Viruses encounter numerous bottlenecks in moving from the primary site of infection to sites that are responsible for disease. Establishment of a productive infection requires the viral population to survive these bottlenecks. What this means practically is that a virus population requires the ability to withstand multiplicities of infection (MOIs) far lower than those used for a one-step-growth analysis, an experimental condition in which genetic complementation can even suppress lethal mutations. We have developed microfluidics-based approaches to study viral infection dynamics and cell pairs at a range of MOIs.  We will share some of the fascinating observations made using our single-cell and cell-pairing methods, including our recent discovery of a viral non-structural protein capable of hijacking cellular secretory macroautophagy pathway to promote non-lytic spread of an enterovirus.