From Model T to AI: The Evolution of Personalisation in Media
In the past decade, the rise of individualisation has marked a significant trend in modern society. This movement towards tailoring products to the unique tastes and needs of individuals has driven growth across many industries.
The demand for personalised experiences in retail and services, combined with increased empowerment, autonomy, and flexible work arrangements spurred by the pandemic, has heightened expectations for products and services. People yearn to stand out, to be unique, to be different, and to be treated as such.
The one-size-fits-all selling approach epitomised by Henry Ford and the Model T — available in any colour you like, as long as it is black — is obsolete. Today, cars offer myriad customisation options where buyers can assemble their product using online configurators.
Harnessing Digital Technology for Personalisation
Digital technologies have actively enabled this trend, allowing intricate customisations across many sectors. Production technologies can now handle complex configurations and variations, while online systems allow customers to configure and visualise products through simple drag-and-drop interfaces.
Online configurators for furniture, apparel, footwear, eyewear, kitchen appliances, and even hamburgers enable people to tailor products to their specific preferences and needs. 3D printing even allows individuals to build their own products at home.
This trend of individualisation not only reshapes retail and other industries but also affects news media, demanding innovative approaches to content delivery.
The State of News Media Personalisation
Considering this backdrop, how have news media companies embraced this trend?
Most news media companies offer some level of personalisation on digital platforms. Readers can choose their favourite topics to create their own content feeds or select a preferred location. They can sign up for newsletters based on their interests, and “smart” algorithms provide recommendations for articles in a “further reads” section. Readers can adjust aesthetic preferences such as font size or night mode.
Some platforms allow users to listen to articles read aloud, and some media houses present a different selection or prioritisation of content based on subscription status. However, personalisation often stops here. It rarely goes beyond deciding what content to present.
News organisations continue to deliver the same story format to all readers — still primarily text — using the same journalistic forms such as reports, analysis, or interviews, with the same length, structure, language, and tone. Yet many additional parameters could be used to create a more distinctive experience for each reader. Imagine customising a news feed not just by topic, but by the situation a reader is currently in or by the kind of day they are having.
Motivation and Life Situation
People have different reasons to consume content, and this motivation can change throughout the day or across different situations in their lives.
Sometimes readers simply want to be quickly informed or updated about events. At other times they want deeper explanations, education, or inspiration. And sometimes they seek help to solve a problem, entertainment, or simply the pleasure of being surprised.
The concept of user needs, popularised by the BBC, is increasingly adopted by news media companies around the world. BBC user needs focus on tailoring content to specific audience motivations identified through extensive research into media consumption. The framework categorises needs into four main areas:
- keeping audiences informed
- providing deeper understanding
- offering new perspectives and stimulating creativity
- providing relaxation and enjoyment
Newsrooms attempt to create and manage content accordingly. For example, the same story about an event can be presented as bullet points, a fact sheet, a summary, a table, an image, or an audio bulletin to fulfil the need to keep the audience informed. If the same story appears as an 800-word analysis, a podcast, or a video, it addresses the need for deeper understanding.
The key is recognising the personal need of the reader at a specific moment. Generative AI (GenAI) and large language models (LLMs) can quickly and cost-effectively create different versions of the same story to serve these varying needs.
Mood Management
Another potential parameter for personalisation relates to mood management, a concept from media psychology that explores how individuals use media to influence or regulate their emotional state. The theory suggests that people often choose media that counterbalances their current mood. For example, someone feeling sad might watch a comedy. However, the concept of **semantic affinity** suggests that people sometimes select media that aligns with their current mood, particularly when the mood is positive or when they wish to sustain a certain emotional state.
While this may seem contradictory, it reflects the complexity of human emotions and the different strategies individuals use to regulate them. As a result, mood-based personalisation is more complicated than other forms of personalisation. Nevertheless, recognising the emotional state of a reader and adapting content accordingly can contribute to a more personalised experience.
Simple examples already exist on social media platforms. Users can disable the visibility of sensitive content on X, hide offensive words on Instagram, or use profanity filters for comments on Facebook. Going a step further, GenAI and LLMs could adapt the style, language, and tone of content to match a reader’s emotional state without changing the factual message of a story. Readers could create a form of “safe space” for themselves without resorting to avoiding news altogether or escaping into cat videos simply to avoid distress.
The Challenge of Knowing
The main challenge is recognising which user need should be fulfilled at a specific moment. Certain assumptions can help guide the process. For example, people may prefer more informational formats in the morning and more reflective or analytical content in the evening.
Direct feedback from readers could further refine this understanding. Interaction mechanisms — such as simple swipe gestures similar to Tinder, allowing readers to signal “this met my need” or “this did not work” — could provide valuable signals to the algorithm. Such interactions would also give readers a sense of control and influence over their personal news experience. Even though mood-based personalisation is complex, expanding personalisation beyond simple “suggested reads,” topics, or geographic preferences represents an important and promising area of innovation.
Next-Level Personalisation Is Not Optional
Sophisticated personalisation depends not only on deep empathy for readers and understanding their needs, situations, and moods, but also on a strong data strategy and technological infrastructure. Systems must seamlessly manage everything from collecting relevant data to processing signals and delivering content nearly in real time.
Challenges remain. These include technological limitations in data processing, algorithm accuracy, and the complexity of interpreting user signals. However, overcoming these challenges is essential. Advancing towards more personalised news consumption is not optional — it is imperative. Embracing innovation in AI will help media organisations align with evolving reader expectations.
Achieving this balance means that the fusion of advanced AI with empathetic design will not only improve the user experience; it has the potential to fundamentally transform the nature of news consumption itself.
This article was also published on the INMA Media Leaders blog.