Public branding strategies can help create a strong identity for a public institution, fostering trust among potentially increasing conscious tourism. This research focuses on understanding the corporate branding strategies, i.e., uniqueness or isomorphism, enacted by Italian public museum by analyzing the content of official Facebook posts. Implementing a topic modelling analysis of the content of posts published over six months shows that generally museums pursue isomorphic branding strategies especially in connection to service-related themes, but uniqueness of branding strategies emerges especially in connection to experiences and knowledge themes for some museums. These results enrich public branding literature as well as provide relevant results for public managers and tourism managers.
Social media public branding is highly relevant for sustainable tourism. An effective branding of
public sector institutions helps create a strong identity for a destination, fostering trust among
potential tourists [
Even if strategies of social media branding in the private sector have been largely studied [
© 2025 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
In line of this need, I focus specifically on the peculiar empirical setting of the 50 top-visited Italian public museums in 2017 and study their branding strategies on Facebook along a six-month period, between September 2018 and February 2019. I answer the following research question “Do top-visited Italian public museums pursue uniqueness or isomorphism as Facebook branding strategy?” by analysing the textual content of official posts generated on Facebook by museum accounts to identify whether there exist topics of discussion unique for specific museums or rather if the topics are shared across museums indicating isomorphism in the branding content. The specific empirical setting of top-visited Italian public museums is selected to support the research for four major reasons.
First, museums are extremely appealing for the sake of understanding whether unique aspects are leveraged in branding strategy as these institutions have the peculiar characteristic of targeting multiple objectives being oriented to sustainability, community and public value development, as clearly shown by the definition of museum provided by the International Council of Museums [19].
Second, public museums usually compete for visitors because the autonomous part of their economic sustainability largely depends on ticketing (see also [20]) but focusing on top-visited public museums alleviates this constraint of high-competition, as these public institutions operate in an environment of over-demand and hence in a non-highly competitive setting. Moreover, selecting top-visited museums guarantees an already consolidated brand, as shown by the fact that municipality may flagship monuments for touristic attraction [21], and guarantees homogeneity in the management problems they must face, like queue management.
Third, I focus on the 50 top-visited in Italy in 2017 not only because Italy is recognized as one of the countries with the highest density of cultural heritage and institutions in the world [22], but also because in the same period there has been institutional attention to digital aspects in cultural service delivery, among which social media adoption (see also [23]). Additionally, this provides homogeneity in governance as all the analysed institutions depend on the Italian ministry in charge.
Fourth, among all the online platforms, I focus specifically on social media as these platforms are the most used by public sector institutions and by citizens and users and have a great influence on external branding, making the study appealing to inform about public branding. Though considering multiple platforms could be interesting to compare branding strategies of museums across channels, the choice of selecting the specific channel of Facebook is motivated by the fact that in the period analysed it was the most used social media platform among Italian public museums and that various accounts on this platform were already strong and consolidated (e.g., [24]). As there is a strong dependency of the branding strategy on the social media platform, the choice of this single platform provides homogeneity in the social media audience and target of the official accounts of museums.
The main data source of this study is represented by Facebook posts published by a subset of the 50 Italian public museums representing the top-visited in 2017.
In this study, for the sake of simplicity, I use the term “museum” to refer to those institutes defined by the Italian national statistical institute as “museums and similar institutions”, i.e., museums, galleries and archaeological parks, therefore excluding institutions that are only monuments [25]. Among public museums, I considered just institutions directly managed by the Italian ministry and excluded those managed by consortia or foundations. I then considered the 50 museum institutions mostly visited according to the official statistics on annual visitors of 2017 [25]. As one museum institution can be composed of multiple individual sites and museums that are managed by the same authority, the complete sample of individual museums most visited in 2017 considered for the study is composed of 71 individual museums listed in Figure 1.
For each of these museums, the official Facebook accounts have been manually identified, and the credibility of the web sources has been directly verified with museum managers. Though at the time Facebook was the most diffused platforms among Italian museums, not all museums indeed managed an official Facebook account and only 39 unique official Facebook existed.
Between September 2018 and February 2019, 35 out of the 39 official Facebook accounts of museums actually published at least one post containing text or image generated by the account and were therefore included in the analysis. Through an ad-hoc implemented automated data collection system, 4888 unique posts have been collected.
The museums analysed have different characteristics in terms of onsite visitors, online followers and number of Facebook posts. Specifically, the average amount of published posts per account is highly variable, ranging from a minimum of 7 posts per account to a maximum of 600, with an average of 139 posts and a median of 102 posts.
Among the 4888 posts collected, 125 posts contain only images and, as the scope of the research is the analysis of the uniqueness and isomorphism in the textual content, these few posts (2.56% of total) are discarded from the analysis. The text of the official posts published by the museums of interest is analysed through a topic modelling technique to detect the semantic themes of discussion that occur. Among the wide set of methodologies for automatic text modelling, I select Latent Dirichlet Allocation (LDA) [26] as it enables the identification of latent dimensions through an unsupervised Bayesian model without making any preliminary assumption on the distribution of posts, other than assuming the existence of hidden semantic concepts within the posts. For the application of LDA, I construct a document-term matrix of 4758 posts and 18380 terms, obtained preprocessing text of posts with R software by means of lowercase conversion, elimination of specific characters (e.g., emojis, URLs, punctuation, and numbers), exclusion of language-specific and context-specific stopwords, and applying language-specific Porter’s stemming algorithms, relying on packages tm, SnowballC, and ldatuning. Then, I fit the LDA model using a Gibbs sampling with the with 2000 Gibbs iterations, 2000 thinning and no burn-in.
The choice of the most suitable number of latent topics discussed has been based on the metrics proposed by [27-30], with the number of topics varying between 2 and 25 and selecting the candidate number of latent topics following the elbow criterion (Figure 2).
The topics are interpreted combining the analysis of the topic-word distribution of the 15 most probable words and of the probability of observing topics within the 30 most probable posts for each topic. The validity and robustness of the LDA results is assessed applying the non-negative matrix factorization method in R [31].
The resulting distribution of topics are analysed to further identify whether a topic is peculiar to a specific museum or, vice versa, if it is shared across different museums. Specifically, for each topic, the upper outliers of the distribution of posts over the topic are analysed. If the frequency of outliers’ posts published by specific museum account is strongly unbalanced towards a specific museum account, the topic is considered peculiar to the specific museum account, and this constitutes evidence of the uniqueness of the topic for the specific museum account.
!"#$%$F$'()*+'
From the visual inspection of Figure 2 it is possible to attest the existence of k=8 topics of discussion across museum posts, namely:
T1) Educational activities with art & artists, e.g., educational activities with art & artists T2) Expositions & heritage, e.g., expositions, history, building, facades, … T3) Calendar, e.g., openings, special events, … T4) Archaeological narratives, e.g., archaeological discoveries, research, life, … T5) Museum for the community, e.g., participatory activities, spirit healing, inclusion, … T6) Collection & Curatorship, e.g., collection, restoration, research, … T7) Cultural anecdotes, e.g., curiosities, ironies, … T8) Special events & offerings, e.g., festivities, discounts for subscriptions, …
The analysis of the outliers of the upper-right tails of the distribution of posts per each topic shows that topics T1, T4, T5 and T8 grasp contents that are peculiar to specific museums.
Specifically, Figure 3 shows that 48.11% of extreme posts in the distribution of topic T1 are published only by gallerieaccademiavenezia. Moreover, the second museum (PinacotecadiBrera) in terms of extreme posts with respect to topic T1 has a frequency of 12.75%. This shows that the discussion of educational activities with art and artists (T1) is not generally shared across different museum accounts but rather it is a highly specific content presented in the Facebook branding strategy of gallerieaccademiavenezia. Similarly, Figure 4 shows that 49.87% of the upper outliers for topic T4 are published only by MuseoArcheologicoRC, with the second account (parcocolosseo) achieving a publishing rate of 6.82% among outliers’ posts. Therefore, archaeological narratives (T4) are not generally shared across different museum accounts but rather are highly specific contents branded by MuseoArcheologicoRC. In similar fashion, Figure 5 shows that 60.26% of the outliers’ posts for topic T5 are published only by museocapodimonte, with the second account (museodisanmartino) publishing only the 11.59% of the outliers’ posts. This shows that the delivering content about the museum as a community (T5) is not generally shared across different museum accounts but rather it is highly specific for the branding strategy of museocapodimonte. The analysis of Figure 6 shows that 63.21% of the extreme posts in the distribution of topic T8 are published only by MANNapoli, with the second account (PilottaParma) publishing just the 7.77% of the outliers’ posts. This shows that discussing special events and offerings is not generally shared across different museum accounts but rather highly specific content to the branding strategy of MANNapoli.
Additionally, the analysis of the outliers of the upper-right tails of the distribution of posts per each topic shows that the other topics – namely topics T2, T3, T6 and T7 - are not peculiar to specific museums but rather are shared across various museums. For instance, the analysis of the upper outliers of the distribution of posts within topic T3 shows (Figure 7) that the 16.62% of these posts are published by scaviostia, the 12.47% by museomiramare, and the 11.91% by parcoarcheoilogicopaestum. As similar shared behaviours are observed considering topics T2, T6 and T7, these results symbolize the existence of isomorphic branding strategy among museums in terms of content of posts published when it comes to communicating about calendar (T3), collection and curatorship (T6), describing the expositions and heritage (T2) and narrating cultural anecdotes (T7).
Since four of the topics emerging from the official Facebook posts published by top-visited Italian public museums are spread across various museum accounts, the analyses shows that museums do not generally pursue a branding strategy which favours uniqueness but rather pursue a branding logic of isomorphism in terms of semantic of textual description of the post. This is evidenced by the topics four topics identified concerned with service-provision aspects, namely expositions and heritage (T2), calendar (T3), collection and curatorship (T6) and cultural anecdotes (T7).
Nonetheless, the analyses also show that there are four topics that are strongly referred to specific museums. This is the case of topics concerned with experiences and knowledge that are provided by some museums as a uniqueness branding value. This is the case of educational activities with art and artists (T1) for gallerieaccademiavenezia, of archaeological narratives (T4) for MuseoArcheologicoRC, of description of the role of the museum for the community (T5) for museocapodimonte, and narratives about special events and offerings (T8) for MANNapoli.
Though the nature of this study is exploratory, this research offers at least two major contributions. First, the empirical findings of the study enrich the debate in public sector branding literature concerned with the tensions between corporate branding and mission and values of the public sector institutions, providing empirical evidence in an empirical case of relatively low competition, thus far underexplored. Second, the findings show that some shared values exist in the branding strategies of different museum institutions and identifies which are the specific themes that are shared across various museums. These results could be leveraged also by other public sector institutions, such as local government institutions and administrations, to establish online connections with extant cultural institutions to interact with the community and to foster collaborations to target specific audiences that could be further attracted to these destinations relevant for the public administrations.
Nonetheless, the explorative nature of this research also presents some limitations that could be further explored as future developments to provide relevant robust statistical results of the measurement of the uniqueness and isomorphism of social media public brand strategies.
First, extensions should consider wider timespans, e.g., one year or more than one year, to compare the potential seasonality of results and public branding strategies connected to special campaigns. Second, extensions of the current study should study what happens to the strategies when more than one channel is considered, e.g., Instagram, X, and TikTok, to establish whether also uniqueness or isomorphisms are channel-specific or preserves across channels. Third, additional analyses could consider the physical distances between museums and with other institutions, e.g., schools, to assess whether there exists a model of public branding strategy specifically connected to some territories. Fourth, the study could be expanded to the analysis of niche museums instead of focusing only on top-museums, to consider also the stronger effect of the variable of economic sustainability and studying the uniqueness and isomorphism of public branding strategies in a highly competitive but understudied setting. Fifth, the study could be extended considering also the role of brand image, i.e., perceptions and reactions of online users to specific strategies, to consider also the reactions of users to the public branding strategies. Finally, the study could be expanded analysing also the content of images in the public branding strategy, to compare the semantic similarities across images used in the public branding strategies and identify uniqueness or the isomorphism also connected to the content of images.
Declaration on Generative AI The author has not employed any Generative AI tools. [4] Lichtenstein, M., & Lyons, R. (2010). Communities, Environments, and Transformations in
Sustainable Tourism. The Annals of Regional Science. [5] Boulding, W. & Christen, M. (2005). Branding and Sustainable Innovation. Journal of Business
Research. [6] Kaplan, A. M., & Haenlein, M. (2010). Users of the World, Unite! The Challenges and
Opportunities of Social Media. Business Horizons. [7] McCarthy, J., & Rowe, S. (2014). Branding in the Age of Social Media. Journal of Brand
Management. [8] Wright, G. H., Chew, C. and Hines, A. (2012) The Relevance and Efficacy of Marketing in Public and Non-Profit Service Management. Public Management Review, 14:4 pp433–50 [9] Leijerholt, U., G. Biedenbach, and P. Hulten. 2022. “Internal brand management in the public sector: the effects of internal communication, organizational practices, and PSM on employees’ brand perceptions” Public Management Review 24(3): 442-465, DOI: 10.1080/14719037.2020.1834607 [10] Waeraas, A., H. Bjørnå, and T. Moldenaes. 2015. “Place, Organization, Democracy: Three Strategies For Municipal Branding.” Public Management Review 17 (9): 1282–1304. doi:10.1080/14719037.2014.906965. [11] Perez, Edgar Oliver Bustos. 2021. Organizational Reputation in Public Administration: A Systematic Literature Review. Public Administration Review. https://doi.org/10.1111/puar.13363. [12] Dolbec, P.-Y., Castilhos, R. B., Fonseca, M. J., & Trez, G. (2021), "How established organizations combine logics to reconfigure resources and adapt to marketization: A case study of Brazilian religious schools" [13] Sataøen, H. L., and A. Waeraas. 2015. “Branding without Unique Brands: Managing Similarity and Difference in a Public Sector Context.” Public Management Review 17 (3): 443–461. doi:10.1080/14719037.2013.841976 [14] Waeraas, A., Bjørnå, H., & Moldenaes, T. (2014), "Place, organization, democracy: Three strategies for municipal branding" [15] Carvalho, D., Demo, G., Scussel, F., Montezano, L., & Amazonas, B. (2022), "How do public service users see public institutions? Development and validation of the public brand personality scale" [16] Liu, Q., & Wang, X. (2023), "The impact of brand trust on consumers’ behavior toward agricultural products’ regional public brand" [17] Powell, S. M. (2019), "Journal of Brand Management: Year end review 2019" [18] Sataøen, H. L., & Waeraas, A. (2013), "Branding without unique brands: Managing similarity and difference in a public sector context" [19] International Council of Museums. (2023, June 5). Museum Definition - International Council of
Museums. https://icom.museum/en/resources/standards-guidelines/museum-definition/ [20] Ore, I. S. 2. (2019, December 8). Musei: quanto sono autonomi dall’ente pubblico o statale? Il Sole 24 ORE. https://www.ilsole24ore.com/art/musei-quanto-sono-autonomo-dall-ente-pubblico-ostatale-ACnZD83 [21] Mabillard, V., Pasquier, M., & Vuignier, R. (2023). Marketing tools in attractiveness policies. In
Place Branding and Marketing from a Policy Perspective (pp. 170-194). Routledge. [22] UNESCO World Heritage Centre. (n.d.). UNESCO World Heritage Centre - World Heritage List.
https://whc.unesco.org/en/list/&order=country#alphaI [23] Franceschini: “Piano Strategico del Turismo governa la crescita del turismo internazionale.” (n.d.). Ministero per I Beni E Le Attività Culturali E per Il Turismo. https://cultura.gov.it/comunicato/franceschini-piano-strategico-del-turismo-governa-lacrescita-del-turismo-internazionale-1 [24] Gronemann, S.T., Kristiansen, E., and Drotner, K. (2015), “Mediated co-construction of museums and audiences on Facebook”, Museum Management and Curatorship, Vol. 30, No.3,pp. 174-190.