1 Introduction

Social networks allow consumers to achieve real contact with specific people within the organization, and good customer service achieves numerous objectives, including increasing customer satisfaction and, therefore, customer loyalty. In this context, satisfaction is a causal factor for recommending and repeating the use or consumption of a product or service. Ribbink, van Riel, Liljander, and Streokens (2004) consider that social networks can complement customers’ perception of satisfaction and motivate them to communicate their experiences to others, which determines an opportunity for companies that wish to use this medium to improve satisfaction and strengthen loyalty.

It has been identified that companies make use of social networks not only to promote their products and/or services or to position their image but also for sustainability (Du, Yalcinkaya, Bstieler, 2016), whose meaning has been developed from different approaches (Brown, Hanson, Liverman, and Merideth, 1987), within the business context, sustainability is understood as environmentally friendly policies and actions in which economic development takes place in an environment of social balance (Martínez-Navalón, Gelashvili, and Saura, 2020), from this perspective the promotion of sustainability has nowadays become an important objective for companies in different sectors.

Sustainability is vital for business and society. Nevertheless, the evidence of its study is limited in developing countries such as Ecuador, a country that, like many others in the region, is highly dependent on the tourism sector for its income and which, in recent years, has been affected by unprecedented crises, such as the SARS-CoV2 pandemic and the consequences of the armed conflict between Russia and Ukraine. Tourism development has generally been considered a positive contribution to economic growth (Che Chou, 2013), so its analysis is always a topical issue.

Tourism activity in Ecuador is essential; during 2015–2018, an increasing influx of foreign tourists was reported, with 2.43 million foreigners visiting the country in 2018. At the end of 2019, this influx was reduced by 16% compared to the previous year (Ministry of Tourism, 2022). Most foreign tourists entering the country in 2019 came from Colombia (38%), the United States (20%), and Peru (15%). Tourism revenues reached USD 2.4 billion at the end of 2018 and represent the third-largest source of non-oil revenues (Ministry of Tourism, 2022). Data for 2020, 2021, and 2022 are atypical due to the consequences of the global impact crises.

Tourism companies materialize their activity in the satisfactory or unsatisfactory experience of their customers, which is why the relationship between the use of social networks and this economic sector is of particular interest (Martínez-Navalón, Gelashvili, Saura, 2020). For tourism, social networks are a fast and powerful way to connect with current and potential customers. Improving customer perception is also vital for developing the localities where tourism activity occurs and where social media’s role is relevant (Zhang and Zhang, 2018).

2 Literature Review

2.1 Social Networking, Tourism, Satisfaction, and Loyalty

Companies have started to use social media to interact, facilitate information search, promote, and improve customer buying behaviors (Zeng and Gerritsen, 2014). By the end of 2019, its use accelerated exponentially in response to the SARS-CoV2 confinement, leaving many questions about its harms and benefits (Yul Lee, Soo Yang, Ghauri, and Park, 2022).

Interactivity is one of the most relevant factors determining customer engagement in social media communities. Social media, in turn, has been shown to increase revenue, business efficiency, and the impact of promotions and advertising (Alalwan, Rana, Dwivedi, Yogesh, and Algharabat, 2017). The use of social media is directly linked to relationship marketing strategies, which seek to create a collaborative relationship with the customer and a greater sense of belonging to the company’s brand. Relationship marketing aims to achieve customer loyalty, which occurs when the individual is a favorable attitude toward the company when buying its products or services (Campón, Baptista, and Hernández, 2009). In the tourism sector, consumers have a higher degree of uncertainty due to the nature of the services and the consequences on their satisfaction and loyalty, which tends to be greater in comparison with other sectors, which is why the relationship marketing approach in tourism companies is particularly appropriate.

Trust has a positive effect on relationship commitment; if customers maintain a good relationship with the company in the long term, the trust they have in the company will generate or be a strong driver of loyalty, commitment, and loyalty in the customer relationship (Kassim and Abdullah, 2010). Therefore, it can be stated that customer satisfaction is strongly and positively related to the trust they have in a company, even more so in the case of tourism services.

2.2 Social Media Marketing and Sustainability in the Tourism Sector

Sustainability in the business world starts its evolutionary path from corporate social responsibility initiatives focused on marketing, being an integral part of the strategic planning system (Fuxman, Mohr, Mahmoud, and Grigoriou, 2022). However, it has several edges. One of these is environmental sustainability, which is related to the protection and preservation of natural resources and the development of renewable energy sources, and a balanced consumption with the environment and human development (Gelashvili, Martínez-Navalón, and Herrera-Enríquez, 2021). Signitzer and Prexl (2008) argue that sustainability relates to environmentally responsible products, social justice, and sustainable awareness.

By integrating the concepts of sustainability and marketing, it can be understood that price, product, place, and promotion actions will be oriented to satisfy three criteria: customer needs, company objectives, and the compatibility of processes with the eco-system (Lee, 2016), consequently relationship marketing focused on sustainability can make use of social networks as a mechanism to reach its customers and establish it as an effective strategy to strengthen its positioning around the concept of sustainable business (Kotler, Kartajaya, and Setiawan, 2010), however, to achieve this it must achieve customer satisfaction that will materialize in consumer loyalty.

Tourism is representative of many developing countries because it generates wealth and distributes it more equitably (Chica, Hernández, and Perc, 2022). Tourism has a direct relationship with the environment, society, and the economy, so the balance of these three components is essential; consequently, consumers of these products and services are aware that this integral and harmonious relationship must be part of the offer (Zhang, Zhong, and Yu, 2022). Evidencing the sustainable exercise of tourism is a challenge for companies dedicated to this sector, so communication through social networks is presented as an alternative to strengthen satisfaction and generate consumer loyalty.

3 Hypothesis Development

The development of the hypotheses considers the different studies that evidence relationships between marketing and environmental sustainability that is achieved through the development of innovative and competitive products and services that generate customer satisfaction and trust (Walch and Dodds, 2017). It would be expected that the results in the case raised confirm these statements. The variables of environmental sustainability and consumer satisfaction allow the following hypothesis to be put forward:

Hypothesis H1. Publications on social networks related to environmental sustainability directly and positively influence user satisfaction. According to Martínez-Navalón, Gelashvili, and Saura (2020), the relationship between the satisfaction of users of social networks related to tourism companies in Spain has a positive effect on the trust of companies, so it would be expected that this relationship is present in the case raised, so the following hypothesis is proposed:

Hypothesis H2. The satisfaction of social network users with tourism companies directly and positively influences their trust in these companies. The literature on environmental sustainability, satisfaction, and loyalty was carried out. The following study model (See Fig. 1) is proposed, as well as the hypotheses proposed (See Fig. 1).

Fig. 1.
figure 1

Research model

Full size image

4 Methodology

4.1 Data

The study uses a questionnaire that has been developed based on the literature review related to the subject. The questionnaire was carried out in person at different points in Quito (Ecuador). The specialized nature of the data collection made it possible to analyze users directly after the pandemic and to gather their evaluations more accurately.

The questionnaire was applied to 2850 people; 50 were discarded because the information was incomplete. It should be noted that the data collection is completely anonymous, and it is impossible to identify the respondents’ origin a posteriori, which allows us to preserve their anonymity and comply with the different data protection regulations.

For the collection of the sample of 2800 valid questionnaires, a questionnaire was used, which is divided into two parts. Firstly, questions were asked to classify the individual, these being classificatory questions. Secondly, individuals are asked questions about the three variables analyzed and which allows us to measure the level of sentiment in each of them.

A questionnaire is used to measure the sentiment in the questions and therefore be able to carry out a complete analysis (Gelashvili, Martínez-Navalón, and Herrera-Enríquez, 2021) used, whose questions are on a Likert scale. This scale is structured with a measurement where five is “strongly agree,” and 0 is “strongly disagree.“ This scale is considered one of the most reliable and valuable in Social Sciences (Alismail and Zhang, 2020).

The variables analyzed in the questionnaire were environmental sustainability, satisfaction, and loyalty. For each of the variables, different questions were asked to allow the measurement of each variable. The questions were obtained from a thorough literature review.

Table 1. Sample characteristics (n = 345)
Full size table

Table 1 shows the characteristics of the respondents. These characteristics are obtained by classificatory questions that allow us to classify the respondents as a posteriori. Firstly, the sample is very even in terms of the sex of the respondents. Slightly more women than men (50.36% and 43.36%, respectively). Secondly, it is worth noting that most of the users surveyed reside in a population of over 100,000 inhabitants (43.37%), followed by 30.11% of the individuals who have their habitual residence in a population of between 20,000 and 100,000 inhabitants. It can be deduced that most respondents reside in large towns. Thirdly, the sample comprises mainly self-employed individuals (40.36% of the sample), followed by employees (35.86%). Fourthly, the level of education of the sample, in general, is high, with 35.89% having a university education. Lastly, an important data point in this study is the number of minutes individuals spend each day on social networks. Most respondents spend more than 2 hours a day on social networks.

4.2 Method of Data Analysis

In analyzing the variables proposed in the model (Fig. 1) and after having carried out the theoretical approach, we proceed to analyze both the variables and the relationships between them. For these analyses, the partial least squares (PLS) technique is used in this study. Using structural equations based on variances is one of social science studies most widely used techniques.

It should also be noted that it is one of the most widely used techniques in studies where the measurement of the respondents’ feelings is carried out using a Likert scale for data collection (Del-Castillo-Feito, Blanco-González, and González-Vázquez, The relationship between image and reputation in the Spanish public university. 2019). The software used in this case is SmartPLS, which allows us to assess the reliability and validity of the different measurement scales and test the structural model (Henseler, Ringle, and Sarstedt, 2015). Its choice is based on its capacity for the graphical resolution of the study and the set of possible statistical methods that can be applied during the research (Cachon Rodriguez, Blanco-González, Prado-Román, and Diez-Martin, 2021).

This study evaluation technique allows measuring the simultaneous behavior of the relationships proposed as a dependent. It also allows for very comprehensive multiple regression and factor analysis studies. Relationships can be studied without having to fix the relationships of the influences in the hypotheses, allowing for a more complex and varied study (Liengaard et al. 2021).

5 Analysis of the Results

5.1 Measurement Model

Once the study sample has been analyzed, we proceed to analyze the proposed model. Before starting, this analysis is divided into two processes. In the first process, the study of the scale of measurement proposed must be carried out. In this analysis, the questions used to measure the variables are checked to ensure that they are correct and, therefore, valid for the analysis (Gelashvili, Vera, Martínez-Navalón, and Saura, 2021). In the second process, once the scale has been validated, the hypotheses set out in the model are analyzed (Fig. 1), as the predictive capacity of the variables and the model in general (Martínez-Navalón, Gelashvili, and Saura, 2020).

In the first of the analyses described above, all the variables in the study are reflective. This characteristic indicates that the analysis process of the measurement scale should have the following studies: individual reliability, composite reliability, convergent validity, and discriminant validity. The data obtained from these analyses will make it possible to validate each of the questions asked in the questionnaire and, therefore, validate the scale. These analyses can be seen in Tables 2 and 3.

In the individual reliability analysis, each question’s loadings (λ) must be studied. To be validated, the value of the loadings must be higher than 0.707 (Carmines and Zeller, 1979); (Del-Castillo-Feito, Cachón-Rodríguez, and Paz-Gil, Politi-cal Disaffection, Sociodemographic, and Psychographic Variables as State Legitimacy Determinants in the European Union, 2020). In this analysis, all the items proposed to pass the test in the case of the composite reliability study where Cronbach’s Alpha should be analyzed following the criteria of Nunnally and Bernstein (1994). The cut-off level of the Alpha is 0.707. Once this process has been analyzed, it could be concluded that there is high reliability in the study. However, as it has a variable above 0.85, it is advisable to carry out the study (Dijkstra and Henseler, 2015). In this analysis, the ratio (rho_A) is analyzed, and the cut-off value of the variable is set at 0.7 (Gelashvili, Vera, Martínez-Navalón, and Saura, 2021). All the items proposed in the study have passed the three analyses carried out so far.

Fourthly, convergent validity analysis is carried out, that is, the analysis known as average variance extracted (AVE). This analysis, carried out according to the criteria of (Fornell and Larcker, 1981), sets a cut-off point for AVE at 0.5. This cut-off indicates that the variables must have at least 50% of the explanation of the underlying variables (Hair, Risher, Sarstedt, and Ringle, 2019).

Table 2. Measurement items
Full size table

To complete the validation of the measurement scale, it is necessary to carry out a final process, that of discriminant validity. This process analyzes the variable’s variance from the items that make it up. The contribution of the items must be more significant than that which these items may share with other variables in the model (Hair, Risher, Sarstedt, and Ringle, 2019). The Fornell and Larcker (1981) criterion has traditionally been used to carry out this study. Many researchers in their work have used these criteria. However, it has recently been found that it is not a very demanding study, so Hair et al. (2019) recommend using another more recent and more demanding criterion that allows for greater robustness in this analysis (Del-Castillo-Feito, Blanco-González, and González-Vázquez, The relationship between image and reputation in the Spanish public university., 2019). This second criterion is the Heterotraitmonotrait ratio (HTMT). This analysis sets the cut-off index at the maximum value of 0.9. These indicate that the variables in the model analyzed would be empirically different (Dijkstra and Henseler, 2015). Once the analyses have been carried out, it can be seen in Table 3 that all items meet both criteria.

Table 3. Measurement discriminant validity
Full size table

Finally, it can be affirmed that the measurement scale proposed in the model and made up of the questions on the variables environmental sustainability, satisfaction, and loyalty is a valid and reliable scale of measurement. The questions asked in the study are valid for measuring the proposed variables and the relationships between them.

5.2 Structural Model Analysis

To finalize the analysis of the results, a study of the model must be carried out, studying the influences proposed in the model and its prediction. In order to carry out this study, a bootstrapping of 50,000 samples will be performed. This analysis allows standard errors and t-statistics to be obtained (Cachon Rodriguez, Blanco-González, Prado-Román, and Diez-Martin, 2021).

Before starting with the analysis, it is necessary to analyze whether there is multi-linearity in the structural model. This analysis applies the VIF value criterion that shows whether such collinearity exists in the model. If values of less than five are obtained, it can be affirmed that there is no multicollinearity in the model (Gelashvili, Vera, Martínez-Navalón, and Saura, 2021). Once this criterion has been applied, the relationships of the study have values equal to 1, so there is no multicollinearity, and we can proceed to measure the relationship between the variables and the predictive power of the model.

In the analysis of the model, we can see how both hypotheses are accepted, as we can see in Table 4. In the same way, we can see how the original sample of the relationship is high, as well as the t-statistics being 37.619 for H.1 and 57.871 for H.2, which indicates the strong influence of the relationship. Similarly, the third criterion for validating the relationships is confidence intervals. These also show the validation of the relationships satisfactorily.

Regarding the explained variance (R2) of the model, it can be seen that the prediction results are promising. Both explained variances have a medium predictive power (Chin, 1998). Similarly, the effect size (f2), the degree to which an exogenous variable contributes to explaining an endogenous variable, has a significant effect, and the predictive relevance of the model (Q2) is obtained by blindfolding analysis. This shows that the model has a medium predictive validity, being very close to a high predictive relevance (Hair, Risher, Sarstedt, and Ringle, 2019) (Fig. 2).

Fig. 2.
figure 2

Proposed research model

Full size image
Table 4. Comparation of hypotheses
Full size table

6 Conclusions

The study confirms the direct and positive relationship between environmentally sustainable content on social networks and the level of customer satisfaction of the companies analyzed so that the decisions of tourism companies related to environmentally friendly actions and their communication through social networks can generate a positive image that could be capitalized with higher revenues because of high levels of satisfaction. Ecuador is a country that presents itself to the world as a natural destination, which is why foreign tourists take sustainability very much into account as a factor in choosing a destination or a tourism company.

Although there is theoretical and empirical evidence on the relationship between satisfaction and trust, this study presents a particular approach by analyzing the satisfaction of social network users and its direct and positive relationship with the trust they may have in the tourism companies studied in Ecuador. The design of communication strategies through social networks should consider that these are a way to develop trust in customers, strengthening their positioning and the quality of their relationship with their customers.

This study has limitations regarding the sample design, which was not probabilistic, and therefore does not allow inferences to be made about the population. Despite this, the number of companies studied is relevant, as the object of study, which was the city of Quito, is considered one of the most representative heritage cities in Latin America.