4.07. Generative AI: A Digital Marketing Tool?

Disclaimer / Warnings:

• This sheet deals with Artificial Intelligence (AI), whose development is still rapidly evolving at the time of writing this note. The evolution of this technology may outpace the update of this page.
• Generative AI is a probabilistic tool that can make mistakes: its results and outputs must be carefully analyzed.
• If you use generative AI to analyze personal data, remember to anonymize it to comply with GDPR.
• The use of AI should be supported by internal training and governed by an AI usage charter.
• More information in this section: Understanding the issues and training for responsible AI use.  

For several years now, so-called generative artificial intelligence (Gen AI) technologies have been developing across all areas of private and professional life, spreading at an astounding pace. According to a 2024 study by Hostinger and FLASHS involving 1,000 business leaders and managers from small, medium, intermediate, and large French companies, more than 60% of respondents reported that their organization used AI. Even more indicative of the rapid adoption of these technologies, only 17% of those surveyed said they had never used generative artificial intelligence tools professionally at the time of the survey. Among the cited application areas, data analysis, customer service, and personalized customer journey actions occupy the top three spots.

Cultural organizations also use generative AI tools to optimize their activities. Many of the solutions they already use have integrated AI into their processes, whether in office automation, ticketing, or graphic design. Generative AI, as an assistant for content creation or data analysis, promises to transform the digital marketing of cultural institutions, as it is already doing in other economic sectors, by accelerating the creation of textual or visual content and facilitating the understanding of customer data. However, its use also entails several limitations, biases, and risks that must be acknowledged to ensure a conscious and context-aware application.  

Generative AI tools are capable of producing text, visuals, and videos. The key is knowing how and when to use them.

The value of generative AI tools lies largely in their capacity for learning and adaptation. They can produce content that reflects an organization's identity by drawing on its previous data and publications, thereby respecting its style, editorial voice, and image.

The challenge, therefore, is to properly collect and structure the data that will allow you to personalize your tools. Using generative AI for content creation requires strong professional expertise-both to design effective prompts (strategic thinking) and to evaluate the quality and originality of the output. Generative AI does not replace skill or training. Understanding its uses, maintaining continuous monitoring, and defining ethical frameworks for application have become essential prerequisites. It also calls for new project management abilities, particularly when working with specialized service providers skilled in AI.

AI can also act as a proofreading or grammar correction assistant, suggesting reformulations to improve tone and address. In this sense, AI becomes a partner in the writing process. Once again, the difference between good and poor content will depend on the quality of training in using these tools and on complementary communication strategy skills.

  Some AI use cases:

• Optimize SEO: AI tools analyze trends and relevant keywords to maximize content visibility on search engines.
• Automatically generate subtitles and descriptions: Ideal for videos and multimedia content to improve accessibility and engagement.
• Personalize messages for different audience segments: Tailor tone and style according to the expectations of regular viewers, new audiences, or potential customers.
• Facilitate translation and multicultural adaptation: Allowing content distribution to international audiences without heavy manual work.
• Automate monitoring and trend analysis: Identify topics that interest the public to adjust communication in real time.
• Analyze past publications to suggest content consistent with an institution's editorial line.
• Analyze successful campaigns to guide communication and propose innovative formats.
• Automatically generate tailored texts based on a content base, editorial line, and audience targeting elements.

It is worth noting that many pre-existing software solutions have developed generative AI features without replacing their original purpose. Examples include Canva, with its "magic" writing, image generation, and visual identity features, or Grammarly, the English spelling and grammar correction tool, which now offers rewriting and text suggestion capabilities powered by AI. Likewise, most Adobe, Microsoft, and Google products have also integrated similar functionalities.

These are often the first tools to use when training teams, as they help to understand how AI systems work and to evolve professional practices-by consciously determining when to use them, and when not to.

Generative AI is well known for its potential in writing and visual creation. However, it is not just a creative tool; it can also simplify the management and analysis of audience data-particularly valuable for marketing and ticketing teams with limited resources that have not yet invested in data analysis tools.

AI goes beyond being a mere technical aid: it can extract strategic insights that support decision-making and help optimize marketing actions. However, this requires that the data be well structured and the objectives clearly defined. The saying "garbage in, garbage out" has never been more relevant!

We remind you again that:

• Generative AI is a probabilistic tool that makes errors: its results and outputs must be carefully analyzed.
• If you use generative AI to analyze personal data, be sure to anonymize it to comply with GDPR.
• The use of AI must be supported by internal training and governed by an AI usage policy.

Here are some AI usage ideas:

It is worth noting that many expectations surrounding generative AI actually concern applications that can be achieved without it. Its impact is not insignificant, and it is therefore essential to define clear parameters for its use.

• Develop audience segmentation strategies: AI can analyze your databases to identify groups with similar behaviors (e.g., levels of loyalty, audiences to re-engage) and recommend targeted actions.
• Detect audiences at risk of disengagement: by spotting spectators whose attendance frequency is decreasing, AI can help design personalized re-engagement and loyalty campaigns.
• Optimize marketing campaigns and pricing: through the analysis of purchasing behaviors, AI can recommend price adjustments based on demand (yield management) or suggest the best times to launch promotional offers.
• Automate reporting and data visualization: some AI tools can generate clear, dynamic dashboards, making data analysis accessible to all teams, even without technical expertise. Once again, thorough data preparation is essential, and teams should assess whether to use generative AI tools or rely on business intelligence and dashboard solutions.

The use of generative artificial intelligence raises several ethical, legal, qualitative, and environmental issues that are important to understand, know, and master in order to use these technologies consciously and responsibly.

The use of AI in the cultural sector must be thoughtful and regulated to avoid abuses that could affect content authenticity, data protection, the environment, and employment.

Avoid dependence and homogenization of content

Studies on the impact of generative AI tend to show its detrimental effect on users' critical thinking abilities¹, which is why it is important to remain in control of the tool and use it in a reasoned and critical way regarding its outputs.

• Generative AI should be used as an assistant tool and not as a substitute for human creativity.
• It is a statistical technology, capable of inventing information, book titles, or scientific articles based on probabilities. Therefore, a degree of skepticism toward its outputs is necessary, and sources should be requested.
• Personalizing and reinterpreting AI-generated content is crucial to preserve a strong and distinctive editorial identity.

Protect personal data and comply with GDPR

• Cultural institutions must choose solutions compliant with European regulations and avoid integrating sensitive data into uncertified tools.
• Inform the public about data usage and prioritize local solutions or AI trained on secure and transparent databases.

Limit algorithmic biases and ensure diversity of representation

• Verify that the tools used do not reinforce stereotypes in marketing campaigns, programming, or audience analyses.
• Combine algorithmic analysis with critical human review to avoid errors and discriminatory biases.

Limit environmental impact through responsible use

• Generative AI usage exacerbates a trend already underway that predicts a massive increase in the carbon footprint of our digital activities-potentially up to 60% by 2040 according to a joint study by ADEME and ARCEP².
• While generative AIs are energy-intensive during training, they do not all remain so during operation³.
• Favor less energy-consuming AI solutions and limit superfluous usage (avoid systematically generating content).
• Raise team awareness about the environmental impact of AI queries and encourage responsible digital practices.

Anticipate work transformations and support teams

• Train professionals for a critical and informed use of AI so that it becomes a lever for innovation rather than a constraint.
• Encourage active monitoring of changes in professions impacted by AI, especially in administrative and creative sectors.
• Reflect on the risk of job replacement by AI (photographers, authors, graphic designers...), the impact on creative quality, and the institution's social responsibility in this area.

Generative AI is a powerful lever for cultural organizations, facilitating content creation and data analysis. However, it is an opportunity to seize with discernment. Its use must fit within a global strategy and take into account the ethical, ecological, and human challenges outlined above. A tool remains a tool: it is up to humans to give it meaning. By training and adopting a critical approach, cultural professionals can harness these technologies while preserving the uniqueness of their institution. Training will be the best way to ensure a reasoned and situated use of these technologies-understanding core functions, recognizing and avoiding algorithmic biases, and identifying relevant use cases. This will allow relying on AI without giving it a blank check.  

• [FR] Café TMNlab : IA et data pour mieux connaître ses publics (external link)(accès réservé aux adhérents TMN Lab), avec Karine Rahmani, consultante en marketing et fondatrice de La formule Marketing, décembre 2024
• [FR] LinkedIn sous l’emprise de l’IA : 54% des contenus seraient générés par des robots (external link), Guillaume Fleureau, novembre 2024
• [FR] Les biais algorithmiques, fatalité ou symptôme ? (external link) "Les biais de l'intelligence humaine et de l'intelligence artificielle", Thibault de Swarte, Jean-Baptiste Janvier, Institut Psychanalyse et Management, nov. 2021
• [FR] Les Entreprises Françaises à l’Heure de l’Intelligence Artificielle et du No-Code (external link), Chaimaa Chakir, Hostinger, 2024
• [FR] « À l’heure de l’IA générative, nous devons nous interroger sur ce que signifie le rapport à la productivité dans notre secteur. », (external link)TMNlab, juillet 2024
• [EN] We should not rely on AI to solve the problems of live performance discoverability (external link), Linked Digital Future, may 2023
• [FR]  Se former à ChatGPT (external link), PMD – Le Podcast du Marketing Digital, mars 2024
• [FR]  Café TMNlab : Découvrabilité à l’ère de l’IA, pourquoi se mobiliser ? (external link) (accès réservé aux adhérents TMN Lab), une heure avec le chercheur Jean-Robert Bisaillon, avril 2024
• [FR] Café TMNlab : IA générative et pratiques créatives (external link) (accès réservé aux adhérents TMN Lab), avec Elisabeth Gravil, consultante en stratégie digitale pour le secteur culturel et particulièrement muséal, fondatrice de Museovation et enseignante, juin 2024
• [EN] How arts and culture organisations can prepare for AI (external link), Joe Perkins & Jocelyn Burnham, 2024
• [FR] Impact de l’IA générative sur le spectacle vivant : les métiers techniques et la communication en première ligne, les effets pour la filière encore difficile à identifier (external link), TMNlab, Juin 2024
• [FR]  L’impact de l’IA sur la découvrabilité dans le secteur culturel (external link), Guy-Philippe Wells, LATICCE, avril 2024