From big data to artificial intelligence: how the audiovisual sector is being transformed
The audiovisual sector is undergoing a silent but profound transformation in which data and artificial intelligence are no longer ancillary elements but tools that redesign the entire supply chain: from creative decisions to programming and distribution. This topic was discussed recently in Rome, at the ANEC headquarters, during a meeting entitled "Cinema Data: data-driven strategies for distribution, programming and innovation", promoted by CrewReel Services and the Cinema Exhibitors Association, from which three crucial issues emerged.
The first concerns the interoperability of data. Between box office, admissions and socio-demographic profiles as well as qualitative research, cinema produces a huge amount of information that is often analysed with different methodologies and is not easy to compare; in this respect, AI can play a significant role. It is Federica D'Urso, Professor of Film and Media Economics at La Sapienza University of Rome, who explains: "The impact of AI on the audiovisual ecosystem differs from previous digital disruptions because it does not only concern creative or production processes, but also the ability to read and connect data in the sector. Artificial intelligence tools, in fact, make it possible to extract much deeper and more granular insights, enhancing information that is often dispersed or underused. A decisive advantage is interoperability: AI can bring together today's fragmented databases and sources, creating shared metrics and integrated overviews. This enhances industry transparency and enables more strategic decisions. In other words, AI does not replace the experience, but amplifies the view'. Therefore, it should be emphasised that data alone do not provide meaningful insights as they need to be mapped, integrated and interpreted: therefore, without a common basis, there is a risk of building decisions on partial and misleading readings.
The second issue is the closing of the 'funnel' that starts with promotion and ends with ticket sales. Currently, those who promote a film know the audience's awareness and interest, while those who manage the cinemas measure conversion, i.e. how many actually buy the ticket: combining data from these two sources would allow more precise campaigns and more reliable evaluations. Davide Novelli, Senior Vice President of Distribution Piper Film, emphasises that technology must serve to make fewer mistakes and not to complicate the supply chain, adding: "From the production and distribution point of view, the fundamental metric remains the ROI of the operation throughout the entire chain of exploitation of the product: the financing and its margin are realised in the different channels that make up the audiovisual world, from the theatrical release to passages in Home Entartainment, TVOD, SVOD, AVOD, FreeTV, as well as the long life that a film produces when it becomes part of a library until the distribution or author's rights are exhausted. The recent comparison between the financing of a film and the result in cinemas does not take into account the audiovisual world as a whole, which generates much larger economies of scale: this is not an easy balance to achieve, which is why public financing of the audiovisual product is necessary in almost all countries of the world to make the market sustainable. From a promotional point of view, the quest of any marketing or programming team is to reach and satisfy the final customer, measured in terms of reach and engagement in the various promotional channels during the launch of a film. Unfortunately, the disconnect between those who promote a film and those who show it in cinemas does not allow for a direct metric of campaign conversion." Today, for example, chains such as The Space Cinema already exploit purchase data to better understand conversion and build sales-oriented campaigns.
Finally, the issue of the conscious use of artificial intelligence should be highlighted. Predictive models and generative tools speed up the creative phases and support programming, but cannot replace editorial responsibility: "The room must remain a place of guidance and surprise, not an algorithmic bubble," recalled Carlo Rodomonti of ANICA, while Silvia Paonessa, CEO of CrewReel Services Srl, commented: "Artificial intelligence is a tool, it depends on how we choose to use it. Generative models tend to repeat what they know: if we just follow them passively, the risk of homologation is high and real. But if we integrate them into a conscious creative process, AI can become an accelerator of imagination. The real danger is to become 'creative automatons', letting the tool take over thinking. Instead, I believe that today, with these tools, stories can emerge that had no place before. What is needed, however, is editorial direction, creative courage and new skills capable of combining artistic vision and technological literacy. Only in this way does AI not risk standardising, but liberates new narrative forms. In my work I see how AI can simplify processes and free up resources, but the central point remains the same: protecting the creative voice. Automation must amplify, not replace. To achieve this we need ethical rules, transparent tools and new skills capable of combining technique and vision. The author must remain at the centre, but in a changing context: his work must be accompanied, enhanced and made adaptable to the evolution of the industry'.
Audiovisual marketing is facing an evolution that will no doubt continue over time, but which changes are expected and which are likely to be averted? "In 2035, Italian cinema will have integrated generative AI into production both to optimise costs and to diversify the type of products. Fantasy, sci-fi will be genres that can be realised even with reduced budgets. Excessive and indiscriminate use in conception and writing should be avoided. The risk is the flattening and trivialisation imposed by linguistic models based on already written stories and the most statistically probable solutions. The risk of losing originality, depth and identity is high. We must avoid creating tailor-made films for viewers based on data. cinema must surprise, make people think and not be condescending to the ChatGpt model,' concludes Carlo Rodomonti, President of the Publishers' Union and Digital Creator - ANICA, on the subject.
