Exploring the Pathways to Tourist Loyalty in Pakistani Tourism Industry: The Role of Destination Image, Service Quality, E-WOM, and Social Media

Abstract

Tourism plays a pivotal role in Pakistan’s economic landscape. Understanding the drivers of tourist loyalty is paramount for sustainable growth. This study delves into the factors influencing tourist loyalty within Pakistan’s tourism sector, examining the destination image, service quality, electronic word of mouth, and tourist satisfaction. Data sourced from a tourist survey at popular Pakistani destinations unveil significant relationships and mediation effects. The results underscore that a positive destination image and superior services directly impact tourist loyalty. Furthermore, electronic word of mouth and tourist satisfaction mediate these relationships, highlighting how positive perceptions and satisfaction bolster loyalty. Social media also exerts a moderating influence on the destination image, service quality, and tourist loyalty dynamics. These findings offer crucial insights for tourism stakeholders and destination marketers to refine strategies. By enhancing its destination image, delivering exceptional services, and leveraging social media, Pakistan can both attract and retain tourists, bolstering its tourism sector’s growth and prosperity. This study contributes to the existing literature and offers practical guidance for promoting tourism in developing nations like Pakistan.

1. Introduction

The ability of a destination’s tourism sector to draw back tourists is crucial to its success. An increasing body of research is being conducted on the connections between the quality, DI, TS, and loyalty [1,2,3]. The atmosphere that visitors interact with while traveling might affect their enjoyment, desire to return, and DI. People choose their vacation locations based on where they think they would enjoy themselves the best. Service quality (SQ) and general place appeal are two of a vacation’s most crucial factors [4].

Tourist loyalty (TL) holds paramount importance in securing the long-term prosperity of a tourist destination. When tourists perceive comfort and a sense of belonging in a place, their inclination to return and advocate for it rises, resulting in cost savings for local tourism industry stakeholders [4]. This is particularly significant for prospective travelers who tend to make choices based on the feedback of past visitors. Consequently, positive encounters for tourists can enhance a tourist destination’s reputation and influence visitors’ intentions to return and advocate for it [5]. Enrique [6] states the ability to attract tourists is determined by its ability to provide high-quality tourism services that can adapt to the changing tastes and needs of international visitors. The effectiveness of a destination’s tourist sector is determined by the first visitor experience and the resources available [7]. Around 50% of a destination’s tourist flow comprises repeat visits [8,9]. Recurring visitors are critical to the profitability of a country’s tourism industry since attracting and retaining them is less expensive than attracting new ones [10]. Therefore, destination marketers must prioritize visitor retention and loyalty building to ensure a destination’s success [11,12]. Although destination loyalty is crucial for effective destination marketing, there has been little research into the factors contributing to TL in developing nations like Pakistan.

Scholars and tourism stakeholders have begun to emphasize the significance of the destination image (DI) and service quality (SQ) in boosting TL for a tourism destination’s sustainable growth. Visitors’ assessments of a location are affected by their perceptions of its attractiveness and facilities [13]. The degree of crime and the reputation of the place impact where tourists decide to travel since they are worried about their safety. Visitors will likely obtain negative impressions of the whole place when they experience bad service, deforested landscapes, and clogged roadways. Tourists’ favorable opinions of their chosen sites can be enhanced by destinations with high SQ (service quality) and DI (destination image) ratings. Providing well-preserved natural, historical, and cultural legacy resources helps achieve this [13].

According to the literature, the service quality experience, DI, E-WOM, and tourism satisfaction are all critical features of TL in a developing country context [13,14,15,16]. Nevertheless, past research findings on the links between the SQ, DI, and TL are uneven, if not contradictory. The DI influences tourists’ aesthetic choices rather than overall enjoyment or TL [17]. Kumail et al. [18] discovered no association between the SQ, tourists’ revisiting intentions, and customer satisfaction. However, they did discover a substantial beneficial influence of the DI on tourist satisfaction. In contrast to these findings, Henseler [19] claims that, in addition to the DI, the perceived service quality and pleasure significantly affect tourists’ revisiting intentions. TL was discovered as a potential predictor of the SQ, TS, and DI in these studies. However, several studies (e.g., ref. [20] have demonstrated that the SQ and DI are precursors of TS and E-WOM, both of which may have a significant influence on the TBI’s decision to revisit a destination. As a result of the contradicting results shown here, more study is needed to reexamine these relationships in a range of cultural and contextual circumstances.

Pakistan, an emerging country, holds significant untapped opportunities within its tourism sector. According to World Travel and Tourism Council forecasts, Pakistan’s tourism field is projected to yield as much as $39.8 billion by 2027 [21]. Magnificent tourist attractions, such as Pakistan’s breathtaking mountains, gorgeous hills, and thousands of kilometers of amazing deserts, can be found all around the country [22]. Furthermore, Pakistan’s religious potential is enormous. Muslim mosques, Hindu temples, Sikh gurdwaras, and Buddhist monasteries are the most important religious landmarks, which may be favorable for business. Moreover, its history as a once-great civilization only adds to its allure as a vacation destination [14]. Pakistan is also a popular mountaineering destination because of its high peaks [23]. In 2016, around 33.82 billion USD were brought into South Asia by international tourists. India (69%), Sri Lanka (10%), and the Maldives all made greater contributions than Pakistan. Understanding the factors that simultaneously shape TL will be aided by an investigation of Pakistan’s tourism industry, which could prove helpful to Pakistan’s tourism managers as they attempt to keep their current visitors satisfied while attracting new ones in light of the effects of COVID-19. This research advances our knowledge of tourism and tourist behavior in developing countries. This study adopts an interdisciplinary approach to amalgamate social media (SM) and destination marketing within the context of visitor behavior, mainly focusing on the tourism industry in developing nations. It addresses the burgeoning interest and the pressing need for further research on relational strategies. The research investigates how tourists’ perceptions of the destination image (DI) and service quality (SQ) influence their loyalty, shedding light on pivotal elements. Additionally, our study contributes to the understanding of tourist behavior by elucidating the influence of social media (SM) on the interplay between the DI, SQ, and tourist loyalty (TL). Furthermore, it explores the mediating role of tourist satisfaction (TS) and electronic word of mouth (E-WOM) in the relationship between the DI, SQ, and TL. While previous research has stressed the significance of the DI, SQ, and TL, there remains a dearth of empirical models quantifying the moderating role of SM and the mediating influence of TS and E-WOM within this triad. To the best of our knowledge, this study is the first to examine the connections between the DI, SQ, and TL, with SM acting as a moderator and TS and E-WOM serving as mediators. As the DI, SQ, TS, and E-WOM are believed to be the most influential factors in determining TL, it is particularly vital to investigate the mediating influence of TS and E-WOM in the unique social exchange context. However, there needs to be more information from studies examining whether SM moderates the relationship between the IQ, SQ, and TL, particularly in a developing country context.

2. Literature Review

2.1. Tourist Loyalty

Tourist loyalty (TL) has piqued the interest of academics and professionals in marketing and tourism management due to its favorable effects on revenue and profitability [24,25]. When we say “loyalty”, we mean the intention to buy or return to a product or service frequently and the persistent effort to maintain a relationship with a specific brand [23]. TL is defined by Govindarajo [26] as “a deeply held commitment to re-buy or re-patronize a preferred product/service consistently in the future, resulting in repetitive same-brand or same-brand set purchasing, despite situational influences and marketing efforts that may cause switching behavior”. According to a literature review, the demand for novelty influences the choice of tourism locations, with the complex selection process. Thus, TL is more difficult to build than traditional consumer loyalty [27]. Destination marketers also view TL as an essential attribute associated with a strong desire for less expenditure. This includes retaining current visitors instead of gaining new ones [28]. However, attracting repeat visitors is less expensive than attracting new ones [29].

2.2. Destination Image

Since the 1970s, tourism researchers have written extensively about the DI. When attempting to explain behaviors such as destination selection, RI, and satisfaction [30] and when developing destination marketing strategies, the idea of the DI can be pretty valuable [30]. The phrase “DI” refers to “a person’s entire collection of preconceived notions about a region” [31]. “A mental image of destinations exists regardless of whether these areas have been visited” [32]. The target image is dynamic, updating as new data are received [32]. According to Herrero-Crespo [33], a tourist’s view of a destination changes after visiting it. Similarly, Wantara [34] specifies that tourists form a more sophisticated image of an area due to their interactions and activities while on vacation. Returning visitors have a more positive view of a destination than first-time visitors [35]. This process of changing a destination’s image is continual, and an image will evolve due to events that occur throughout the visit [36]. A pleasant and memorable travel experience directly impacts a destination’s image [37]. Tourism-related events, such as sporting events, can also help to develop a positive DI [38].

2.3. Service Quality

Service quality (SQ) in service marketing refers to a client’s satisfaction with a service’s pricing, features, and overall value [29]. The spirit of quality is captured alongside standards, and expectations are met or exceeded [13,39]. SQ is increasingly recognized as a competitive advantage and tourism source in developing countries [3,37].

SQ is assessed to provide value for consumers by comparing their pre-service expectations to their post-service comprehension [27,38,40]. Various things might influence one’s expectations, including prior experience, word-of-mouth recommendations from other consumers, and promotional materials [41]. Several studies have been conducted on SQ to investigate the elements that influence consumer happiness and loyalty [42]. SERVQUAL has usually been accepted as the standard for measuring SQ in several service sectors [41]. This method has been chastised for supposedly misrepresenting tourism SQ [12]. As a result, we created a model tailored to the tourism industry in a developing country like Pakistan [27,39].

2.4. Theoretical Underpinning and Development of Hypothesis

The tourism sector frequently uses the quality–satisfaction–behavior paradigm to examine TL and intentions towards travel destinations [3]. The present study employed the Theory of Reasoned Action and the Theory of Planned Behavior [38] as theoretical frameworks to examine the phenomenon of TRI in diverse tourism contexts. In addition, researchers have examined the DI, SQ, E-WOM, and loyalty in conjunction with both frameworks [15]. Further study reveals that TS [2,40], SQ [41], DI [2], and E-WOM [42] recommendations can all increase TL [43]. The relationship between the SQ, DI, tourist satisfaction, E-WOM, SM usage, and tourists’ loyalty has yet to be examined empirically in the tourism literature. In order to evaluate the role of the DI, SQ, TS, and E-WOM on TL with the moderating effect of SM in the novel setting of the Pakistani tourism industry, hypotheses and a conceptual framework have been developed, as shown in Figure 1.

2.4.1. Destination Image, Tourists’ Satisfaction, and Tourists’ Loyalty

E-WOM has been used in both direct and indirect studies of the effects of the DI on TL [15]. Academic studies have shown that E-WOM is an effective advertising method for various products and services [44,45,46]. According to Król [47], E-WOM is an essential channel for consumers to learn about relevant topics such as SQ, brand products, travel experiences, and food. Many travel companies now use E-WOM to better serve their customers by disseminating information about their tours, hotels, and other services [48]. The literature on tourism services indicates that E-WOM positively affects repeat visits and customer loyalty [49]. In addition, E-WOM handles all marketing and advertising expenses, allowing you to increase the quality of your services and attract more customers [15].

E-WOM impacts tourists’ inclinations to plan trips, revisit places, and endorse tourism offerings, as it is recognized as a powerful motivator for forming intentions related to buying, revisiting, or suggesting [38]. Due to its association with visitor retention, E-WOM has emerged as a reliable predictor of the DI [50]. Some well-established relationships are highlighted between the DI and E-WOM and the DI and customer loyalty via E-WOM [48].

Despite the well-established relationships between the DI, E-WOM, and loyalty, there is disagreement about which aspects of the DI should be considered, how the concepts should be measured, and so on. Buhali [21] discovered that the DI influences E-WOM and, by extension, RI. Jeong [51] discovered that the DI significantly positively affects E-WOM, which influences TL and TS. In comparison, the DI has been shown to have a negative effect on E-WOM and TL [52].

Researchers have attempted to invert the equation or include additional variables. Using the consumer-based brand equity framework to test a multi-level model, [53,54] determined that affective and conative images directly influenced the perceived quality, customer satisfaction, recommendation, and brand loyalty. Other studies discovered that the DI influences happiness indirectly via variables such as DS and trust [36,55] or place attachment and emotional solidarity [30]. Nonetheless, there is consensus that the DI has a direct effect on TL [48,52]. Several studies discovered that the DI directly impacted TL, as measured by future visits, referrals, and word of mouth [48,53]. Based on this overwhelming evidence, the hypotheses of the present study are:

Hypothesis 1 (H1): 

DI has a significant association with TL.

Hypothesis 2 (H2): 

DI has a significant association with E-WOM.

Hypothesis 3 (H3): 

E-WOM has a significant association with TL.

Hyopthesis 4 (H4): 

E-WOM mediates the relation between DI and TL.

2.4.2. Service Quality, Tourists’ Satisfaction, and Tourists’ Loyalty

Several travel and tourism literature studies have found a strong correlation between PSQ, satisfaction, and TL in a range of tourism contexts [41,56]. Empirical proof that PSQ positively affects TS, which in turn affects loyalty, was found by Hamanishi [29]. Furthermore, TS mediates PSQ and RI [29]. Previous studies [9,57,58,59] have established a link between the TSQ experience, satisfaction, and BI or loyalty in a variety of tourist settings.

However, the study of SQ, TS, and loyalty in the context of the tourism industry in a developing nation has yet to produce conclusive results. Contrary to Widayati et al.’s [35] assertions, the likelihood of a revisit or referral from a traveler with a positive BTI is independent of the quality of their trip. According to Li et al. [56], PSQ did not influence TL’s attitude towards the Cox’s Bazar beach in Bangladesh, but it did influence their satisfaction. According to Wu et al. [60], visitors were more satisfied with Bangkok if they had good experiences with service quality, facilities, safety, and sanitation. Those with a high PSQ score are more likely to return to a destination with sun and sand, such as Bangladesh or Spain [61]. Thus, the following hypotheses are proposed:

Hypothesis 5 (H5): 

Service quality has a significant association with TL.

Hypothesis 6 (H6): 

Service quality has a significant association with TS.

Hypothesis 7 (H7): 

Tourists’ satisfaction has a significant association with TL.

Hyopthesis 8 (H8): 

Tourists’ satisfaction mediates the relation between SQ and TL.

2.4.3. The Moderating Role of Social Media (SM) in the Link between DI and TL

The importance of a DI to tourists, as measured by a destination index, heavily influences their decision to visit or return. Numerous studies have shown that a positive DI is critical to the success of the tourism industry because it increases TS, positive E-WOM, and repeat visits [61]. For example, Beerl [28] discovered that visitors’ perceptions of a DI influenced their loyalty and desire to return. Because of the prevalence of social media (SM), user-generated content such as photos, reviews, and ratings can impact tourists’ perceptions and impressions of a destination [62]. Furthermore, the impact of the DI on TL can be influenced by social media. According to Moon et al. [63], social media can improve or degrade TL depending on usage patterns. Positive SM reviews and recommendations, for example, can boost the DI and TL significantly. Negative reviews and comments, on the other hand, can have the opposite effect, lowering the DI and TL.

Multiple recent studies support the hypothesis that SM use can weaken the link between the DI and TL [62]. Oliver [64] found that SM significantly moderated the association between DIS and TL. The positive effects of the DI on TL were amplified, in particular, among tourists who used social media more frequently.

Hyopthesis 9 (H9): 

Social media moderates the relationship between the DI and TL, such that at a higher level of SM usage, the association between the DI and TL will be stronger.

2.4.4. Moderating the Role of Social Media in the Relationship between SQ and TL

The tourism industry has expanded significantly over the years, and SQ has been identified as a significant factor in TL [65,66]. Researchers have started looking into how social media affects the relationship between SQ and TL in light of its growing significance. According to Philipp [67], using social media amplifies the benefits of SQ on TL. The study found that when tourists interact with other tourists and share their travel experiences using SM, SQ’s beneficial effects on customer loyalty are amplified. Social media provides a platform for tourists to express their opinions, obtain local insight, and connect with others with similar interests. Podsakoff et al. [68] found that using social media for trip planning and preparation strengthens the relationship between SQ and TL. Similar findings were made by Park [69], who discovered that SM usage positively moderates the relationship between SQ and TS, which affects loyalty. Businesses in the tourism industry must emphasize providing excellent customer service and engaging with customers on these platforms as more and more tourists rely on social media for travel planning and socializing.

Hypothesis 10 (H10): 

SM moderates the connection between SQ and TL, such that at a high level of SM usage, the association between SQ and TL will be stronger.

2.4.5. The Moderating Role of Social Media (SM) in the Relationship between DI and E-WOM

Technology’s advancement and the emergence of social media have drastically altered how people communicate and share information. E-WOM has become a powerful tool for consumers to exchange product-related information and convey their opinions. In the tourism sector, E-WOM has a significant impact on the DI, which influences tourists’ decision-making processes. However, the impact of E-WOM on the DI is not straightforward. Several factors may moderate this relationship, such as the source of E-WOM, the type of information provided, and the context in which it is shared. Social media, as a platform for E-WOM, has become increasingly popular in the tourism sector. The moderating effect of social media on the relationship between E-WOM and the DI is of interest to tourism scholars.

Academic studies have recently examined the moderating effect of SM in the link between E-WOM and the DI. According to Poon [70], the utilization of social media platforms such as Trip Advisor or Yelp for communication purposes resulted in a more pronounced influence of E-WOM on the DI. Consistent with this perspective, the study conducted by Oliver [64] revealed that SM serves as a moderating variable in the link between E-WOM and the DI. Furthermore, it is possible that SM platforms could exert an impact on the nature of the data that tourists disseminate. According to the study conducted by Priporas [71], review sites tend to emphasize functional aspects of travel, such as lodging and transportation, while travel blogs primarily emphasize experiential travel elements, such as cultural immersion and social interaction.

Moreover, the context in which E-WOM is shared on SM platforms may also moderate its impact on the DI. For example, Lua [22] found that the influence of E-WOM on the destination image is stronger when it is shared by credible sources, such as experts or well-known bloggers. The literature suggests that social media platforms may influence the type of information shared, the credibility of the source of the DI, and the context in which it is shared, which may moderate its impact on E-WOM.

Hyopthesis 11 (H11): 

Social media moderates the relationship between the destination image and E-WOM, such that at a higher level of social media usage, the association between the destination image and E-WOM will be stronger.

2.4.6. Moderating Role of Social Media in the Relationship between SQ and TS

Social media platforms have changed how people communicate, share information, and plan vacations. Travelers may now instantly exchange travel advice, ideas, and reviews with people worldwide thanks to the spread of social media platforms. Numerous studies have looked into how SM affects TSQ and TS severity. According to Enrique [24], giving tourists access to information about tourism services via social media significantly impacts their happiness. To improve SQ, tourism enterprises can use SM to gather customer feedback about their experiences. Another study by Chen [8] found that SM plays a moderating role in the relationship between SQ and TS. The research illustrated that SM strengthens the impact of SQ on TS by enabling the seamless exchange of information and feedback between tourists and those in the tourism industry.

Furthermore, a recent study by Parola [66] examined the moderating role of SM in the link between SQ and TS. The study found that SM enhances the effect of SQ on TS by providing tourists with personalized and interactive experiences through the SM platform.

Hypothesis 12 (H12): 

Social media moderates the relationship between SQ and TS, such that at a higher level of SM usage, the link between SQ and TS will be stronger.

3. Research Methodology

3.1. Research Setting

This study selected Murree, Gilgit-Baltistan, Naran Valley, and Swat Valley as Pakistan’s four most popular tourist destinations. The natural beauty, highest peaks, glaciers, freshwater, lakes, and valleys of these areas make them among the country’s most popular tourist destinations [1]. The chosen regions are steeped in rich cultural and ethnic diversity, drawing tourists globally. These localities boast some of the world’s highest resorts, natural lakes, ancient forts, and numerous pristine natural sites. The recent developments in the CPEC road and train networks promise heightened accessibility and, consequently, an anticipated increase in tourist influx. Five of the world’s fourteen tallest mountains are found in Pakistan’s Gilgit-Baltistan province, including the second-highest peak, K2. These regions are home to some of the world’s highest resorts, as well as natural lakes, ancient forts, and many other untouched natural locations that are sure to draw many tourists. Historically, these regions have registered robust tourist traffic. For example, Gilgit-Baltistan alone recorded over 0.4 million international tourists annually before the COVID-19 pandemic. However, the pandemic induced a sharp decline, with current estimations showing a decrease by 50%. To assist readers in visualizing the geographical context of these regions, we have included a map as Figure 2, pinpointing each area’s location. We chose these locations to conduct our survey in order to put our theoretical model to the test in the real world.

3.2. Measurement Instrument

This study focused on six critical areas of investigation among the respondents, including E-WOM, social media usage, TL, DI, SQ, and TS. Based on the provided scale, the response options are assigned numerical values as follows: “strongly disagree” corresponds to 1, “disagree” corresponds to 2, “doubtful/neutral” corresponds to 3, “agree” corresponds to 4, and “strongly agree” corresponds to 5.

SQ was measured using eight items from [72]. Enrique Bigné et al. [24] used four-item scales to measure the DI. TS was measured using five items [73]. Five items were used to measure TL [73,74,75] by asking respondents to rate their agreement with statements regarding their revisit, recommendation, and WOM. The E-WOM variable consists of five items adapted from the study of Aiken [62]. Black [42] used a three-item scale to assess social media usage.

3.3. Sampling Methods and Data Collection

We utilized a convenience sampling approach for this study, directly engaging tourists visiting the selected destinations between June and August 2022. Our team was positioned at various strategic points of tourist interest, including hotels, local attractions, and community centers in the regions of Murree, Gilgit-Baltistan, Naran Valley, and Swat Valley. Before participating, each tourist was briefed on the study’s objectives and assured of the anonymity and confidentiality of their responses. Out of the initial 748 responses, 274 came from the Murree region, 110 from Gilgit-Baltistan, 149 from the Naran Valley, and 215 from the Swat Valley. In total, 748 questionnaires were collected initially, but 28 were discarded due to missing data, incomplete sections, or non-participation. The remaining 720 sets were used as independent units in the final analysis.

3.4. Respondents Profile

Table 1. Demographic information.

Background	Categories	No. of Respondents	Percentage of Respondents (%)
Gender	Male	378	52.6
	Female	342	47.4
Marital status	Married	426	59.3
	Unmarried	294	40.7
Age	18–30	324	45
	31–50	352	49
	Above 50	44	6
Education	Secondary education	213	29.6
	Undergraduate	221	30.8
	Graduate	194	27.1
	Postgraduate	92	12.6
Income	Less than 50,000	295	41
	51,000–80,000	318	44.3
	81,000–100,000	107	14.7
Nationality	Domestic	631	87.68
	International	89	12.32
Employment status	Full time	396	55
	Students	207	28.8
	Retired	117	16.2
Frequency of visits	4 times or less	632	87.9
	5 times or above	88	12.1

presents the demographic characteristics of the respondents who completed the questionnaire. Within the group, the proportion of women (52%) significantly exceeded that of men (47.4%). The data indicate that 55.1% of the population was in a married status, and 38.7% of the population was in an unmarried status. In terms of economic stability and travel experience, individuals aged 31 to 50 exhibited higher levels compared to those aged 18 to 30 (45%) and individuals above the age of 50 (6%). A mere 12.6% of the population possessed doctoral degrees, although 29.6% had completed primary or secondary education. Additionally, 30.8% had attended college to some extent, while 27.1% held a bachelor’s degree or lower. In the survey, it was found that 41% of the participants reported a monthly income of less than 50,000 Pakistani rupees (PKR). Additionally, 44.3% of the respondents reported earning between PKR 51,000 and PKR 80,000, while 14.7% reported earning between PKR 81,000 and PKR 100,000. It is important to note that the exchange rate used in this study was USD 1 = PKR 210. The majority of visitors originated from the United States, accounting for 86.68% of the total, with the remaining 13.22% being visitors from other countries. The results of the survey indicate that a majority of the respondents, specifically 100 percent, were engaged in full-time employment in various roles. Among the remaining participants, students accounted for 28.8 percent, while retirees constituted 16.2 percent of the sample, representing the second- and third-largest groups, respectively. A significant proportion of participants (87.9%) reported visiting these destinations a maximum of four times, but a minority (12.1%) indicated having visited them five or more times.

4. Research Results

4.1. Non-Response Bias

This study investigates non-response bias by comparing respondents’ ages and genders. The survey received 378 male responses and 342 female responses. Furthermore, the respondents are divided into two age groups: those under 30 (324 cases) and those over 30 (312 cases) (396 cases). A multi-group analysis reveals no statistically significant difference between the two age groups (p > 0.05). The DI and E-WOM, on the other hand, have a statistically significant gender difference (p < 0.05). A moderation test is performed to investigate this difference further, which yields no significant results for either gender (p > 0.05). This finding eliminates the possibility of non-response bias in this study.

This study also divides respondents into those who use social media infrequently (n = 311) and those who use it frequently (n = 409). The findings show significant differences between the two groups (p < 0.05) and that these differences impact the relationship between the SQ and TS, and DI and TL. As a result, social media use is assumed to impact the study model moderately.

4.2. Common Method Bias Assessment

The data gathering methodology utilized in this study exclusively involved the use of binary yes/no inquiries. Harman’s single-factor test (CMB) was employed in order to investigate potential signs of methodological bias. An exploratory factor analysis (EFA) was performed on the complete set of 27 questions utilizing Harman’s single-component approach [6]. Based on our research findings, the variables that were obtained accounted for 82.11% of the total variation seen in the six rotational components. The initial component accounts for a mere 34.72 percent of the observed variation, falling short of the customary 50% threshold required for completing a variance analysis via traditional means. Hence, the data gathered for this study remain essentially unaltered by non-response bias.

4.3. Measurement Model Evaluation

A comprehensive analysis of the precision of the measurement model was performed using maximum likelihood estimation. Table 1 summarizes these findings. The CFA results confirm that the measurement model fits the data from the current study well; 2 = 873.369, df = 309, p < 0.001; 2/df = 2.826; RMSEA = 0.050, CFI = 0.969, SRMR = 0.026, TLI = 0.965; 2/df = 2.826. Convergent validity investigates an item’s convergent validity in relation to other constructs in the theoretical model. Standard factor loadings, CR values, reliability test results, and mean extracted variance were used to determine convergent validity [76,77]. The standardized loadings of all items exceeded the recommended threshold of 0.80, ranging from 0.824 to 0.938 [46,78] and exceeding the recommended level of 0.80. All model constructs’ calculated average variance extracted (AVE) values were deemed acceptable. AVE values for SQ were 0.718, TL was 0.744, TS was 0.823, DI was 0.783, E-WOM was 0.826, and SM was 0.803 (see

Table 2. Items, loadings, composite reliability, and average variance explained.

Constructs	Items	Loadings	CR	AVE
Destination Image	The scenery and natural features of vacation destinations	0.880	0.935	0.783
	Climate and weather at a tourist destination	0.886
	Tourists’ perceptions of the quality of a destination	0.880
	A place that is both exciting and fascinating	0.894
Service Quality	Cleanliness of tourist destinations and the neighborhood	0.852	0.953	0.718
	Outstanding accommodations	0.824
	The standard of food and drink	0.849
	Sanitation and the tidiness of services	0.849
	Well-groomed and hygienic service staff	0.847
	Courtesy and respect for servants	0.858
	Options for public transport to get there	0.849
	The standard of transportation services	0.853
Tourist Satisfaction	I truly enjoyed visiting this beach	0.919	0.949	0.823
	My needs have been fulfilled by coming to this beach	0.890
	This tour has surpassed what I anticipated	0.938
	I’m happy with the time and money I invested in this place	0.881
E-WOM	I frequently rely on the travel reviews of others to assist me in selecting captivating attractions	0.938	0.934	0.826
	Before embarking on trips to specific tourist spots, I commonly gather insights from online travel reviews written by fellow explorers	0.848
	Travel reviews are posted online to boost my confidence in visiting these tourist destinations	0.937
Social Media Usage	My friends on social media frequently ask me for recommendations on new vacation destinations	0.928	0.925	0.803
	I have interacted with friends on various social media platforms regarding potential holiday destinations	0.857
	The most remarkable holiday advice I’ve ever received came from my friends	0.903
Tourist Loyalty	I intend to return to the beach	0.867	0.936	0.744
	I am ready to invest both time and money to return to the beach	0.865
	My family, colleagues, and I are planning a trip to the beach soon	0.893
	I intend to share positive word about the beach	0.842
	I would advise others to explore this beach	0.845

). As shown in

Table 3. Discriminant validity.

Constructs	SQ	TL	TS	DI	E-WOM	SM
SQ	0.848
TL	0.320 ***	0.863
TS	0.317 ***	0.439 ***	0.907
DI	0.266 ***	0.409 ***	0.327 ***	0.885
E-WOM	0.308 ***	0.331 ***	0.338 ***	0.350 ***	0.909
SM	0.307 ***	0.317 ***	0.240 ***	0.325 ***	0.271 ***	0.896

Note: *** = p < 0.001.

, all of these values confirmed convergent validity. In addition, the squares of the correlations between the constructs were all less than the reported AVE values in

Table 4. Heterotrait–monotrait results.

	SQ	TL	TS	DI	E-WOM	SM
SQ
TL	0.319
TS	0.317	0.440
DI	0.265	0.411	0.327
E-WOM	0.317	0.335	0.345	0.358
SM	0.310	0.324	0.243	0.325	0.288

, indicating discriminant validity [79,80].

Table 5. Results of direct relationships.

Hypothesis	Path	Beta Weight	SE	T-Value	p-Value	Result
H1	TL ← DI	0.258	0.039	6.616	***	Supported
H2	E-WOM ← DI	0.350	0.046	9.155	***	Supported
H3	TL ← E-WOM	0.118	0.031	3.082	**	Supported
H5	TL ← SQ	0.141	0.034	3.747	***	Supported
H6	TS ← SQ	0.317	0.043	8.289	***	Supported
H7	TL ← TS	0.297	0.031	7.799	***	Supported

Note: *** = p < 0.001; ** = p < 0.01.

summarizes the results of this study’s use of the heterotrait–monotrait (HTMT) method for evaluating discriminant validity [81]. The distribution of ratios less than 0.0 is considered valid for discriminatory purposes, and the upper limit of the ratio is 0.90. All six constructs can be discriminators because none of the value distribution numbers for the construct HTMT test is greater than 0.90.

The internal consistency of each latent construct was calculated by indicating the composite reliability (CR) values for all model constructs. SQ’s composite reliability values are 0.953, TL’s are 0.936, TS’s are 0.949, DI’s are 0.935, E-WOM’s 0.934, and SM’s are 0.925; these values are higher than the suggested cutoffs of 0.70 [82]. This suggests that the model’s latent constructs are internally consistent. The outcomes of this phase showed that every scale item in the suggested model is reliable and valid.

4.4. Structural Model Evaluation

In the structural model analysis, all path coefficients were estimated at once. Utilizing a mediator enabled us to isolate the indirect effect [83]. The results indicate that the structural model meets the proposed threshold indices of model fitness [83]. CFI = 0.956, TLI = 0.950, and RMSEA = 0.064; χ² = 965,716; df = 246; p < 0.001.

The results of testing the hypotheses were presented with a t-value and p-value (level of significance). In a quantitative empirical study, an idea is more likely to be accepted if its t-value is greater than 2.58 and its p-value is less than the significance level [83]. We observed a direct and statistically significant relationship between TL and the DI, consistent with H1 (0.258, p < 0.001). This lends credence to the idea that people will stay put longer if they have a positive impression of a place. Regarding H2, we found a direct and statistically significant effect of the DI on positive E-WOM (0.350, p< 0.001), demonstrating that a better opinion of a tourist destination is connected to overall positive word of mouth. The results thus supported the importance of a clear perception of the tourism destination in enhancing E-WOM for the destination as a whole. The findings demonstrated that E-WOM affects TL (0.118, p < 0.002) and that the higher the positive E-WOM, the more likely visitors will want to return to the location. Thus, H3 is supported. Hypothesis 5 is supported by the statistically significant (p < 0.001) correlation (0.0141) between SQ and TL. The direct and considerable impact of SQ on TL supports hypothesis 5 in this case. Additionally, the findings showed that SQ had a direct and significant impact on TS (0.317, p < 0.001), demonstrating that tourists are happier when they travel to a place that offers excellent service. This gives H6 much solid backing. As expected, the direct effect of TS on TL was statistically significant and positive (0.297%, p < 0.001). This helps H6 in some way.

R2 indicates how the independent variables can explain a given construct. With an R2 value of 0.236, we can conclude that the variables of interest (DI, E-WOM, SQ, and TS) explain 23.6% of the total variance in TL. The DI explains 12.3% of the conflict in the model, as indicated by the R2 values of 0.123 for E-WOM and 0.100 for TS, whereas SQ explains 10% of the variance.

This study used the AMOS mediation analysis with the bias-corrected bootstrapping method developed [78] at a 95% confidence interval to estimate the indirect effects and test the mediation hypotheses H4 and H5. The bootstrapping method (5000 iterations) was chosen because it is a non-parametric test that calculates indirect effects by repeatedly sampling the data set. The mediation analysis results (

Table 6. Results of mediation analysis.

Hypothesis	Path	Direct Effect	Indirect Effect	Lower Limit	Upper Limit	Result
H4	TL ← E-WOM ← DI	0.258 ***	0.041 **	0.016	0.071	Partial Med.
H8	TL ← TS ← SQ	0.141 ***	0.094 ***	0.066	0.132	Partial Med.

Note: *** = p < 0.001; ** = p < 0.01.

) confirm the indirect relationship between the DI and TL (indirect effect 0.041, CI 0.016 to 0.071) through E-WOM and SQ and TL (indirect effect 0.094, CI 0.066 to 0.132) through TS was significant, thus supporting hypotheses H4 and H8. As shown in Table 6, E-WOM partially mediated the effect of the DI on TL. Similarly, SQ affects TL through partial mediation of TS.

This study uses a moderation analysis using the [84] interaction approach in AMOS 24.0 to further explore the integrity of hypotheses 9 and 10. According to Chin. [85] and Beerli [86], the mean-centered dependent variables (E-WOM, TS, and TL), independent variables (DI and SQ), and moderating variable (SM usage) are used to address the potential for multicollinearity in the moderation outcomes. We first entered the data on the means of the independent variables before generating interactions (product terms) between the independent variables and the moderator to investigate how SM moderates the relationship between the DI and TL, SQ and TL, DI and E-WOM, and SQ and TS in AMOS 24.0.

The findings revealed that SM significantly and negatively moderates the relationship between the DI and TL, and SQ and TS, confirming this study’s moderation hypotheses. As shown in

Table 7. Results of moderation analysis.

Hypothesis	Moderation Analysis Relationships	Estimates	S.E	C.R	Decision
H9	ZC_TL ← SM_X_SQ	−0.037	0.033	−1.023	Not Supported
H10	ZC_TL ← SM_X_DI	−0.092	0.032	−2.467	Supported
H11	ZC_TS ← SM_X_SQ	−0.097	0.034	−2.599	Supported
H12	ZC_E-WOM ← SM_X_DI	−0.030	0.033	−0.805	Not Supported

, the interaction (β = −0.092, CR = −2.467, p < 0.014) of SM and the DI moderates the link between the DI and TL negatively, and the interaction (β = −0.097, CR = −2.599, p < 0.009) of SM and SQ moderates the relationship between SQ and TS, thus supporting hypotheses 10 and 11. However, the interaction (β = −0.037, t = −1.023, p < 0.306) of SM and SQ does not moderate the relation between SQ and TL, and the interaction (β = −0.030, t = −0.805, p < 0.421) of SM and the DI does not moderate the link between the DI and E-WOM, thus rejecting hypotheses 9 and 12.

Following the recommendations of Almeida-Santana [87], we also estimated the effects of low and high usage of SM on the relationship between the DI and TL, SQ, and TS (the accepted hypothesis). We drew short slopes for low and high SM usage to achieve this. The slope test reveals a significant relationship between the DI and TL in Figure 3 and Figure 4, but only at high levels of SM usage (β = 0.345, p < 0.001) and at low levels of information sharing (β = 0.135, p < 0.05). The relationship between SQ and TS is significant when there is a high level of SM usage (β = 0.386, p < 0.001) and a low level of information sharing (β = 0.185, p < 0.001). However, in both cases, the positive relationship between the independent variables (i.e., DI and SQ) and dependent variables (i.e., TL and TS) is stronger for the high level of SM usage and weaker for the low level of SM usage.

5. Discussion

The primary goal of this study was to investigate the impact of the SQ, DI, E-WOM, and TS on the retention of foreign tourists in a tourism environment typical of a developing country. This study’s empirical findings indicate that the SQ and DI have a significant, positive effect on TL (H1 and H5). Based on the available evidence, it can be inferred that offering exceptional services and additional benefits has the capacity to enhance the positive perception of a tourist destination among its visitors. The results of this study confirm the importance of the SQ and DI as dependable measures of TL with respect to the tourism industry in Pakistan. The significance of service quality in fostering tourist loyalty is paramount as it possesses the capability to impact the overall tourist experience. Tourists exhibit a tendency to exhibit loyalty towards destinations that offer high-quality services, as this has a favorable impact on their overall level of contentment. The referenced study of Andriani [88] revealed that SQ in China exhibited a noteworthy ability to predict TL. A research investigation carried out in Taiwan [83] revealed the existence of a favorable association between SQ and TL. The significance of the destination image (DI) in forecasting tourist loyalty is noteworthy. The depiction of a travel destination in visual form possesses the capability to establish an affective connection with tourists, consequently impacting their inclination to revisit. The loyalty of Korean travelers was found to be significantly impacted by the DI [89]. A positive correlation between the DI and TL was discovered in a study conducted by Mohamad et al. [84].

Numerous studies have been conducted in Pakistan on the variables influencing TL. A substantial association between SQ and TL was found in Pakistan by Hussain and colleagues [4] during their inquiry. In Pakistan, Widayati [35,90] found that the DI had a beneficial impact on TL.

The study revealed that the perception of the DI had a significant and favorable effect on E-WOM, as per hypothesis 2. As per the study, E-WOM holds substantial importance in shaping the behavior of tourists, including their decision to visit, level of contentment with the destination, and inclination to endorse the destination to others. E-WOM can have a significant impact on a potential visitor’s decision to visit a website, both positively and negatively. Several studies have linked E-WOM favorably with recurrent visitors. For example, E-WOM increased customer loyalty [88,90]. Similarly, it was discovered that E-WOM positively and substantially influenced TL (H3). In today’s digital era, when an increasing number of tourists rely on the Internet to plan trips and make purchasing decisions, the relationship between E-WOM and tourist loyalty is more vital than ever.

Several studies have identified a positive relationship between E-WOM and customer loyalty. For instance, a study conducted by [91] provides evidence that E-WOM significantly influences the likelihood of repeat visits to Hong Kong by travelers. Furthermore, Azhar et al. [92] have found a noteworthy association, indicating that positive E-WOM has a considerable impact on tourist satisfaction, which in turn affects TL.

The study also found that for Pakistani tourist destinations, SQ directly affects TS (H6), and TS indirectly affects TL (H7). These results align with previous research that has empirically demonstrated the relationship between TS and TL and between TS and SQ [11,13,29,65,70,92]. The results of the study show that SQ is a critical determinant of TS, which in turn influences TL towards tourism destinations in Pakistan. High-quality service can help create positive experiences for tourists, increase their satisfaction with a destination, and encourage them to become loyal customers.

The findings of this study also indicate that E-WOM mediates the relationship between the DI and TL within the context of Pakistan’s tourism industry (H4). Several prior studies have highlighted the significance of the DI in influencing tourist behavior, such as loyalty. For instance, a study discovered that the DI substantially affects TL in Spain [32,85].

The study also confirmed H4, the mediation effect of E-WOM in the link between the DI and TL. Therefore, considering online reviews and advice from other travelers can improve the usefulness of TL in a particular destination. These findings align with earlier research [82,92], demonstrating how Internet communication significantly affects tourists’ perceptions and actions.

It was found that TS mediated the link between SQ and TL in the Pakistani hotel sector (H8). The results are in line with the examination of travel. Research shows that TS is a critical mediator between SQ and TL [85]. The value of TS and its role as a mediator between TL and SQ has been shown in numerous studies [43,87,90].

Overall, these findings suggest that tourist satisfaction plays a crucial role in linking SQ and TL. SQ is a critical determinant of TS, which in turn influences TL towards a destination. Therefore, it is essential for tourism stakeholders to focus on providing high-quality services that meet the expectations of tourists and create positive experiences for them. Additionally, it is crucial to ensure that tourists’ satisfaction is continuously monitored and improved to maintain their loyalty towards the destination.

5.1. Theoretical Implications

There is very little analysis of the relationship between the SQ, DI, E-WOM, and TS and customer loyalty in the tourism literature. As a result, this study has several significant theoretical ramifications for studying tourist behavior, particularly tourism in developing countries. This study shows that TL is significantly influenced by both the SQ and DI. Although the SQ, DI, and TL have previously been used to investigate TBI in tourism contexts, little empirical research has been conducted to examine the effect of the SQ and DI as antecedents of TL in BI and the loyalty process in the tourism industry [80,88,89,92]. The current findings provide theoretical support for the idea that the SQ and DI are significant factors in shaping a positive perception of tourism destinations in Pakistan, even though their relationship with TL has received relatively little attention. This is true even though relatively little research has been conducted on this topic.

Secondly, the data show that SQ significantly affects TS regarding visiting destinations. Few studies have claimed that SQ significantly predicts TS behavior [72,73]. The impact of visitors’ satisfaction on their attitudes towards returning to travel destinations has not been thoroughly studied. In addition, the empirical analysis of the impact of the DI and E-WOM on the propensity of tourists to return to their previous destinations is novel and unmatched.

Additionally, for the two direct paths connecting the DI and SQ to revisit a previously visited tourist destination, the coefficient of the DI was higher than that of the SQ. This proposes that the DI is the most vital factor influencing visitors’ intentions to visit a particular destination again. The current research findings suggest that visitors’ satisfaction levels favorably impact their propensity to visit well-liked tourist attractions again. Despite the common assumption that satisfaction and attitudes are similar, few studies have examined their relationship [64]. These findings establish a theoretical basis for user satisfaction and attitude to revisit as independent variables in behavioral intention models by demonstrating a causal relationship between the two concepts under consideration. The significance of E-WOM in the decision-making process of tourists is demonstrated by the study’s fourth finding: the significance of the destination image in shaping tourists’ perceptions of destinations and their loyalty to them. Also emphasized is the significance of E-mediating WOM’s role in the relationship between a DI and the loyalty of its visitors. Fifthly, this finding emphasizes the significance of comprehending the interactions between offline and online travel experiences. A vacation involves more than just what tourists can do and buy. They also use digital platforms to communicate with and learn about the destination. This finding suggests that SM usage can affect the way tourists perceive and evaluate the SQ and DI, which, in turn, influences their satisfaction and E-WOM. It was also found that the literature, particularly about the moderating role of SM, hardly addressed the integrated relationship between the DI, SQ, TS, and E-WOM in a single model for the loyalty process. Although the relationships between these variables have previously been examined separately, the current study’s findings support the idea that they should instead be included in a single model.

5.2. Practical Implications

Currently, most tourist destinations compete for market share locally and internationally [60]. As a result, this study’s findings would provide policymakers and destination marketers for tourism in Pakistan and other developing countries with new information and practical implications. A higher service quality and DI are critical for increasing tourist revisit intentions (TRIs) in Pakistan’s tourism destinations. According to Saini. [74], enhancing the destination’s image and improving visitors’ perceptions of service quality are essential for retaining current visitors. In order to improve the reputation of tourism destinations, destination marketers must pay close attention to crucial service indicators, such as hygiene and cleanliness at the tourist destination and in the neighborhood, employee-oriented services, and food and lodging options. Additionally, this study found that E-WOM and TS were effective predictors of tourists’ destination loyalty. Tourism stakeholders should prioritize the promotion of positive E-WOM about the destination because visitor loyalty is crucial for the sustainability of the tourism industry. This can be performed by using a variety of tactics, such as encouraging visitors to share their positive online interactions, interacting with visitors on social media platforms, and responding to unfavorable remarks and reviews. This further implies that without improving the DI, TS, and E-WOM, the destination loyalty would decline [59,86,91,92]. The tourism destinations and general perception of developing countries like Pakistan as travel destinations are insufficient to attract foreign visitors. As a result, destination marketers should plan to enhance their destination’s reputation by highlighting its reputation as one of the world’s most beautiful mountainous regions, site tour attractions, and tourism destinations with distinctive beauty. This would boost visitor satisfaction, favorable attitudes, and place loyalty.

Given the demographic skew towards domestic responses in our study, practitioners aiming to attract foreign tourists should be cognizant of the diverse cultural expectations and preferences this demographic entails. Our findings, particularly those regarding the service quality and destination image, may require nuanced adaptation to cater effectively to a global audience.

5.3. Study Limitations and Future Research Avenues

Initial limitations on the scope of this study are imposed by the proposed loyalty model and suggested directions for pursuing it. Due to inherent cultural differences, every paradigm can only be universally successful and globally competitive [75,76]. The hypothesized relationship between variables in the proposed model should be reexamined in TD’s research. Returning customers may have a different experience than newcomers regarding perceptions and loyalty development. Any future studies on the variables influencing DL must include both types of visitors. Thirdly, people’s perceptions of the DI can vary depending on their political stance, religion, ethnicity, and culture [91]. As a result, future studies should classify images according to demographic factors like ethnicity, age, education, religion, and more. Fourthly, future research should incorporate separate measurements of the DI and SQ to increase the explanatory power of TL prediction models. As evidence suggests that social media and the Internet can induce behavioral and attitudinal loyalty, future studies should examine the relationship between the information sources used and destination loyalty [92]. Finally, each respondent was treated as an independent unit of analysis, so the tourist demographic profile and the multi-group sample were ignored. Shortly, another loyalty model in the same context will be empirically examined using mediating and moderating variables.

This study’s scope was initially confined by the demographic composition of our sample, which consisted predominantly of domestic tourists, with foreign tourists accounting for only 12% of respondents. The limited representation of foreign tourists, particularly the insufficient diversity in their countries of origin, introduces a potential bias that future research should aim to rectify. Subsequent studies could enhance the validity of the findings by involving a more globally representative sample, thereby capturing the multifaceted nature of foreign tourists’ expectations and experiences.

Moreover, our survey did not fully explore the cultural dynamics and specific regional preferences of foreign tourists, an aspect that future studies could investigate in depth. Diversifying the sample to include a broader spectrum of nationalities would provide more comprehensive insights and aid stakeholders in formulating strategies that resonate on a global scale.