Communication changed a lot over the past decades and many elements triggered the swing: among others, the spread of smartphones drastically reduced the costs of producing multimedial contents in real-time and allowed for the diffusion of the so-called citizen- or participatory journalism. Social media, on the other hand, multiplied the potential channels where campaigns and topics, rather than being merely consumed, are commented, possibly criticized, and shared by users that actively concur to the success or failure or communication strategies. The new “habits” urge companies to innovate their own Communication, in order both to adapt to such challenges and opportunities, and to efficaciosly compete for nowadays new oil, i.e. people’s attention. Unsurprisingly, data turns out to play a pivotal role in this scenario.

Eni is at the forefront of this, as it can count since 2015 on the so-called Eni Datalab, a dedicated unit for the systematic analysis of data in support of communication activities and beyond, both for an effective measurement of its reputational KPIs, and for an enhanced awareness of its stakeholders and audiences.
In more details, thanks to its methodologies, tools and experties, the Datalab aims at allowing a deeper and proper understanding of its stakeholders’ needs, attitudes and curiosity, in order to correctly direct its own communication activities and be responsive to actual demands.

Three main chapters are interconnected in the Datalab: measurement, insights and data science. As for the first, the Datalab actively creates monitoring systems that allow a real-time scan of Eni’s communication activities, aiming at measuring its effectiveness through custumized KPIs. The monitoring methodologies applies to the communication ecosystem overall, in order to gain the necessary insights to support content strategy and media production on topics of interest. Data Science support trasversally both pillars and the examples are numerous: Machine Learning algorithms and advanced NLP (Natural Language Processing) models are developed in house in order to facilitate and potentiate the processing and analysis of thousands of public documents published by media, national and international organization and all the main stakeholders. Deep Learning neural networks are trained in order to have an accurate forecast of audiences’ interests on the main topics of Eni’s editorial plan. Advanced data visualization techniques are employed to create nested dashboarding systems that are custumized for operational and strategic decisions.

Social Network Analysis (SNA) is obviously an important chapter for Eni Datalab, given its intrinsic interest in discovering interaction patterns in discussions and to extract thematic communities and relevant influencer on specific issues. Community detection algorithms are employed on a daily basis in order to detech whether new publics are joining the online debate on Eni and on its main topics of interest, by examining the typology and the intensity of intra-community linkages. Further algorithms are deployed in order to highlight emerging influencers that should be considered as such not in the light of some static and uninformative characteristics (e.g. number of followers), but rather on the basis of their position within the network of the stakeholders.

Among other things, data science techniques are employed for gaining the main insights in the data journalism chapter hosted by eni.com. Topic detection algorithms are employed in order to understand what are the main themes by Italians under the holiday umbrella: Main destination for 2020 vacations in Italy. Social network algorithms are employed in order to create a semantic map of the two discourses of the US presidential candidate: Major themes in US elections.