The cellular replication dynamics of SARS-CoV-2 are incompletely understood, in particular on a quantitative single-cell and single-molecule level. Lee et al. used single-molecule fluorescence in situ hybridization (smFISH) in three different cell lines to investigate where and when SARS-CoV-2 RNA is produced in cells. They found that initially mainly subgenomic RNA is produced in viral factories and genomic RNA is mostly produced later on. Furthermore, only a fraction of cells (~5–10%) produce substantial amounts of viral RNA, leading the authors to call these cells ‘super-permissive’. The mechanism is unclear at the moment but smFISH analysis of infected hamsters also showed patchy RNA levels. In the cell lines, the authors noted that the SARS-CoV-2 genomic RNA is long lived, which suggests that it evades degradation by nucleases. Interestingly, experiments with the Alpha variant showed slower replication kinetics in the cells than the ancestral variant, which led the authors to speculate that it might trigger the innate immune response less, potentially thereby contributing to its higher transmissibility.