Nowadays, Statistica estimates that Facebook monthly active users are approximately 2,2 billion around the world. Other social networks are following the same exponential rise.

As a consequence, social networking platforms are gathering a huge amount of data by making connections between people and reaching them through generated content.

The role of Artificial Intelligence (AI) takes place when all the collected unstructured information has to be managed.

The rise of machine learning and its impact of brands strategies and user experience

Machine learning has definitely been a key trend of the past decade. It has had a tremendous impact on social media and companies are starting to understand what is at stake in using AI in their social media strategies.

According to Mia Tawilé, Freelance Digital Consultant,

the most interesting thing of observing these trends is understanding the way businesses and brands can use them to build stronger relationships – and even loyalty – with their clients.  

Thanks to machine learning, brands can easily identify their target. This type of AI allows to extract information from social media and make it highly valuable for companies. Thanks to the data they collect, a brand can convert prospects into customers by pushing the right product to the right person at the right time.

The interview we had with Pascal Nguyen, a former data analyst at LVMH & Sephora allowed us to get some insights on the use luxury brands can make of these technologies. In this sector, “it is all about creating a tailored experience for customers”. Artificial Intelligence allows them to deliver a personalized message, at the right moment to their prospects on the customer journey. On the other hand, the automated analysis of the performance of each channel helps them optimize their campaigns on a large scale which is usually consuming a lot of human resources.

Social media users, on their side, can also benefit from the rise of machine learning.

It allows us to optimize the publications and to sort the informationChristophe Tricot, Artificial Intelligence Expert

Social networks can decide which information brands want to push to users and the user is not overwhelmed by the amount of information. Having personalized content delivered to users has enhanced brands’ e-reputation. This has surely changed the brands’ and users’ experience on social media.

How machine learning helps social media to be more ethical

AI can solve violence and aggressivity issues that have been growing on social media: moderating violent, insulting and aggressive content for example, or as I mentioned in my article on Siècle Digital, decrease bullying issues by identifying the content that causes it. Mia Tawilé

The rise of social media has brought harassment and hatred behavior to a new scale. A famous yet sad example would be the misuse of Periscope as a tool to live stream tragic events such as suicides.

Machine learning has helped social networks prevent this kind of use of their platform. Recently, Instagram has created a filter that detects when published content causes harm to users, their appearance or their mental and physical health.

Facebook has also used Artificial Intelligence for suicide prevention: AI is able to know if a post on Facebook has a suicidal tendency. The technology will make it possible to process these distress messages faster, more efficiently and to prioritize them.

To conclude, such examples show how machine learning and AI can bring an ethical dimension to social media and, again, change and improve the user experience. As last example, this video is presenting how AI can help persons with sight deficiencies to listen to description of images on Facebook.

Louise ChapuisMarion Lefébure | Lucas Brossier | Sergio Gonzalez Andrade | Marine Lextrait


This article was written by students on the Grenoble École de Management Advanced Masters in Digital Business Strategy. The task was to create an article within 24 hours including varied media and to share it. The students were responsible for all aspects of writing, editing and uploading with no input from the lecturers. It has, therefore, not been checked by us before uploading. Thanks to everyone who took part by agreeing to be interviewed, answer questions and share the content.

James Barisic

Lecturer, Grenoble École de Management