We study the early stage coagulation kinetics for a charged colloidal dispersion which is here modeled by an effective two-body colloid-colloid potential. The colloidal system was physically prepared by choosing sets of colloidal parameters varying in particular the Hamaker constant and the particle's size. The kinetics of coagulation process was driven by the addition of an indifferent electrolyte and assumed to proceed in two quasi-steady steps. In the first step, colloidal particles are destabilized by the presence of a second potential minimum to diffuse from a bulk-stabilized liquid phase to a flocculated phase. In the second step, we assume that different entities are found in the second potential minimum. The entities comprise secondary dimers, secondary dimers undergoing redispersion, and monomers still in singlet states. If, under favorable condition, this kind of interaction-driven diffusive motion continues, a fraction of the secondary dimers will be induced to undergo primary dimers formation in the first deep minimum. Whether or not the latter process occurs is determined either energetically by the potential barrier falling below a prescribed value, say of 15k(B)T, or/and the second potential minimum becoming negligibly small (with a magnitude < k(B)T). A prototype example to exhibit this kind of the coagulation kinetics phenomenon is an aqueous dispersion of polystyrene latex particles. Our detailed analysis on this system showed the connection between the change of rate constants and the reversible flocculation reversible arrow irreversible coagulation transition and would throw a fresh light on the use of both the energy and the kinetic criteria for understanding the colloidal stability such as those observed in the liquid-liquid coexistence.