves the comparison of the discounted sale price with the higher original price, often in order to (re)define the customer’s frame of reference (McKechnie et al., 2012). According to extant academic literature, the customer’s internal reference price can be defined as the price the consumer expects to pay and is important in price evaluation and the subsequent perceived value of the purchase (Montaner & Pina, 2008; Chandrashekaran & Grewal, 2006; Grewal et al., 1989). Based on stimuli in the price advertisement, buyers either adjust their internal reference price or accept the advertised reference price (ARP) to evaluate the value of the promoted deal (Grewal et al., 1989). Providing the frame of reference by stating the original selling price (ARP) will help to control (or influence) the actual monetary saving that the customer will perceive (Chandrashekaran & Grewal, 2006). However, according to Ariely et al. (2006), consumers are sensitive to arbitrary information and use random information that influence the valuation of goods and experiences, which in turn manipulates their decision-making process that is built to not violate the rules of consistency. Whereas comparative pricing helps to increase perceived value and thus increase purchase
13
intention on the short-term, Johnson and Cui (2013) confirm that the suggested discounted price also creates downward pressure on the internal reference price (IRP) and based on the theory of Ariely et al., (2006), this new anchor might also influence (future) decision-making. This would imply that a price discount could decrease the internal reference price, which then affects (lower) the amount that one is willing to pay for a certain product and/or service since price expectations have decreased. Therefore, this study will aim to explore the negative effects of a price promotion on the customer’s willingness-to-pay due to lower price expectations.
Hypothesis 3: There is a negative relationship between the presence of a hotel price promotion and the customer’s willingness to pay.
McKechnie et al. (2012) explain the increase in purchase intention as a result of the customer’s perception that the specific purchase offers superior value as a consequence of the reduced price. Based on these findings, it is expected that the increase in purchase intention positively influences the willingness-to-pay due to the fact that they have perceived superior value. Therefore, assuming that the previous relationship of hypothesis 3 is present, this study also presumes a mediating role of purchase intention for this direct relationship.
Hypothesis 4: The negative relationship between the presence of a hotel price promotion and the customer’s willingness to pay is mediated by purchase intention, so that this relationship is weaker for higher values of purchase intention.
14
On the contrary, when the price promotion leads to a decreased perception of quality as posited in hypothesis 2, customers will believe that the product they receive in return for what they pay become less valuable (Chang, 2009; Chiang and Jang, 2006). Assuming that the expected negative effect of hypothesis 3 is present, this study also presumes a mediating role of perceived quality for this direct relationship.
Hypothesis 5: The negative relationship between the presence of a hotel price promotion and the customer’s willingness-to-pay is mediated by perceived quality, so that this relationship is stronger for lower values of perceived quality.
2.5 Conceptual Model
A visual representation of the theoretical framework was created in Figure 2.5. This conceptual model shows the positive and negative relationships of the different constructs. Furthermore, an overview of the hypotheses can be found in Table 2.5.
15
Table 2.5 – Overview of the hypotheses
H1 There is a positive relationship between a hotel price promotion and the level of purchase intentions.
H2 There is a negative relationship between the presence of a hotel price promotion and the perceived quality of the hotel services.
H3 There is a negative relationship between the presence of a hotel price promotion and the customer’s willingness to pay.
H4 The negative relationship between the presence of a hotel price promotion and the customer’s
willingness to pay is mediated by purchase intention, so this relationship is stronger for lower values of purchase intention.
H5 The negative relationship between the presence of a hotel price promotion and the customer’s
willingness to pay is mediated by perceived quality, so that this relationship is stronger for lower values of perceived quality.
16
3. METHOD
This section on methodology presents the research methods and approaches that were used for the empirical study with regards to design, sampling and measurements. Furthermore, it shows how the data was prepared for further statistical analysis.
3.1 Research Design
In order to test the posited hypotheses, it is important that the exact relationship between the proposed variables can be examined. In order to do so, the researcher should be able to apply an intervention while other conditions remain constant (Saunders et al., 2012). For this reason, an experimental vignette study was executed that offers the advantages of high internal validity by means of controlled manipulation, as well as external validity through multivariate measurements of the variables of interest (Aguinis & Bradley, 2014; Atzmüller & Steiner, 2010). This is done through an online questionnaire with a between-subjects design that rules out possible spillover effects (Saunders et al., 2012). This implies randomization of participants to either the experimental or the control group so that differences between the groups’ characteristics are prevented.
First, every participant was asked when he or she last stayed in a hotel. Afterwards, a scenario was presented that described the same situation for all participants, including a fictive luxurious hotel brand (see appendix 2). This scenario was previously used in a study by Yang et al., (2016) and already controlled for potential confounding factors like attitude towards the brand. Then, subjects were randomly allocated to either the control or the treatment group. A fictive design of the hotel’s website (as introduced in the scenario) was presented for both groups with a price offering. However, the treatment group were offered a discounted room price based on the frequently used percentage by hoteliers as shown by Piccoli and Dev (2012). All other conditions were equal. Following, the subjects were asked
17
to answer all the questions that are used to measure the variables needed for testing the hypotheses. Lastly, demographic questions were asked to sketch a profile of the sample. In order to decrease the chances for missing values after completion of the experiment, the option to force responses was used when applicable (Saunders et al., 2011).
To check for clarity of the scenario and the effect of the manipulation, a pre-test was executed for a small group of people (n = 14). Also, a small-scale preliminary study was conducted in the form of a pilot-test to gather data that was used for testing reliability of the measures and completeness of the design (n = 28). The final experimental survey is to be found in appendix 2.
3.2 Sampling
The population for this study is all people that are considered to be consuming since they are relevant potential guests of a hotel. However, it is not possible to test for the population as a whole and therefore a sampling method is required. According to Veal (2006), a sample should be a valid representation of the population and this can be done through random sampling. However, due to the lacking sampling frame, a non-probability sampling method had to be used (Saunders et al., 2012). It is important to realize that for this reason, implications of the findings are limited to the sample only and cannot be interfered to the population as a whole. Wilson VanVoorhis & Morgan (2007) state that for measuring group differences, a minimum of 30 participants per condition is required in order to decrease the chances of incorrectly rejecting the null hypothesis. Also, they state that in the case of regressions and correlational testing, the sample size is required to be around 50. Therefore, when adding up the two groups and to account for possible non-response, the sample size that was aimed for was 140 (Saunders et al., 2012).
18
In order to recruit the required amount of participants, a convenience sampling method was used. This is a relevant strategy when the individual cases are not hard to identify, when the required sample size is relatively big and lastly, when little variation is expected in the population (Saunders et al., 2012). As a result, the participants were acquired directly through e-mail and WhatsApp (direct approach) and by means of snowball-sampling through Facebook. Moreover, participants were approached in relevant public areas like hotel lobby’s. Inspired by the study of Yang et al., (2016), it was decided to add a screening criterion that would decrease possible response bias by selecting only the respondents for whom the most recent hotel stay was no longer than 12 months ago.
3.3 Measurements
The techniques that were used for measuring the constructs were based on former research. The reason for this is that previous studies already tested the scales on internal consistency by aiming for a Cronbach’s Alpha of at least 0.7 (Field, 2013). This increases the chances of having a measurement scale that results in reliable findings within this particular research. Following, the measured variables are discussed and the former use of measurement scales is summarized in appendix 1.
3.3.1 Purchase intention
This variable aims to define the level of purchase intention after being introduced to a certain price offer. The posited hypothesis assumes that the measured level of purchase intention is higher in the case of a price promotion (treatment group). This variable was measured using a 7-point Likert scale (Strongly Disagree to Strongly Agree) on a 4 multi-item question, previously used by Chiang & Jang (2006) who reported a Cronbach’s Alpha coefficient of .92. None of the items are reverse-coded.
19
3.3.2 Perceived Quality
This variable aims to define the quality perceptions of the participant, after being introduced to a certain price offer. It is expected that the measured level of perceived quality is lower for those exposed to a price promotion (treatment group). The variable was measured using a 7-point bipolar scale on a 4 multi-item question used by Chiang and Jang (2006) who reported a Cronbach’s Alpha coefficient of .96. None of the items are reverse-coded.
3.3.3 Willingness-to-pay
This variable aims to define the price a customer is willing to pay for a certain offering. The willingness-to-pay was researched by using the direct approach as discussed by Hofstetter and Miller (2009). This question aimed to elicit WTP directly by means of an open-ended question, asking for the amount (in €) that the participant is willing to pay for the offering which is similar to the study of Homburg (2005). It was expected that in the absence of the price promotions, subjects are willing to pay higher amounts of money than those that were previously exposed to a price promotion. However, there are some downsides to this measurement technique (Wertenbroch & Skiera, 2002). Subjects may be aware of the fact that there are no actual financial consequences of their answers, which might prevent them from stating their actual willingness-to-pay but reveal an amount that is higher. This is also defined as ‘hypothetical bias’. Nevertheless, this is still the most often used approach when time and money constraints apply (Hofstetter and Miller, 2009). Also, for this study, the mean difference will not be affected since both control and treatment group will suffer this hypothetical bias.
3.3.4 Control Variables
The experimental vignette study provides the advantage of systematically controlling for the confounding factors through conditional manipulation (Aguinis & Bradley, 2014; Atzmüller
20
& Steiner, 2010). Though it was previously mentioned that the convenience sample is an appropriate way to gather the required sample for this study, it makes it harder to control for the sample’s characteristics (Saunders et al., 2012). In case of a between-groups design, variety based on demographics between the groups is not desired and randomization does not guarantee the absence of potential bias (Saunders et al., 2012). The described approach for participant acquisition is vulnerable for self-selection, which might result in for example gender-biased information and thus decreased validity (Saunders et al., 2012). Therefore, this will first be examined manually when analyzing the sample demographics, and additionally two control variables are included in the correlation analysis. The two control variables that are assumed to be greatly diverse and therefore chosen for this research are Gender and Age. This should indicate if any further statistical findings are potentially subject to either one of these two characteristics (Field, 2013).
3.4 Data analysis
The obtained data was explored to look for any errors that could substantially influence further statistical testing. This section will discuss the findings and potential adjustments within the dataset.
3.4.1 Reliability of scales
The described multi-item scales were tested for internal consistency. The results are shown in Table 3.4. The Cronbach’s Alpha coefficients were above 0.7 and the results for ‘Alpha if item deleted’ showed that this would not increase internal consistency (Field, 2013). In terms of corrected item total correlations, values lower than 0.3 may indicate that the item is measuring something different from the scales as a whole (Pallant, 2010), but this was not the case. In conclusion, no changes were needed to increase reliability of the scales.
21
Table 3.4 –Analysis Data for Internal Consistency Cronbach’s Alpha Alpha if item
deleted
Corrected Item Total Correlations
Reliable? Purchase Intention .904 lower for all items all > .724 Yes
Perceived Quality .955 lower for all items all > .871 Yes
3.4.2 Assessing outliers
Prior to statistically testing the data, it is important to check for the presence of extreme scores that are well above or well below the majority of other cases (Pallant, 2010). These so-called outliers have been investigated first by calculating the 5% trimmed mean. Table 3.5 shows that removing the top and bottom 5% of all cases did not affect the mean with more than 1% (Pallant, 2010). On top of that, boxplots were created that did not show any individual cases of extreme scores. Only the data for ‘perceived quality’ included one outlier but after examining the z-scores, it was not larger than |3| and thus again proves non-significance in influencing the mean (Field, 2013). Based on these findings, it was decided that no changes within the dataset were needed.
Table 3.5 – Outlier analysis
Mean 5% Trimmed
Mean
Std. Deviaton Minimum Maximum Purchase Intention Control Treatment 3.64 5.04 3.64 5.08 1.47 1.08 1.00 2.00 6.25 6.75 Perceived Quality Control Treatment 5.39 5.14 5.47 5.19 1.16 1.08 1.75 2.25 7.00 7.00 Willingness-to-Pay Control Treatment 144.56 115.38 143.74 114.80 37.134 34.815 70 45 240 225
Note: Control: N = 54, Treatment: N = 50. Scores of purchase intention and perceived quality are measured on a 7-point scale.
3.4.3 Assessing normality
The independent-samples t-test that will be used later on assumes that the data that was gathered for the continuous variables are normally distributed. For this reason, the data was checked for the level of symmetry (skewness) and the shape of the distribution (kurtosis) (see Table 3.6). Scores for skewness and kurtosis for willingness-to-pay (.460; 057 respectively)
22
and purchase intention (-.426; -.876) were considered acceptable: according to Leech, Barrett, & Morgan, (2005), values between -1 and 1 are considered to be minor deviations from normal distribution and therefore reliable for further parametric testing. Furthermore, Pallant (2010) states that if sample sizes are reasonably large the results will not be vulnerable for small deviations. However, for perceived quality the kurtosis score exceeds this acceptable range (1.079 > 1) and indicates that the data is rather peaked and clustered in the center (Field, 2013). To double check for this finding, the normal Q-Q Plot and the histogram were observed and again confirmed this moderate level of non-normality. This was taken into consideration for further statistical testing with measures of perceived quality.
Table 3.6 – Outcome for Skewness and Kurtosis
n = 104 Skewness Std. Error Kurtosis Std. Error
Purchase Intention .-426 .237 -.876 .469
Perceived Quality -.955 .237 1.079 .469
23
4. RESULTS
This chapter will first describe the characteristics of the results for this particular sample by presenting the sample in general, followed by the average ratings per group. Also, the overall correlations are discussed between the relevant variables and those that are considered important as control variables. Afterwards, the relationships posited in hypotheses 1 to 5 will be tested.
4.1 Sample Demographics
As a result of the convenience sampling method, 139 participants were recruited to take part in the experiment. Only 113 of them actually completed the questionnaire (81%). Nine participants answered that their last hotel stay was more than one year ago and were therefore excluded from further analyses. The remaining 104 participants are still equally divided between the two experimental groups (control = 54, treatment = 50). Furthermore, the sample consists of 71 female and 33 male participants. This substantial difference could be explained by the sampling method (convenience sampling) and might indicate that females are more willing to participate in such an experiment, thus self-selection bias applies (Saunders et al., 2012). However, type of gender is equally divided between the two experimental groups. In total, 80 participants were employed from which 20,2% worked part-time. Others indicate to be retired (2,9%) or currently studying (19,2%). Also, most of the participants were between the age of 21 and 35 (65,4%) and also the majority indicated to have a HBO degree (higher professional education) or higher. A summarized overview of these statistics is to be found in Table 4.1.
24
Table 4.1 - Descriptive statistics of demographics
Frequency Percent Gender Male Female 33 71 31.7 68.3 Age > 21 21-35 36-50 51-65 66 or older 6 68 16 12 2 5.8 65.4 15.4 11.5 1.9
Education High school graduate MBO (Sec. Voc. Ed.) HBO (High. Prof. Ed.) WO Bachelor WO Master Doctorate 9 13 49 14 18 1 8.7 12.5 47.1 13.5 17.3 1.0 Total 104 100 4.2 Descriptive statistics
In Table 4.2, the calculated means are presented for both the control group that were not exposed to a price promotion, and those of the treatment group that were presented a promotion. In the case of a price promotion, the mean difference for purchase intention is +1.401 which is substantial on a 7-point scale. Also, those that were exposed to the price promotion perceived the quality as lower, however this mean difference is only 0.249. The results also show a negative difference (Mdiff = -29,18) of the amount that people are willing
to pay for the presented hotel room when a price promotion was previously introduced. Table 4.2 – Descriptive results per group
Mean SD N
Purchase Intention Control Treatment 3.639 5.040 1.467 1.084 54 50
Perceived Quality Control Treatment 5.389 5.140 1.158 1.079 54 50 Willingness-to-Pay Control Treatment 144.56 115.38 37.134 34.815 54 50 4.3 Correlation analysis
In order to discover the basic patterns between the variables, a correlation matrix was created and is to be found in Table 4.3. Here, the control variables are also checked for any influence. The results show that neither Gender nor Age are significantly correlated with any of the
25
measured variables. The expected relationship between the presence of a price promotion and increased purchase intentions is moderate1 and significant at 1% (r = .478, p = .000). Though there is a small mean difference, it is not significant for perceived quality (p = .260). However, again when looking at the presence of a price promotion, there is a significant negative correlation with the amount that the person is willing to pay at the significance level of 1% (r = -.378, p = .000). Between the measured variables, only purchase intention seems to indicate a small positive correlation with the amount that a person is willing to pay (r = .232, p=.018).
Table 4.3 - Means, Standard Deviations, Correlations and Reliabilities
M SD 1 2 3 4 5 6
1. Gender (0=male, 1=female) 0,68 0,47 x
2. Age (1=<36, 2=36-50, 3=>50) 1,42 0,72 -,088 x
3. Price Promotion (0=no, 1=yes) 0,48 0,50 -,130 -,085 x
4. Purchase Intention 4,31 1,47 ,050 -,005 ,478** (.904)
5. Perceived Quality 5,26 1,12 ,118 ,035 -,111 ,027 (.955)
6. Willingness-to-Pay 130,53 38,74 -,023 ,155 -,378** ,232* -,043 x
Note: N=104, reliabilities are reported along the diagonal when applicable. **. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
4.4 Statistical testing
In the following paragraphs, the previously stated hypotheses are tested with the help of different statistical techniques. For both purchase intention (H1) and willingness-to-pay (H3), it is important to examine actual differences between two experimental conditions that both had different participants which is possible through an independent-samples t-tests (Field, 2013). Due to previously concluded non-normality, the effect of the experimental manipulation on perceived quality (H2) is analyzed through a Mann-Whitney U Test that is not dependent on normal distribution (Field, 2013). Finally, the Process Macro by Hayes (2016) helped to execute a regression of the total model and examine if the direct effect on
26
willingness-to-pay is (partially) mediated through purchase intention (H4) and/or perceived quality (H5).
4.4.1 Hypothesis 1
Previously, it was concluded that there is a substantial mean difference of the level of purchase intention between the subjects that were not exposed to the hotel price promotion (M = 3.639, SD = 1.467) and those that were presented a discount (M = 5.040, SD = 1.084). An independent-samples t-test was conducted to compare the means. First, Levene’s test for equality of variances showed that the variance of the two groups were not the same. As a result, a significant difference in the mean scores of purchase intention was reported, t (97,315) = -5.57, p = .00. According to Cohen (1988) the magnitude of the differences in the means is very high (eta squared2 = .233). In conclusion, the hypothesis that a price promotion
relates positively to the level of purchase intention is supported.
4.4.2 Hypothesis 2
To test the relation between the presence of a price promotion and the expected decrease in perceived quality, a non-parametric alternative to the independent-samples t-test was necessary because of the non-normality of the data (Field, 2013). Instead of mean differences, this Mann-Whitney U Test compares the median and converts the scores to ranks so that the actual distribution has become irrelevant (Pallant, 2010; Field, 2013). The results revealed no significant differences in values of perceived quality between those that were offered a price discount (Md = 5,25, n = 50) and those that were not (Md = 5,50, n = 54) , U = 1158, z = -1.26, p = .209. Therefore, the negative relationship between a hotel price promotion and perceived quality is not proven and hypothesis 2 is rejected.
27
4.4.3 Hypothesis 3
The assumed relationship between the presence of a price promotion and the decreased amount that the subjects are willing to pay was again examined by assessing the mean differences between the control and the treatment group with an independent-samples t-test. Levene’s test indicated that equal variances of the two groups can be assumed. Based on this knowledge, findings showed a significant higher mean for those that were not presented any discounts (M = 144.56, SD = 37.134) than those that were manipulated with a discount M = 115.38, SD = 34.815; t (102) = 4.125, p = .00. The magnitude of the differences in the means (Mdiff = -29.18, 95% CI: 15.146 to 43.205) is large (eta squared2 = .14). Therefore, it shows
that as a result of communicating price promotions, the amount that subjects are willing to pay is substantially lower and therefore, the negative relationship as posited is proven. Thus, hypothesis 3 is supported.
4.4.4 Hypothesis 4 & 5
It was proven that there is a direct relationship between the presence of a price promotion and the willingness-to-pay. To examine this direct effect, hypothesis four and five consider a possible mediation through both purchase intention and perceived quality. Therefore, both hypotheses are examined simultaneously by means of the PROCESS macro developed by Hayes (2016). The results of the direct and indirect effects are to be found in Table 4.5 and 4.6, respectively. The used number of bootstrap samples for bias corrected bootstrap confidence intervals was 5000, for a confidence interval of 95%. The direct effect of the price promotion on the willingness-to-pay that was found for hypothesis 3, is also proven by this regression (B = -50.571) and was statistically significant (t = -7.239, p = .000). However, the total effect of the model is lower (B = -29.176), meaning that the variables together predict 29.176 units less in terms of willingness-to-pay for those that were presented a price
28
promotion earlier. This effect is statistically different from zero, t = -4.125, p = .000 with a 95% confidence interval lower than zero (-43.205 to -15.146). When looking at the variance of WTP in Table 4.4 , the direct effect explained 38% whereas the total model explained 14.3% of the dependent variable willingness-to-pay (R2 = .143: F(1, 102) = 17.015, p = .000). Table 4.4 – Model Summary for change in variance
Outcome: WTP R R2 MSE F df1 df2 p
Direct effect .6170 .3807 957.2589 20.4925 3 100 .000
Total effect .3781 .1430 1298.7952 17.0148 1 102 .0001
The direct effect between a price promotion and purchase intention was again present within this regression (B=1.401, p = .000). Also, the purchase intention can predict the amount that the subject is willing to pay as an independent variable (B = 14.466) which means that the higher the purchase intention, an increase of 14.466 follows for WTP. When looking at the indirect effect of a price promotion on the WTP, mediated by purchase intention (taking into consideration the parallel presence of perceived quality), we can conclude that this indirect effect exists (B = 20.269) and is significantly different from zero (BCI: 11.516 to 32.042). This implies that when the price promotion applies, the WTP changes
with 20.269 units, but solely as an indirect result of purchase intention. Therefore, hypothesis 4 is supported.
The results show again that perceived quality is not significantly related to both the presence of a price promotion, or the willingness-to-pay. The indirect effect (B=1.127) indicates the change in WTP for those where the price promotion applies, but solely as a result of measures in perceived quality. However, this effect is not statistically different from zero, as revealed by a 95% BC bootstrap confidence interval where zero is included (-.3631 to 6.421). Therefore, it was concluded that the effect of a price promotion on the amount that the
29
subject was willing to pay, is not mediated by the perception of quality. Thus, hypothesis 5 was rejected.
Note: N = 104. LLCI: lower level of CI. ULCI: upper level of CI. Effect scores are unstandardized and represent the original values of the DV which is recommended by Hayes (2012) when IV is dichotomous.
4.4.5 The results
Three of the five posited hypothesis were supported by statistical testing (see Table 4.6). First of all, it was strongly proven that the application of hotel price promotions helps increase the customer’s purchase intentions. On the contrary, price promotions do also decrease the amount that the customer is willing to pay for a hotel room. Also, this effect is mediated by purchase intention, so that the direct effect is weaker for higher values of purchase intention, since this positively correlates with the willingness-to-pay.
The two hypotheses that tested for the (low) effects of perceived quality within the model were not significantly proven. However, the direction of the effect (though not significant) was according to the expected relationship.
Table 4.5 – Testing for mediation / direct effects
Variables Effect SE t p LLCI - UPCL
Direct effects IV: Price Promotion DV: Purchase Intention
1,401 .254 5.513 .000 .897 1.905
IV: Price Promotion DV: Perceived Quality
-.249 .221 -1.124 .264 -.688 .190
IV: Price Promotion DV: WTP
-50.571 6.986 -7.239 .000 -65.15 -35.99
IV: Purchase Intention DV: WTP
14.466 2.649 5.460 .000 9.210 19.72
IV: Perceived Quality DV: WTP
-4.529 3.167 -1.430 .1558 -10.81 1.754
Note: N = 104. LLCI: lower level of CI. ULCI: upper level of CI. Effect scores are unstandardized and represent the original values of the DV which is recommended by Hayes (2012) when IV is dichotomous.
Table 4.6 – Testing for mediation / indirect effects
Variables Effect SE LLCI ULCI
Indirect effects
(mediation through:)
Purchase intention 20.269 5.266 11.516 32.042 Supported
30
Table. 4.7 – Outcome of hypotheses testing
H1 There is a positive relationship between a hotel price promotion and the level of
purchase intentions.
supported
H2 There is a negative relationship between the presence of a hotel price promotion
and the perceived quality of the hotel services.
rejected
H3 There is a negative relationship between the presence of a hotel price promotion
and the customer’s willingness to pay. supported
H4 The negative relationship between the presence of a hotel price promotion and
the customer’s willingness to pay is mediated by purchase intention, so this relationship is stronger for lower values of purchase intention.
supported
H5 The negative relationship between the presence of a hotel price promotion and
the customer’s willingness to pay is mediated by perceived quality, so that this relationship is stronger for lower values of perceived quality.
31
5. DISCUSSION AND CONCLUSIONS
Based on the findings of the previous chapter, this section will first discuss the main findings and compare it to what previous literature have stated. Following, conclusions are drawn and theoretical contributions are discussed. Finally, the managerial implications are discussed for business practice.
5.1 General discussion
An experimental vignette study was conducted to first examine the effects of hotel price promotions on purchase intention and perceived quality. The results suggest that hotel price promotions lead to higher purchase intentions than when monetary discounts are absent. This finding is consistent with previous studies that argue that consumers have higher value perceptions due to perceived beneficial monetary savings which leads to higher intentions to purchase the product and lower intentions to search for alternatives (Yang et al., 2016; Chang, 2009; Chiang & Jang, 2006; Chandon et al., 2000; Oh, 2000).
On the contrary, the results do not show significant effects as hypothesized for perceived quality and therefore do not lend academic support for previous literature stating that perceived quality is affected by the communication of lower prices through price promotions (Ye et al. 2014; Chang, 2009; Chiang & Jang, 2006; Suri et al., 2002; Grewal et al., 1998). This field of research generally explains this effect a simply attribution of the lower price to poorer quality of the product. However, the current experiment of this paper made use of comparative price advertising which involves comparing the discounted (lower) price with the original selling price. This advertised reference price (ARP) serves as a tool to raise consumer’s internal reference price (IRP) that determines the consumer’s evaluation of the offer (McKechnie et al., 2012; Chandrashekaran & Grewal, 2006). Hence, though this study does not support the negative relationship with perceived quality as concluded in previous research, it possibly complements to previous research that with comparative price
32
advertising, the original selling price (ARP) is used as the quality cue instead of the discounted price and therefore does not affect the perception of the product quality.
In addition, the study did show that introducing a hotel price promotion results in a significant decrease of the amount that the consumer is willing to pay for this discounted product. Previous studies have demonstrated that consumers’ price assessment depends on their internal reference price (McKechnie et al., 2012; Chandrashekaran & Grewal, 2006; Grewal et al., 1989) and also exhibited proof that price promotions affect this internal reference price (Johnson & Cui, 2013). The customer’s willingness to pay after being exposed to a hotel price promotion was not researched earlier and therefore this paper provides new insights in the field of hotel price promotions and its negative effect on price expectations. This finding is closely related to the field of consumer behavior, for which former research proposes that any (arbitrary) information can serve as an anchor that will be recalled from memory in future decision-making (Ariely et al., 2006). The discounted price affects the internal reference price that, in turn, is used as a comparison anchor to define expectations that influence the actual customer’s willingness-to-pay. Lastly, this paper investigated how the so-called immediate effects (short-term) on purchase intention and perceived quality, would influence the more long-term effect on willingness to pay. Similar to the non-significant effect on perceived quality, there was not found any reason to believe that perceived quality mediates the direct effect for willingness-to-pay. In contrast, this was the case for purchase intention so that the direct effect is weaker for higher levels of purchase intention. The study found that the decrease in willingness-to-pay due to the former price promotion becomes less for higher values of purchase intentions. Still, the total effect would remain negative so that the willingness-to-pay is always lower for the discounted room offer. By means of this finding, one could argue that based on theories of previous research (Johnson and Cui, 2013; Ariely, 2006; Suri et al., 2002) the internal reference price strongly
33
lowers the (future) price expectations and thus affects the amount that the customer is willing-to-pay. Moreover, the level of purchase intentions is not sufficient to compensate for this negative effect.
In conclusion, findings of this paper contribute to existent academic literature by providing new insights in previous researched relationships, in particular within the field of hotel price promotions. Firstly, it confirmed the previous stated relationship that price promotions in general are effective tools to increase purchase intentions and also complements to previous studies by means of the online context for the service industry (e.g. Yang et al., 2016). Furthermore, it raised the awareness of how perceived quality might only be affected in absence of the advertised reference price (ARP) which could also be used as a cue to evaluate the quality of the discounted product and/or service (e.g. Suri et al., 2002). Most important, the study shed new light on how the consumers’ internal reference price is not only affected by the price promotion, but that this also has direct implications for the customer’s willingness-to-pay and that this negative effect is reduced (mediated) by the direct increase in purchase intention due to the price promotion.
5.2 Managerial implications
Next to its theoretical contributions, this study also provides hoteliers with directions for future pricing strategies, based on the short-term goal to occupy the unsold inventory of hotel rooms through comparative price advertising. Again, this study demonstrates that the primary reason for hoteliers to work with hotel price promotions is present (Piccoli & Dev, 2012). This involves the introduction of a remarkable discount with the goal to acquire new customers by increasing their purchase intentions. However, it is known that hoteliers need to consider the threat of cannibalization so that revenue is not decreasing due to existing customers switching to sales promotions (Piccoli & Dev, 2012; Kang et al., 2007). Also,
34
previous studies have demonstrated that quality perceptions are dependent on price evaluation so that a higher price is associated with higher quality and vice versa. This study hypothesized that discounted products would result in lower perceived quality, but no significant results appeared in the case of comparative price advertising. Though not researched in comparison with other discounted pricing formats, this finding might imply that using comparative price advertising has the potential to not affect the quality of the hotel room perceived by the consumer. Based on previous literature (Chandrashekaran & Grewal, 2006), this research assumes that this is only due to the presence of an original selling price (ARP) that serves as the quality cue.
Further, this study distinguishes between immediate short-term results of price promotions and expected future implications. As mentioned above, the investigated short-term effects on purchase intention and perceived quality was considered to be a tradeoff but in the case of this research, where comparative price advertising was used, only the immediate positive increase in purchase intention was discovered. However, a more long-term consideration should be made by practitioners about the detrimental effects on the hotel’s future profitability and survivability (Kang et al., 2007). Besides the possibility that hoteliers do not know the potential implications of offering remarkable discounts, the popularity amongst hotel price promotions clearly show that they simply do not think about the long-term results either. Within the field of human psychology, this could be explained by the construal level theory (CLT) developed by Trope and Liberman (2010) that more distant events are thought of in more abstract ways. This study provides proof that customers that were exposed to the discounted hotel room, were willing to pay a remarkable lower amount of money for the offering, than when no discount was presented. Hence, this could hurt the profitability on the long run since profit margins will be lower for the same hotel room. In
35
terms of unsold capacity, future promotions become less effective because of the smaller deviation from the internal reference price.
As formerly explained by McKechnie et al. (2012), reduced prices make the consumer perceive superior values that partially accounts for the increase of purchase intention. This paper demonstrates that this superior perceived value compensates for the detrimental decrease in willingness-to-pay so that the higher the intention to buy the product, the lower the damage on willingness-to-pay. However, for practitioners, this finding would only be of interest when knowing the exact discount that involves the highest purchase intention, but the lowest damage to the amount that the customer is willing to pay. In conclusion, this paper helps to raise awareness amongst hoteliers and other practitioners within the field of marketing with regards to both short-term beneficial effects and long-term implications. However, further research is desired in order to be able to generalize the findings to real business practices and to provide exact values that lead to business optimization. The previously described shortcomings and other limitations are discussed in the next chapter followed by future research suggestions.
36
6. LIMITATIONS AND FUTURE RESEARCH
Though this study demonstrated interesting findings that provide academics with additional insights, one should be careful in terms of generalizability of the results due to potential limitations of the research method. This section will discuss these limitations and will provide future research suggestions
6.1 Study Limitations
Firstly, with regards to research design, the experimental vignette study involved a scenario-based condition that limits the external validity of the research outcomes scenario-based on its high hypothetical characteristic. After all, participants have to hypothetically imagine the situation they are in which might not result in measuring actual behavior or responses (Saunders et al., 2012). This applies also for the hypothetical website where the offer was shown and the hypothetical brand to control for any brand equity that could influence participant’s responses. In addition, the measurement of willingness-to-pay is known for its hypothetical restriction based on the lack of actual financial consequences (Wertenbroch & Skiera, 2002). This also counts for measuring purchase intention, since participants may find it irrelevant to stay in a hotel and therefore simply are not motivated when they are aware of alternative options (e.g. budget hotels, hostels, Airbnb). Though it would certainly improve generalizability when purchase intention or WTP would be measured in a real-life setting (field study), the limitation was applicable for both experimental and control group and therefore, the measured differences are still greatly internally valid.
Secondly, regarding the manipulation, the results of this study are limited to one type of sales promotions which was based on the monetary saving. Besides, the findings are difficult to infer to other industries outside the hotel sector. Also, the amount of the discount did not vary within the study and was only limited to one substantial discount of 45% based on the average percentage found by Piccoli & Dev (2012). Also, the framing of the discount
37
was not investigated and was solely limited to comparative price advertising with percentage-framing. In addition, the manipulation of this current study only researched the effects in the case of a luxurious hotel offering.
Thirdly, it is important to understand that the research method was cross-sectional, which implies that it tries to measure phenomena at one particular point in time (Saunders et al., 2012). For this reason, measuring the willingness-to-pay happened shortly after the exposure to the price promotion. This indicates another limitation with regards to generalizability since the long-term effect on WTP might provide different insights when measured for a different time horizon (e.g. longitudinal research). Also, Krishna et al. (1991) showed that the perceived frequency of the deal could influence the willingness-to-pay for the promoted product or service. In the case of this study, findings are again limited because the constructs were measured after only a one-time promotion of the product.
Another limitation of the study is the representativeness of the sample due to the majority of the participants being female (68,3%) and that it consisted merely of younger people between 21 and 35 (65,4%). This can be explained by the convenience sampling method where self-selection bias often occurs which results in limited representativeness (Saunders et al., 2012). Additionally, Yang et al., (2016) showed that the consumer’s need for status moderate the effect in returning intentions for a hotel after being informed that the hotel of their interest have introduced price promotions. These personal characteristics of the sample were not taken into consideration. Neither was the acceptable price range of each participant for a luxurious hotel stay, that is considered helpful in assessing and explaining deviations of the general patterns (Dodds et al., 1991).
Also, it is known that in case of small effect sizes, the sample size becomes more important to prove the existence of a relationship (Field, 2013). It may be for that reason that
38
the posited hypotheses with regards to perceived quality are correct, but that a higher sample size is needed to gather significant results. After all, small effect sizes can still have large implications for the population (Cohen, 1988).
6.2 Future Research Suggestions
The findings of this study suggests several interesting directions for future research. In general, the majority of existing literature within the field of sales promotions focuses on the more utilitarian products instead of hedonic goods and/or experiences. For this reason, it is recommended to further extend this research area, particularly for the services industry. To overcome the limitations of cross-sectional research, longitudinal research designs are desired to potentially gather new insights and to see if the effects of price promotions would hold for a different time horizon.
In particular for this study, replication is suggested while controlling for possible limitations as discussed earlier. To begin with, new conditions could be added based on frequency of the deal, discount size or discount framing. Also, the consumer characteristics could be considered like Ye et al., (2014) did with business versus leisure guests. The findings could be extremely helpful for practitioners to define the effectiveness of hotel price promotions and to aim for the least possible damage on the long-term, based on the most advantageous conditions or target group.
The non-significant effect of a price promotion on perceived quality is contradictory to the majority of research that state that price functions as a quality cue. For this reason, other pricing formats are recommended to see if different effects occur than in the case of comparative price advertising where the advertised selling price might be used as the quality cue. Also, future research might find significant outcomes for perceived quality when sample
39
sizes are greater, so that in the presence of a small effect size, it might also be systematically proven.
In addition, the hypothetical biases as mentioned previously might be solved by conducting a field study, with a real brand and real hotel customers (guests) that booked a price promotion earlier. This is particularly important for measuring the actual amount that the customer would be willing to pay, controlling for the fact that they have stayed in the hotel earlier against a price promotion. Nevertheless, a field study would have other limitations to overcome, especially in terms of internal validity.
