Introduction

We are living in the midst of a gold rush of artificial intelligence (AI) development, of which ChatGPT is one indicator. ChatGPT is an AI-based language model and a type of technological science that can be used to mimic human intelligence and generate text based on a large corpus of data, the aim being to enhance productivity and creativity through automation (Sterne 2017). It has been claimed, and it may be tempting to imagine, that AI can multiply work efficiency and creativity, while simultaneously saving time and reducing errors, stress, and cost. But to what extent can ChatGPT really substitute for human thought?

This study addresses that question by investigating whether ChatGPT could provide the basis for a constructive slogan invention process for a city. The issue is empirically explored in two phases: the first utilizes ChatGPT in the creation of slogans for the 20 largest municipalities in Finland. The second introduces an expert human perspective on the matter by interviewing representatives of three major Finnish cities to garner their thoughts on the use of AI in place branding and on AI-generated slogans.

Adopting ChatGPT to support their branding efforts could save municipalities time and money. Place branding is, however, more complex than mere promotional elements, requiring branding efforts to extend beyond logos and slogans, both of which can be expressions of a place brand but not the brand itself (Zenker 2021; Govers 2013). Nevertheless, slogans are being widely used by municipalities, as they have traditionally been by companies. The question is how much energy and effort places should commit to crafting slogans (Zenker 2021).

Why use slogans? While some people are suspicious of them, slogans can play an important role in differentiating and creating awareness of a place and its image, and a great deal of emphasis, time, and budget are allocated to their development (see, e.g. Wilson 2021; Govers 2013). Several authors view them as often being the most visible and memorable of municipalities’ branding elements (e.g. Zenker 2021; Wilson 2020; Medway and Warnaby 2014; Govers 2013; Hildreth 2013; Kohli et al. 2007). At their best, slogans can concisely capture a municipality’s unique selling proposition, which makes their linguistic form interesting and worthy of examination. Linguistic form is also a topic of interest from an artificial intelligence (AI) point of view.

The number of marketing studies related to artificial intelligence has peaked in recent years. However, that growth in interest has rarely touched upon its creative capabilities. A rare exception is a study by Pankka (2023) investigating the effectiveness of AI-created corporate slogans based on their performance in comparison with human-created equivalents. Research combining linguistics with place branding and investigating the capabilities of AI in creating municipality slogans is thus far non-existent. This study aims to begin to address that issue.

Finland has 309 municipalities, most of which utilize one or more slogans, and therefore, the process of crafting slogans concerns practically every city or town. A recent study conducted among Finnish municipalities (Hakala et al. 2024) found that the current human-oriented slogans do not accurately differentiate and position the municipalities in relation to each other. The strategy work undertaken by municipalities seems to have failed to identify their core values that can be concisely reflected in a slogan.

Literature review

The wider foundation of this study lies in the notion of place branding. Over the last 20 years, place branding has achieved an established position in branding research and also in practical place management and development (Kavaratzis and Florek 2021; Eshuis and Edwards 2013). Compared to companies and product brands, however, the ability of places to use large-scale promotional elements and to brand themselves is limited due to financial constraints on marketing budgets, the complexity of the place as a product, the complexity of the organizational mechanisms steering the brand, and diverse target groups and stakeholders. (Cassinger and Eksell 2017; Zenker and Braun 2017; Warnaby 2009; Roper and Davies 2007). Places cannot be owned in a conventional sense and nor do they have a single identity that can be clearly branded; they can evoke different associations in the minds of diverse target groups and stakeholders (Zenker 2021). As Hubbard and Hall (1998, p. 7) suggest, places are “constituted through a plethora of images and representations” that can be mediated by visual symbols and by verbal expressions, such as slogans.

Such promotional materials—including slogans—are criticized for creating calculated, idealized images that do not reflect reality; nevertheless, Wilson (2020) shows that they are still widely used. Anholt (2008) emphasizes that places cannot build or change their reputation via a catchphrase; they can only be influenced to some degree. According to Zenker (2021), slogans can be useful elements in a place’s secondary communication. At best, slogans can leave a touchpoint in people’s memory, increase awareness of the place, and help it stand out from the crowd—but they cannot comprise the entirety of the brand. (ibid.)

Slogans have been widely examined in the context of tourism (e.g. Richardson and Cohen 1993; Pike 2004; Lee et al. 2006). These previous studies report that tourism slogans could help in developing a destination brand (e.g. Galí et al. 2017) to enhance the destination’s image (Kohli et al. 2007) and to differentiate it from other destinations (Pike 2004). Little attention has been paid to the rhetorical nature of slogans and the persuasive power of language (Huadhom and Trakulkasemsuk 2017) despite slogans being verbal expressions and strongly based on meanings or semantics. Accordingly, it is important to scrutinize municipal slogans more closely from a semantic perspective.

This study reports on using AI to create slogans for Finnish municipalities. First, in order to grasp what AI means, we need to understand what is meant by intelligence (Rahman 2020). A commonly used definition describes intelligence as the ability to obtain and apply knowledge (Gupta and Mangla 2020). There is no exhaustive list covering all its features; however, the following characteristics repeatedly occur in the literature (see Rahman 2020): reasoning (ability to apply logic), perception (making sense of the sensory information), natural language communication (using language that has developed naturally over time), motion and manipulation (one’s ability to move and control movement), learning (the competence to acquire knowledge), representing knowledge (the ability to place meaning on to environments, activities, or items), planning (a set of actions to meet an initial aim), social awareness (comprehending the behaviour of others and adapting to it), as well as general intelligence (combining all the intelligence characteristics to solve issues).

Rahman (2020) asserts that the above list assists in understanding the concept of AI because machines pursue the same features in order to become intelligent. Even though AI has attracted increasing interest due to its varied targets of application during the last few years (Huawei Technologies Co 2023), it is not a concept that has been recently discovered (Akerkar 2018). In fact, the first advancements of the technology can be traced back to well over 60 years ago. The concept has been defined in many plausible ways, the earliest probably being in 1956: “Artificial intelligence is about letting a machine simulate the intelligent behaviour of humans as precisely as it can” (Huawei Technologies Co 2023, p. 1). Notwithstanding the definition’s age, its main message regarding human simulation can still be used to describe one of the recent AI establishments, namely ChatGPT (Pankka 2023).

ChatGPT, the research company OpenAI’s chatbot, and its version GPT-3 is an autoregressive language model aimed to instruct a large neural network to perform a variety of text-generation tasks. It can learn and understand language and produce human-like responses. While AI aims to mimic human behaviour, it also has distinct qualities that people do not have. The following Table lists the differences between humans and machines (Table 1).

Table 1 Humans versus machines (modified from Gupta and Mangla 2020)
Full size table

The above differences between humans and machines give rise to capabilities that both lack or excel at. Firstly, a few authors have found that AI content can be vague and general (e.g. Mitrovic et al. 2023; Dwivedi et al. 2023; Ma et al. 2023), whereas human texts tend to include more divergent and specific thoughts. Secondly, humans have been described as inserting hidden meanings into text, whereas AI tends to produce more literal creations (Guo et al. 2023). Thirdly, AI’s capability to produce meaningful content has been questioned, whereas humans are praised for producing text with deeper meanings (Sterne 2017). Fourthly, some state that AI utilizes more unusual words than humans (e.g. Mitrovic et al. 2023), and others argue that AI content is more objective in comparison with that of humans (Guo et al. 2023). Lastly, there are claims that AI-generated content lacks creativity and novelty, whereas humans are more capable of incorporating such traits into their texts (Dwivedi et al. 2023).

This exploratory study reports attempts to utilize ChatGPT to create slogans for Finland’s 20 largest municipalities and compares the outcomes with the existing, human-generated slogans.

Data and methodology

The purpose of this study was to investigate whether ChatGPT could provide a premise for a constructive slogan invention process for place branding. The basic assumption was that ChatGPT could ease the work of those responsible for place branding and, at the same time, save time and reduce costs. The study employs a two-stage empirical method to address its research aim: the first involved utilizing ChatGPT to create slogans for the 20 largest municipalities in Finland. The second phase introduces an expert perspective on the matter involving interviewing representatives of three major Finnish cities to harvest their thoughts on the use of AI in place branding and on AI-generated slogans.

The slogan data are composed of three sets. The first is the current, human-generated slogans of the 20 largest Finnish towns and cities, which were extracted from the municipalities’ websites. Second and third, the prompts “Create a slogan for the city of […]” and: “Create a slogan according to the values of the city of […]” were input to chat.openai.com. All data sets were analysed by qualitative content analysis screening the vocabulary of the slogans and their lexical items, that is, their nouns, verbs, and adjectives. Given the focal role of verbs, nouns, and adjectives in the meaning of a text, these words were isolated in the slogans, and their incidence was noted. Then, the existing human-generated slogans were compared with the ones created by AI.

The human expert perspective was obtained by interviewing the marketing director of Kuopio, Finland, and the mayors of the Cities of Vantaa and Pori, Finland. The selection of the respondents was based on non-probability sampling by selecting interviewees who had long-term expertise in and commitment to place branding and who were available (see Ghauri et al. 2020). The interviews were semi-structured, meaning that the topics and issues to be covered—the use of AI in place branding and AI-generated slogans—were determined beforehand (Berger 2016). Due to time constraints and geographical distance, the interviews were conducted via Zoom and recorded.

After the interviews, the recordings were transcribed in full. All interviews lasted between 50 and 55 min. To ensure time concurrence with the other interviews (Timonen 2024; Inna 2024), the first interview (Soininen 2023) was supplemented by email correspondence in March 2024. The transcriptions were analysed inductively in line with qualitative textual analysis procedures to identify the interviews’ key issues, themes, and concepts. Member checking, that is, clarifying tentative findings with the participants, was used to determine credibility (Lincoln and Guba 1985; Koelsch 2013).

The research ethics of the interview phase were safeguarded by informing the interviewees about the purpose of the research and the data collection. Moreover, data protection provisions were carefully adhered to. The interviewees gave permission for their names, positions, and organization details to be published (Daymon and Holloway 2011). The research material is carefully preserved and will not be accessible to anyone other than the researcher.

Findings on the slogan data

Human-generated slogans

Among the currently used human-generated slogans (see Table 2), the most common denominator was capital: For example, Oulu is referred to as the capital of northern Scandinavia, Lappeenranta as the climate capital of Finland, Seinäjoki as the space capital, and Rovaniemi as the arctic capital. The wording and sentence structure were quite versatile, and expressions often employed metaphor (e.g. Kiss my Turku; Pori gets under your skin; One Hel of an impact (Helsinki)). How well these slogans convey special features, create an image, or differentiate the city from others remains unclear.

Table 2 Human-generated municipality slogans
Full size table

Slogans created by AI

In category AI 1 (“Create a slogan for the city of […]”), the most typical feature was the structure: all slogans started with the adverb ‘where’ (see Table 3). Another characteristic element was the use of verbs; every slogan contained a verb, the most common ones being meet (5/20), ignite (3/20), and thrive (3 /20). The most typical nouns were innovation (6/20), nature (6/20), history (3/20), and heritage (2/20). The word ‘future’ was mentioned in only one slogan, whereas in the second AI category (“Create a slogan according to the values of the city of […]”), it was the most common noun.

Table 3 AI-generated slogans (AI 1)
Full size table

At best, a municipality’s core competence is rooted in its long-term strategy and realized in the slogan. Without the connection to the deeper, unique values and strategy, and also the culture and history of the place, slogans remain superficial (Wilson 2021). Values are considered a central part of the value foundation of corporate brands (Urde 2003). As place brands have been equated with corporate brands (e.g. Kavaratzis 2009) we can say that place branding should also be based on the values of the place. The core values define the brand and affect an organization at every level—its leadership, strategy, organisation, and communications—at all times. Urde (2003) found the development of core values is synonymous with the development of the brand. That is why it is important for places to understand their core values and build the place brand upon them. Accordingly, the second question to chat.openai.com was: “Create a slogan according to the values of the city of …. The findings are presented in Table 4.

Table 4 AI-generated slogans (AI 2) based on the values of the municipalities
Full size table

In this category (see Table 4), 17 of 20 slogans included a verb, and they were all in the active voice. The most common verbs were embrace (9/20) and nurture (4/20). The communal aspect was emphasized in this category, the word together being quite common (5/20). Future was expressed in various forms: endless possibilities; empowering tomorrow; inspiring progress; inspiring growth; together for a brighter tomorrow; energizing futures; sustainable futures; inspiring dreams; pioneering progress; and crafting tomorrow.

Findings of the expert interviews

The interviews commenced with three questions to lead the conversation to the wider context of the study, place branding.

  1. 1.

    How do you see/define the overall purpose and importance of place branding in your city?

  2. 2.

    How do you conduct place branding in your city? and

  3. 3.

    What role do slogans play in your branding efforts?

The answers can be found in Table 5.

Table 5 Questions and answers to the interview questions
Full size table

All interviewees emphasized the role of place branding in their cities. They were aware that cities are competing to attract residents, investors, travellers, and other stakeholders, just like companies and product brands. Finland is a large country in terms of area (338,462 km2) but is one of the least populated countries in Europe, with about 18 people per square kilometre. The population of 5.6 million is very unevenly distributed, with the majority living near the coast, in the southern and south-western parts of the country. More than a million people live in the capital, Helsinki or its environs.

In contrast, the population of the northernmost regions is, on average, only two people per square kilometre. There are currently 309 municipalities which all strive for their independence and to safeguard their existence by trying to differentiate themselves from each other and trying to attract people and businesses. The low national birth rate of 1.32 is an issue and older people are steadily coming to dominate the population (Statista 2024).

In the cities of the interviewees, place branding is the responsibility of the communications department. The Mayor of Pori, in particular, emphasized the importance of grounding place branding operations in the city strategy: “Behind the success, there is real, serious brand-strategy work” (Inna 2024). Of the three interviewees, he was the most positive about using slogans in place branding. “Slogans are very important…catching the way people think and talk” (Inna 2024). The Mayor of Vantaa raised the importance of the internal and external functionality of a successful slogan:

If you find a good slogan, it is very powerful. It communicates your message but also has an impact on what you do. It crystallizes the soul but also connects and inspires your own staff. If it doesn’t work internally, it doesn’t work externally (Timonen 2024).

Subsequently, the discussion moved to the role of AI in creating slogans. The first question was: “How do you see the role and possibilities of AI in place branding? Do you think it could replace human creativity in the future?” All three interviewees saw potential in using AI but not as something that would replace human input. “I see more opportunity than threat. AI may create a lot of different versions, but it’s the people who have to make the final decisions” (Inna 2024). Similarly, the Mayor of Vantaa commented:

Artificial intelligence will affect everything, communication and marketing, including slogans. AI enables an almost endless pool of options as it mines it from the vast dataset. But the limit is that it will not recognize that certain human tone—history, modern times, many such things that are very difficult for AI to reach. But it produces a huge number of options and perspectives that a human workshop cannot draw from such material—to go through huge materials and find ways to implement them. We need that human part, which can only be perceived with human senses. AI is one tool among others—it’s not worth taking a more dramatic approach to it. (Timonen 2024)

The Marketing Director of Kuopio shared similar views:

The use of artificial intelligence challenges the background research of brand work in its own way, as it is able to analyse the content of the area produced from numerous sources. So, in practice, you could even reach a wider group of respondents than in a normal survey. But ultimately, choices must be made by people, in the strategic management of the cities. (Soininen 2023; 2024)

The same informant also expressed concern over a potential loss of resident involvement and collaboration: “This can be a stumbling block for community commitment and participation if artificial intelligence is used instead of having residents participate”.

Finally, all interviewees were prompted to comment on the AI-generated slogans produced for their own city (see Table 6).

Table 6 AI-generated slogans for Kuopio, Vantaa and Pori
Full size table

The Mayor of Pori was impressed with the results for both the create a slogan for the city and the create a slogan according to the values of the city results, but the other two interviewees preferred the value-based versions. Marketing Director Soininen (2023) commented: “The value-based, AI-generated slogan suits Kuopio very well as it emphasizes the communal values of the city”.

The Mayor of Vantaa also found embedded meaning in the AI2 version: Embracing Diversity, Nurturing Community!

I prefer this one; the verbs (embracing, nurturing) make it active. It has some action in it. A slogan has to be value-based. In one way, we can say, maybe exaggerating a little but not much, that a slogan is the shortest way of communicating your strategy. It has to work both internally and externally. Vantaa is the most diverse city in Finland, perhaps even in the whole of Europe: over 120 languages are spoken. Diversity really differentiates Vantaa from other cities. And we’re a happy community, which means we’re inclusive, many immigrants move to our city. (Timonen 2024)

Discussion on the findings

The idea for this exploratory study originated from the rapid invasion of artificial intelligence and its alleged potential for work efficiency, cost savings, and creativity. During its brief existence, OpenAI’s language-based model GPT-3 has shown that language can be used to instruct a large neural network to perform a variety of text-generation tasks. However, the findings from the experiment on municipality slogans in Finland were not as expected: The slogans did not differ markedly between the two AI slogan sources. To a large extent, the slogans generated by AI adopted similar wording rather than creating unique associations with the cities. The finding aligns with those of Pankka’s recent study (2023) on corporate slogans, which reports only minimal differences between slogans. The reason for this may be that the questions (in phases AI 1 and AI 2) were presented in English, but the amount of English data on municipalities’ websites is limited compared to that on the Finnish websites. For comparison, a few checks were made in Finnish. These produced more versatile slogans (see Appendices).

One reason for the similarity between the AI-generated slogans may also stem from the fact that GPT-3 is free of charge, and thereby, the access to big data is more limited than in its next version, GPT-4, which is subject to a charge. GPT-4 is forecast to surpass ChatGPT in its advanced reasoning capabilities in the near future (OpenAI 2023). Moreover, Image GPT has shown that the same type of neural network can also be used to generate high-fidelity images.

In spite of the parities between the AI-generated slogans, the interviewees from Kuopio, Vantaa, and Pori saw potential in using ChatGPT for place branding with regard to work efficiency, cost savings, and creating unique associations, but in slightly different ways. Citing Inna (2024), “AI can save money. It is fast, and it can create a lot of versions in just a few seconds”. However, he also raised a question concerning the creativity factor, which is especially relevant given that Pori is known for using an original and unique dialect (see Hakala et al. 2020):

When we think about AI, we think about language models as well. How do the AI-based language models succeed in taking the various dialects, which are important in appealing to regional audiences, into account? Dialect could be something that makes a place different.

Timonen (2024), in contrast, challenged the cost savings aspect but saw potential in the work efficiency and creativity factor: “AI produces more options than savings. AI should produce better results, but it doesn’t produce much saving. The result is better marketing and better slogans”. The same informant pointed out that many municipality slogans do not stand out. The main differentiator is just the name of the city at the beginning, “[When we look at the slogans], we are looking at universally positive adjectives that do not differ or differentiate the place in any way”. (ibid.)

Artificial intelligence is creating a buzz across various businesses and organizations, including municipalities. Commentators forecast that AI can facilitate and speed up the generation of marketing communication ideas and the outline of content for different channels. The city representatives interviewed for this study were also intrigued by AI’s potential. However, they cast doubts on its usability. The insights presented here, both critical and positive, should benefit other place managers and practitioners in the future and also potentially extend the understanding of the options for AI applications. Summarizing the essence of a city in a few words is challenging due to the creativity factor, and thus far, the creative capabilities of AI tools have not attracted as much interest from researchers as their mechanical attributes (Pankka 2023). Whether human- or AI-generated, slogans should be based on the strategic values of the city, as aptly put by Timonen (2024): “A slogan is the shortest way of communicating your strategy”. Any brand’s soul, including a place brand’s, comprises the core values that define the brand. It serves as the guiding principle of the brand-building process. These core values should be built into the place brand, expressed through behaviour and reflected in communications. Ideally, the same should apply to slogans that convey the essence of the brand.

The study at hand offers an initial idea of the opportunities AI presents for place managers. They are the people with first-hand knowledge of their municipality brand. To fully benefit from automated slogan generation, they should be trusted to safeguard the distinct traits that ultimately distinguish their city from others. Brand managers are uniquely capable of and responsible for ensuring that any slogan adopted reflects citizen involvement and the location’s local, unique features.

Concluding remarks, limitations and avenues for future research

While there are conflicting opinions on generative AI and its use in various fields of marketing and branding, it looks like it is here to stay. This study’s respondents thought AI could assist and enhance the creative process in place branding and slogan creation; however, human input into consumer-centric creative ideas and their execution remains critical to its success.

Without proper evidence, AI’s capabilities cannot be expressed. Therefore, AI’s role in creative marketing warrants attention in the future research. This study offers only a starting point for examining the issues around utilizing ChatGPT in creating slogans. It could not address questions of profitability and true effectiveness or AI’s other capabilities. One such capability relates to imagery that municipalities might employ in their logos, symbols, and other promotional material. OpenAI’s DALL·E 3, a new text-to-image generation system, can generate images precisely reflecting the text provided. DALL·E 3 is built natively on ChatGPT, which allows the use of ChatGPT as a brainstorming partner and refiner of prompts (DALL·E 2023). That development would certainly provide an interesting avenue for future research.

Further research is also needed as the content analysis encompassed only 20 Finnish cities. Moreover, as just three expert interviews were conducted, the sample is by no means representative. Similar studies could readily be conducted in other cities and countries. Moreover, municipalities frequently need different slogans to target different stakeholder groups. Here, only one slogan per municipality was considered. Future studies should include and test a range of slogans for municipalities. That advance would call for well-prepared strategy work and carefully formulated questions being posed to ChatGPT.