The human circadian timing system depends on the light/dark cycle as its main cue to synchronize with the environment, and thus with solar time. However, human activities depend also on social time, i.e. the set of time conventions and restrictions dictated by society, including Daylight Saving Time (DST), which adds an hour to any degree of desynchrony between social and solar time. Here, we used Google Trends as a data source to analyze diurnal variation, if any, and the daily peak in the relative search volume of 26 Google search queries in relation to the transitions to/from DST in Italy from 2015 to 2020. Our search queries of interest fell into three categories: sleep/health-related, medication and random non sleep/health-related. After initial rhythm and phase analysis, 11 words were selected to compare the average phase of the 15 days before and after the transition to/from DST. We observed an average phase advance after the transition to DST, and a phase delay after the transition to civil time, ranging from 25 to 60 minutes. Advances or delays shorter than 60 minutes, which were primarily observed in the sleep/health-related category, may suggest that search timing for these queries is at least partially driven by the endogenous circadian rhythm. Finally, a significant trend in phase anticipation over the years was observed for virtually all words. This is most likely related to an increase in age, and thus in earlier chronotypes, amongst Google users.