Abstract
This thesis investigates how behavioral data from trade fair event applications can be transformed into actionable insights to support decision making, and the valorization of digital products. The work is conducted in collaboration with Italian Exhibition Group (IEG), a leading player in the trade fair industry, and focuses on the official mobile application used by visitors and exhibitors during major events. The main objective is to design a structured, scalable, and repeatable data-driven analytics pipeline capable of replacing the current Excel-based reporting flow with a more robust, standardized, and reproducible process. The methodological framework adopted is CRISP-DM, with particular emphasis on the Business Understanding, Data Understanding, and Data Preparation phases, which constitute the core of the contribution. The empirical analysis is centered on a flagship event used as a case study and starts from raw event logs exported from the application. The proposed pipeline includes data cleaning and normalization, event log modeling, feature engineering for behavioral analytics, and the definition of a consistent set of metrics and KPIs describing app engagement, exhibitor visibility, performance of digital advertising formats, and usage of advanced functionalities such as lead scanning. The results show that the transition from a fragmented, manual, and operatordependent workflow to a codified pipeline enables IEG to systematically extract insights from existing tracking data, improve the communication of ROI to exhibitors, and support strategic decisions on the evolution and monetisation of digital products. From an academic perspective, the thesis provides an applied example of CRISP-DM in the context of mobile event applications and trade fair behavioural analytics, highlighting how a structured approach to data preparation and process standardisation can generate significant and lasting value for organizations.

Login