Research on regional linguistic variation typically involves data collection in the field. This process itself can take up several months if not years. In the present study we demonstrate how we can use web interactives in collaboration with media outlets for a fast gathering of regional, sociolinguistic data. In collaboration with SPIEGEL ONLINE and Tagesanzeiger, we developed a web interactive that predicts users’ regional backgrounds from within German-speaking Europe. More than 1.9M people have participated in the interactive, more than 770K users have provided metadata. Said metadata allowed us to capture regional variation in language as of today, which we can compare to historical survey data–enabling us to track the evolution of German in Europe over the past 40 years. We report on regional levelling of lexical variants, a process which appears to be particularly prevalent in the northern parts of German-speaking Europe. We further found an effect of (former) national and regional borders on language use. This innovative paradigm allows us to collect sociolinguistic data of an unprecedented scale–at the same time it presents significant challenges, both of which–benefits and challenges–will be discussed in this contribution.