Social media, such as Facebook and Twitter, offer new ways of communication for companies, politicians, or celebrities to get in direct touch with their customers, the public, or fans. Thereby, online posts undergo different information diffusion processes. While the majority of posts van-ishes in the vast flood of data, single posts – called buzzes – receive an extraordinary attention and become viral. Previous research indicates that different types of buzzes, i.e. firestorms, lovestorms, and hot topics, exist. However, researchers have not yet distinguished between them in their analyses. This study examines firestorms, lovestorms, and hot topics by analysing varia-bles describing their characteristics and comparing them. Drawing on a manually classified da-ta set of more than 60,000 Facebook posts, we emphasise that a differentiation between the buzz types is necessary for future analyses. We underline that content-related variables, like the mean comment’s length, are useful to describe firestorms and hot topics. Temporal variables, like the activity within the first hours, help to distinguish firestorms from lovestorms and hot topics. Us-er-related variables, like the number of new participating users, are good predictors for all three buzz types while engagement variables, such as comments, likes, and shares, yield surpris-ing differences among the three types.