Nowadays millions of people all over the world use social media platforms to keep up with the news, to express their feelings and ideas, as well as to share or debate opinions on virtually every conceivable topic. This justifies the interest of National Statistical Institutes (NSI) towards social media as a means for “measuring” the public mood. In recent years, the Italian National Institute of Statistics (Istat) has been investigating whether social media messages may be successfully exploited to develop domain specific sentiment indices, namely statistical instruments meant to assess the Italian mood about specific topics or aspects of life, like the economic situation, the European Union, the migrants’ phenomenon, the terrorist threat, and so on. To this end, Istat researchers have developed procedures to collect and process only social media messages containing at least one keyword belonging to a specific filter, namely a definite set of relevant Italian words. Domain-specific filters have been designed by subject-matter experts with the aim of filtering out since the beginning messages that would very likely turn out to be off-topic for the intended statistical production goal. Istat has recently released a new experimental statistic, based on Twitter data: the Social Mood on Economy Index. The index provides daily measures of the Italian sentiment on the state of the economy. These measures are derived from samples of public tweets in Italian, which are captured in real time. Similar initiatives have been put in place by other NSIs in recent times, notably Statistics Netherlands’ attempts to “mimic” the time evolution of the Dutch Consumer Confidence Index by means of a sentiment index based on social media, and to derive daily measures of social tension from Twitter. This paper provides an overview of the production pipeline of Istat’s Social Mood on Economy Index.