Writing our own chapter in data & media history

The media business is growing more complex by the day, in turn increasing the trend towards a data-driven company culture across all fields and countries. This type of shift in a company‘s mentality demands not only time and resources, but also the courage to think and act differently, all the while facing the challenge of making sure everyone is on board. 

Artificial intelligence, machine learning, data governance, and metadata management all play a central role, a role which is here to stay. That’s why we’ve made a timely investment in a strong combination of data, analytics, and media expertise in order to develop and maintain our competitive edge as well as be fit for the future. 

Our cross-media setup, where TV, print, digital, audio and ATV all intersect, provides us with billions of data points we analyse on a daily basis and use to create models. We’re not waiting for trends to appear. Rather, we detect them and use our strengths to play an active role in shaping them.

We rely on a data mesh logic, making us an important interface for all business units at RTL Deutschland. We gear everything to their needs, and we also develop tailor-made data solutions and strategies for our users and market partners with the help of machine learning and AI. 

Here are two examples to make it a bit more tangible: 

Together with our marketer, Ad Alliance, we have developed BrandAID and created a USP that benefits our colleagues in their consultations with clients. BrandAID is designed to increase the impact campaigns can make. We provide important insights for this, because the AI-based tool measures the advertising impact across media and can also be used in the future for forecasting and simulating advertising impact. 

We are also developing data solutions that underpin and ideally increase the appeal to our user base, as we are doing with RTL+. Our streaming offer serves the ongoing trend of mobility and individualisation and, in the future, will also serve hyper-personalisation.

In short, it’s an entirely new level of data-led recommendations. Hyper-personalisation is based on our fundamental understanding of content and usage. Thanks to the combination of various media genres, user profiles at RTL+ are content hubs and central points of discovery. 

Based on data, we connect what we understand from user profiles to what we know of our content and are able to pinpoint our users’ interests, offering them the right content at the right time and in the right format, whether that means reading, listening, or watching. It’s a milestone in terms of recommendations and catering to one’s individual interests! 

We‘ll continue optimising it and keep an eye on all developments, for which a couple of ideas regarding visual presentation have already been tossed around. As part of our goal to actively shape the global technological and cultural shift, we’re not only keeping our eyes and ears open for other global developments, but also establishing smart partnerships in order to give things a try.

The most recent example is Microsoft. We‘re working together with them on DALL∙E 2. This project is centered around a machine learning model trained in Microsoft Azures OpenAI infrastructure which is able to generate images based on descriptions provided in natural language. 

We at RTL Data were able to use and explore this technology as test clients. Trying out how we could use our metadata to create individual and tailored visuals for audiobook chapters or podcast episodes got many of us very excited and we have lots of ideas about how and where we can implement this sort of technology in our RTL universe. 

It would allow us to design images for RTL+ that match our subscribers’ preferences. With over three million subscribers, that’s a task that would be impossible without artificial intelligence.

We want to write our chapter in the history of data and media. 

But what we want above all else is that handling data becomes part of every single one of our 7,500 employee’s DNA. And that begins with working with data. That’s why we’re currently working on making our treasure trove of data accessible for all internal employees, for which we’re also developing a complex reporting centre. 

This should allow everyone to easily and quickly extract the widest range of analyses. “Integrated thinking and working” is our motto and goal.

Regardless of the increasing fragmentation, it pays to know one’s target groups precisely, especially in times of crisis; to know what interests them and how they consume media. Accordingly, we invest in audience surveys, ratings analyses and even in-depth psychological studies to find out how the world one lives in and your needs, values and views change; what role and importance brands have in this structure and what effect is achieved. 

The fact is: people follow content, not channels. 

With our know-how and our cross-media approach, we are well positioned even in the crisis. We are also well positioned for the future, because here the picture of cross-genre thinking is emerging. A silo mentality is already outdated and therefore the trend in media planning is moving towards cross-media campaign planning. In marketing, we are increasingly working on solutions that do away with the boundaries between genres – both in advertising products and in playout logics and reporting. Data is a great help here – in terms of ex-post analyses or possible impact forecasts.

Communication needs attention, only then can the effect unfold. That has always been the case. However, the ongoing crisis has brought about a return to essentials, fuelling discussions about attention and its measurement criteria. At the same time, it’s a discussion about quality, because the two central drivers are advertising creation and the advertising environment – both of which are thus moving back into the focus of analyses. With the “cookieless future” in our line of sight, the environment is gaining importance again. 

However, measuring the construct of attention is not trivial and can best be depicted by a combination of methods from eye tracking as well as qualitative and quantitative surveys. Basic studies are essential here, such as those conducted by Screenforce. The genre initiative of the German-speaking TV marketers has currently initiated a broad study that examines the connection between creation and environment on the advertising impact. In addition, we at RTL Data are working on the development of various tools that predict the advertising impact of creations or support the selection of suitable environments.

Karin Immenroth, Chief Data & Analytsics Officer at Mediengruppe RTL Deutschland