Keywords

1 Introduction

Currently, Social Media (SM) is fully integrated as an indispensable part of the commercial strategy of tourist destinations. Following the decision-making process model proposed by Goodall et al. (1991), it is known that in the final phase, the traveler chooses among destinations that achieve a comparative advantage among those known to them. Therefore, the mere presence of destinations on SM, while essential for a proper communication strategy, is not sufficient without an audience receiving the message, that is, without a critical mass of followers on various social media profiles. In fact, in some studies, social media presence is precisely measured through the number of followers or the level of engagement it receives from its community (Giraldo & Martínez, 2017).

Given the importance of the number of followers for the effectiveness of the social media strategy, this study aims to analyze the trend in the growth of the number of followers in the 15 cities that make up the CEPH brand. Firstly, the starting situation of each city is analyzed through a convergence ratio that allows positioning each city. Secondly, a time series trend analysis is conducted, allowing us to see which model best captures the trend in each destination and subsequently rank the 15 destinations based on the highest growth in their audience.

The results obtained through this analysis constitute an important output for management, knowing which destinations achieve better performance on social media with the benefits that this information implies for management. To achieve these results, this study conducts a literature review to frame the research problem. Subsequently, the methodology to be used is described. Chapter 4 lists the main results obtained, and finally, it concludes with a series of conclusions and recommendations for management.

2 The Presence on Social Networks of Tourist Destinations: A Strategic Objective

The importance of social media and new technologies has been growing exponentially in recent years in all areas of life, but perhaps even more so in the tourism sector. In this regard, Pan et al. (2021) conducted research in which a series of respondents had to choose a tourist destination from a list of options. They made this choice before being informed about the destination's image on social media and repeated the choice after obtaining that information. The results showed that the image of the tourist destination influenced the choice behavior of the respondents, demonstrating the importance of the social media positioning of destinations for attracting new tourists.

In the same vein, Lee et al. (2019) emphasize the importance of providing quality tourist information through social media. They see this factor as crucial for tourism growth in less-known regions of South Korea, as these regions currently do not provide adequate tourist information to meet the needs of international tourists. In a similar study, Molinillo et al. (2019) analyze ten Spanish smart cities, finding acceptable engagement results through social media such as Facebook, Twitter, and Instagram. However, they showed weaknesses related to promoting their image and brand, as well as the effectiveness with which they communicated their smart features.

On the other hand, Chopra et al. (2022) mention that the rise of social media has revolutionized the tourism industry. Currently, many people post comments about different destinations on the internet, and many others gather information about tourist destinations through these comments. In their analysis, they indicate that the quality and quantity of information offered by tourist destinations are the two most important factors for obtaining good reviews online.

Stojanovic et al. (2018) investigate the effects of the intensity of social media use on the destination brand value, analyzing how communication on social media affects the tourist's perception of the destination. They observe a positive relationship between brand awareness, destination quality, and customer value. They emphasize that tourists who obtain information about the tourist destination on social media have a better perception of it, influencing the tourist's behavior by recommending it through social media as well as outside of it.

3 The Importance of Audience Growth (New Followers) on Social Networks for Tourist Destinations

In their article, Kiralova and Pavlicek (2015) discuss the importance of social media in both tourism demand and supply. They mention that social media allows destinations to interact directly with visitors through various online platforms. They emphasize the positive relationship between destination visitors on social media (followers) and tourists who directly experience the destination. They indicate that a higher number of followers on social media increases the chances of the destination boosting its tourism demand.

On the other hand, Song et al. (2021) point out that in the context of tourism marketing, it is essential to establish effective strategies for engaging with social media followers and strengthening relationships. They highlight the role of social media image and how it influences tourists’ destination choices. In this regard, the increase in the number of followers is crucial for better destination promotion.

Regarding the analysis of social media data, Linnes et al. (2021) stress the importance of reviewing the behaviors and interests of social media followers, as it provides a wealth of information and valuable data for tourism organizations. They emphasize that a higher number of followers on social media means more data available for analysis, enabling better insights to help tourism companies market their products. In their study on the city of Fredikstad in Norway, they use Facebook data to understand the needs of local tourists. This allows tourism businesses to be prepared to meet those needs. The study is also valuable for understanding the evolution of tourists’ preferences and needs, enabling quick adaptation to them.

4 Analysis of the Audience on Social Networks in Spanish World Heritage Cities

4.1 Data

The data used in this study were obtained from the Application Programming Interface (API) of Optimizadata's barometer (www.optimizadata.com). Specifically, various daily metrics were obtained from the profiles of the 15 Spanish Cities of World Heritage (CEPH) on the social media platforms Facebook, Twitter, and Instagram throughout the year 2022. Among all the metrics obtained, the only one utilized in this study is the number of followers on these three social media platforms.

Given that the number of followers is likely to be a metric that changes less from day to day (unlike the number of posts, comments, or impressions), it has been considered that the most suitable temporal unit of analysis in this case is the week, rather than the day. For this reason, a weekly metric of the number of followers has been generated by averaging the daily information generated by the API. In this way, we have worked with a time series of 52 weeks, all corresponding to the year 2022.

4.2 Methodology

Firstly, we aimed to analyze, from a purely descriptive perspective, the starting point of the audience on social media for the 15 Spanish Cities of World Heritage (CEPH). To do this, the number of followers on the three analyzed social media platforms as of January 1, 2022, was considered, and a convergence/divergence ratio was constructed with respect to the maximum audience of the studied group. This ratio (calculated as the quotient between the number of followers of a CEPH and the maximum number of followers among all of them) generates a variable with a maximum value of 1 (corresponding to the CEPH with the maximum number of followers on the social media platform), and where the rest of the values are always less than 1.

Secondly, also with a purely descriptive intention, the distribution of followers for the profiles of each CEPH on each social media platform was calculated. This was done by simply dividing the number of followers on each social media platform by the sum of followers on the three analyzed platforms. This simple descriptive analysis will reveal which social media platform is preferred by CEPH for promotion, to showcase their activities, etc.

Finally, leveraging the weekly information on the audience of CEPH on social media, and given that the primary interest of this study is to understand how the growth of this audience has been distributed throughout the year 2022, we have resorted to statistical time series analysis, specifically, the analysis of the trend in the number of followers. Among the different trend modelings of a time series, three have been considered in this study, as they, in most cases, explain it better:

Linear trend:

$${Y}_{t}={\beta }_{0}+{\beta }_{1} t+{e}_{t}$$
(1)

Logarithmic trend:

$${Y}_{t}={\beta }_{0}+{\beta }_{1} {\text{ln}}t+{e}_{t}$$
(2)

Exponential trend:

$${Y}_{t}={\alpha }_{0} {exp}^{{\alpha }_{1} t} {u}_{t}$$
(3)

Model (3) can also be expressed, alternatively, in the following way:

$${\text{ln}}{Y}_{y}={\beta }_{0}+{\beta }_{1}t+{e}_{t}$$
(3b)

where \({\beta }_{0}={\text{ln}}{\alpha }_{0}\); \({\beta }_{1}={\alpha }_{1}\) and \({e}_{t}={\text{ln}}{u}_{t}\)

The first of these models (linear) assumes that the growth of the audience on social media occurs more or less steadily throughout the entire year. The second trend model (logarithmic) assumes that audience growth is concentrated mainly in the early months of the year and that afterward, this growth is much smoother (or even non-existent). Finally, the third model (exponential) presupposes that during the early months of the year, the growth in the number of followers is very modest, and as the year progresses, this growth becomes increasingly pronounced, to the point that the majority of the increase in the audience occurs in the later months of the year.

To determine which of these trend models best fits the evolution of the audience on social media for the CEPH, these models have been estimated for each of the three social media platforms using the statistical package “stats” (version 3.6.2) of the R software. The model with the highest coefficient of determination (R2) was chosen. Subsequently, it was verified whether the corresponding parameter β_1 is statistically significant or not (at 5%). Finally, the value of this parameter was recorded as an estimate of the average weekly increase in the number of followers on the analyzed social media platform. This was done to rank the 15 CEPH studied based on the greater or lesser growth of their audience throughout the year 2022.

4.3 Results

Table 1 presents the convergence ratio to the maximum audience on the three considered social media platforms for the 15 Spanish Cities of World Heritage (CEPH). In the case of Facebook, the city with the highest number of followers is San Cristóbal de la Laguna, and only four cities have more than 75% of the followers of this city. These cities are Salamanca, the city that comes closest to the number of followers of San Cristóbal de la Laguna (0.915), Toledo (0.780), Úbeda (0.780), and Córdoba.

Table 1 Ratio of convergence to the maximum audience on CEPH social networks as of January 1, 2022
Full size table

On the other hand, Tarragona has the highest number of followers on Instagram, and it seems to be the only CEPH clearly investing in this social media platform. With the sole exception of Santiago de Compostela (0.637), the other cities have a convergence ratio below 0.4. In the case of the social media platform Twitter, Córdoba leads in the number of followers, with Santiago de Compostela (with 10% fewer followers than Córdoba; ratio equal to 0.901) and Tarragona (with 23.6% fewer followers than Córdoba; ratio equal to 0.764) following in the ranking behind the Andalusian capital.

At the opposite end of the cities leading in audiences on the three analyzed social media platforms are the cities with the fewest followers. Therefore, they should express some concern about their limited social media audience compared to the rest of the CEPH. Among these cities, whose average convergence ratio on the three social media platforms does not exceed 15%, are Mérida (average ratio of 0.145), Cuenca (0.103), Baeza (0.090), and Cáceres (0.042).

Table 2 presents the distribution of the total audience for each Spanish City of World Heritage (CEPH) across social media platforms. In this table, the focus is on each social media platform, and it is possible that the same person follows a CEPH on more than one social media platform. Therefore, the percentages in Table 2 do not account for possible duplications or triplications of the same user following a CEPH on multiple social media platforms. It is evident that Facebook is the preferred social media platform for these 15 CEPH, as the majority of their followers are located on this platform. This is the case for Ávila (53.3% of its followers are on Facebook), Baeza (64.8%), Cáceres (65.3%), Córdoba (54.2%), Toledo (55.9%), and, especially, Mérida (76.7%), Úbeda (77.8%), San Cristóbal de la Laguna (80.1%), and Salamanca (85.2%). However, there are some exceptions to this majority preference for Facebook. Instagram leads in online audience for Ibiza (57.9%) and Tarragona (49.1%), while Twitter is the social media platform with the highest percentage of followers for Alcalá de Henares (41.7%) and Cuenca (38.3%).

Table 2 Distribution of the number of followers on Facebook, Instagram and Twitter of the CEPH as of January 1, 2022
Full size table

Moving on to the analysis of the observed trends in the evolution of social media audiences for the Spanish Cities of World Heritage (CEPH), and after estimating the three aforementioned models (1), (2), and (3b) for each city and the three considered social media platforms, it is observed that, in most cases, the model that best fits the analyzed data is the exponential trend model (3b), whose estimations are presented in Table 3. The only exceptions to this general behavior occur with the linear trend model (1), which provides the best fit for 7 CEPH on Instagram and for the cities of Ávila and Baeza on Twitter (Table 4).

Table 3 Estimation of the exponential trend of the audience (number of followers) on Facebook, Instagram and Twitter of the CEPH during 2022
Full size table
Table 4 Estimation of the linear trend of the audience (number of followers) on Instagram and Twitter of the CEPH during 2022
Full size table

Focusing on the exponential trend model (Table 3), the adjustments made result in high values of the coefficient of determination, and thus, with β1 parameters statistically significant at 1%, with only exceptions for Alcalá de Henares on Facebook and Baeza on Instagram. In model (3), the parameter β1 represents the estimation of the expected weekly growth of the natural logarithm of the number of followers of the CEPH on the social media platform. Analyzing the estimates from Table 3, it can be observed that the highest growths in the Facebook audience are recorded in Cáceres and Cuenca, followed at a great distance by Ávila and Mérida. In the case of Instagram, audiences have grown mainly in Cuenca, although Cáceres, Córdoba, and Mérida have also recorded truly significant weekly increases in their audiences. On the other hand, on the Twitter social media platform, the most significant estimated weekly growths have been recorded in Cáceres and Salamanca. As can be easily observed, the highest audience growths have been recorded in the CEPH with fewer followers on these three social media platforms (especially notable in the cases of Cáceres and Mérida). Furthermore, it is also clear that the highest audience growths are recorded on Instagram, as the estimated β1 parameter for this social media platform is higher than the one estimated for Facebook and Twitter. This can be clearly seen in the 7 CEPH where the best model is the exponential trend for all three studied social media platforms.

Finally, considering the estimates of the linear trend model presented in Table 4 (referring only to Instagram and Twitter), it is observed that the β_1 parameter (representing, in this case, the estimation of the expected weekly growth of the number of followers) reaches the highest estimated values on the Instagram social media platform in the city of Tarragona, and to a lesser extent, in the cities of Ávila, Salamanca, and Ibiza. In the case of the Twitter social media platform, for which the linear trend model is the best fit for the evolution of followers in only two cities (Ávila and Baeza), it is observed that the estimate of weekly growth is very similar in these cities (around 10 more followers per week on average).

5 Discussion and Conclusions

Tourist destinations are obligated to provide sufficient information for travelers to make informed decisions, utilizing all available means, including their presence on social media. This level of presence can be measured by the number of followers a profile has. Therefore, monitoring and comparing the evolution of the number of followers becomes essential for proper management of this important communication tool.

This study aims to analyze the trend of the number of followers in a designation that acts as a brand for tourist destinations, the distinction as a Spanish City of World Heritage (CEPH).

The results have identified which CEPHs lead in terms of their number of followers on each of the analyzed social media platforms. This information allows the identification of cities whose social media marketing strategies are yielding better results and provides a relative perspective on the distance between them. This information can be used to guide benchmarking strategies, especially useful for destinations identified as distant from leading positions on each platform.

Regarding the ongoing competition for leadership in different social media platforms, there is a clear preference for the most established network, Facebook. While San Cristóbal de la Laguna is the best-positioned, many other cities show convergence rates close to each other. In contrast, despite the favorable growth trend of Instagram, which has become a popular photographic platform for promoting cultural tourism, CEPHs show some stagnation. Only Tarragona and Santiago de Compostela seem to be successful on this platform, with convergence ratios for the rest of the cities below 50%.

Finally, the analysis of trend in time series reveals that, for the most part, the exponential model is the best fit for the data. This implies that the highest growth rates occur in the middle and late months of the year. This result makes sense as most trips and events concentrate during this time.

It is noteworthy that the linear model is the best fit in some cities, both on Instagram and Twitter. This could reflect that strategies on these platforms are more stagnant, failing to increase the number of followers despite increased activity on the accounts. Destinations should reflect on the effectiveness of strategies developed on these platforms.

As a future line of research, it would be interesting to expand both the geographical scope of the study to see if the pattern is similar in cultural destinations internationally and to continue introducing metrics from new social media platforms, such as TikTok.

Similarly, one of the main weaknesses of this study is represented by the use of a single variable to explain a concept as multivariate as the evolution of the number of followers. Therefore, adding new metrics from the analyzed social media platforms would be useful to identify factors that may help explain the identified trend.