First, destination managers should consider remarketing topics in the repositioning strategy, which we could name the “3 Rs”: religion, relaxation and renaissance (culture). The idea is to enjoy a mix of spiritual, cultural and relaxation experiences.

Even more, given that nowadays it’s important to position a destination considering elements related to social identity, community and sustainability, these topics should not be obviated, as recent comments in Trip Advisor have slightly shown.

Second, although a repositioning strategy can be followed to improve the perception of the Vatican City among its visitors, we cannot forget that places need be pleasant also for their residents because if they are happy, they will contribute to making the repositioning of the destination work.

Third, to avoid the weak points that visitors have commented on (i.e. queues, crowded, a lot of people and waiting times), some ideas from smart tourism destinations could be borrowed. For example, the use of technologies to improve connectivity, information systems and sensorization to solve overcrowding.

Fourth, given that topics such as photos and post office are also mentioned in visitors’ comments, this kind of information could be included on flyers and social networks to promote the destination.

METHOD

Tourism destination and sample

Vatican City was chosen because this destination is visited by international pleasure travelers as well as authentic pilgrims from all over the world (Kim and Kimm 2018). In addition, this destination has been one of the most damaged by the pandemic (Mosier et. al, 2020). In addition, the justification for studying a specific city is because following Bock (2015), city tourism is one of the fastest-rising travel segments worldwide, and the changing nature of city tourism becomes increasingly apparent in many cities that need to be investigated. In addition, Vatican City was chosen because massively visited tourist destinations have suffered a superior image deterioration due to the severe restrictions that Covid has brought in terms of capacity. That’s because holy monuments and other religious spaces have a special place in the hearts and minds of many people (Mosier et al., 2020). Vatican City has been gaining interest in recent years, because is one of the three major pilgrimage destinations for Christians, along with Jerusalem and Santiago de Compostela (Kim and Kim, 2018). In terms of Timothy (2020), Vatican City is the smallest country on earth but probably the most visited. All these points lead us to focus our proposal on this tourism destination. According to cumulative attraction and gravity theory, “travelers tend to visit a number of attractions in and around a destination in order to maximize value (money, time, and effort) and reduce risk (Park et al., 2020), and Vatican City is an excellent example of this.

       To study this destination, big data was used, due to the growing importance of user-generated content in tourism. Trip Advisor was used because it is a website where most information posted is autonomously generated by its users (Miguéns et al., 2008). Among review sites, Trip Advisor is the world’s largest travel site (Murakami, 2018). This source of information is also used in the study of the city of Venice done by(Ganzaroli et al., 2021) or in the study of India done by Gour et al. (2022). More specifically, 7152 online travel reviews (OTRs) written in English on Trip Advisor about “things to do” in Vatican City were retrieved in July 2021. That is, given that the English site of the Vatican City was analyzed, English reviews were analyzed. The data was separated into two main blocks: globally and since the beginning of Covid-19. 10% of these comments were after the declaration of pandemic status in March 2020. The results for this sub-sample were compared with the whole period to identify new trends in travelers’ comments. Recent research by Guerrero-Rodriguez, et al., (2021) compared differences in OTRs by examining different information attributes in the dataset. More specifically, these authors developed three analyses: (1) dividing OTRs based on their geographical origin (national vs. international); (2) identifying temporal changes in OTRs (holiday vs. non-holiday seasons), and; (3) measuring the evolution in the number of OTRs over time (periods between 2002 and 2020). In their study, “Mexico was the country that contributed to a larger number of OTRs, 52.5% of the total. The USA and Canada were second and third, reporting 22.7% and 4.66% respectively. However, other countries also have a considerable number of reviews on the TripAdvisor platform, such as Japan (3.32%), Colombia (2.16%), France (2.05%), Brazil (1.96%), Costa Rica (1.71%), Spain (1.69%), Australia (1.61%), England (1.58%) and Germany (1.47%)” (Guerrero-Rodriguez et al., 2021, p. 9). In the same vein Marine-Roig and Huertas (2020) analyzed 152,704 Airbnb OTRs in Barcelona in several languages, each sub-sample of a different size (English: 108,064; Spanish: 22,132; French: 13,805; German: 4611; and Italian: 4092), from different periods (being 2017 an inflection point due to a terrorist attack and the Catalan independence process.

Statistical tools

To carry on our research, the same steps of Marine-Roig and Clavé (2015) were followed. As these authors recommend, the a-processing paradigm was developed (data are first stored and then analyzed). To this end, three free software were used: Cyotek Web Copy, Note Pad ++, and KH Koder 3 Folder.  First, Cyotek is a software that permits web data extraction. It was used to download the comments in the English web of the Vatican City in Trip Advisor. This information unstructured or only semi-structured information was organized with this software to be easily analyzed more systematically. Second, Note Pad ++ was used to open the extracted data in an easily understandable format. Only texts were analyzed in our study. Mariné-Roig (2022) recommends using this kind of software to give a standard format (ISO: International Organization for Standardization) to date. Third, KH Koder 3 Folder was used to analyze the data, because this software is recommended to relate and interpret people’s comments concerning a specific theme. This free software for analyzing textual data was used, for example, to analyze the user-generated content about Japanese gardens (Yingming and Katsunori, 2019). It permits unveiling how frequently occurring words are interrelated. Also, Murakami (2018) uses this statistical tool to study the image of Japan in order to identify which tourist attractions should be recommended. In the same vein, Tussyadiah (2017) used to KH-Koder to identify content and themes from online reviews to explain major service attributes of peer-to-peer (P2P) accommodation sought by guests That is, text mining has been used to study destinations' image by identifying frequent words and the network of the words (Murakami, 2018; Mariné-Roig, 2022). We have focused on our analysis in the textual body Following Tussyadiah’s (2017) procedure: (i) some words were specified as stop words, such as articles or connectors (i.e. “a”, “an”, “and”, “already”, “the”, “because”, “during”, “each” etc.); (ii) a long list of words were included in the “force pick up” box (for example, cannot miss, can’t beat, can’t miss, can’t recommend, cannot be missed, do not, etc.) and (iii) part-of-speech (POS) tagging were selected (such as noun, proper-noun, verb, adjective etc.). The box other was not marked.