In the broader context of a research project about political participation on Facebook during the Portuguese electoral year of 2019 - for the European Parliament in May, and for the National Parliament in October -, a dataset of more than 9,000 posts and corresponding reactions and comments had been previously retrieved using Facepager. Participating in a data sprint came up as an opportunity to explore some specific questions in an intensive and time-bound context, using a subset of the data tailored to the purpose of discovering: 1) how differently official parties tried to engage with Facebook users before, during, and after the campaigns for the major elections periods (“supply side”), and 2) what type of content received the most attention and engagement (“demand side”). Our results show that regardless of the party’s age and position in the political spectrum, the persistent main categories are “National Politics”, “Fundamental Rights”, and “Self-Promotion” for both elections. Also, we found that posts with images attract more attention, and apparently this may be leveraged if the textual content of the post is in topics of “National Politics”. However, this finding requires further investigation. Along with details about the research during the data sprint and the main findings, this paper is also a testimony about the singularities and learnings of a process built upon the constraints of taking data sprints situation as a workflow