Grain growth in polycrystalline material is influenced by numerous factors which convolute the understanding of the process. Present work intends to delineate certain aspects of this complexity by quantitatively analysing the topological evolution of the grains and its first-neighbours resulting from multiphase-field simulations. Upon verifying the consistency of this approach with the existing 2- and 3-dimensional analytical predictions, face-switching events of about 75,000 grains in 3D domain is extensively analysed. Despite the expected numerical dominance of the face-loss event, we find the affinity of this event to decrease with increase in face-class. A transition in ‘switching affinity’ is observed above face-class 14, wherein, the face-gain events are preferred over loss events. The pathway of the topological evolution is comprehensively analysed to identify its influence on the life-span of the grains. This analysis indicates that the topological evolution around the average face-class, ‘dead zone’, invariably shortens the life-span of the grains and considerable gain in the face-class is seldom observed. The topological behaviour of the first-neighbours during grain growth is quantitatively captured through a well-known statistical tool called heat-maps. And for the first time, the formation of ‘topological clusters’ which account for the time-invariant behaviour of the grains is elucidated.