Abstract: Embodiments of present disclosure relate to systems and methods for predicting potential entities to switch from first to second payment mode. Each entity of a plurality of entities is profiled using transaction pattern of corresponding entities. Clusters for the entities is generated based on profiling. Each cluster includes entities with similar transaction pattern. Adapted entities that are adapted to use the second payment mode is identified. Adapted entities are identified based on tracking spend behavior of each entity. Further, clusters including the adapted entities are identified to be target clusters. Upon identifying the target clusters, predicted entities in the target clusters are determined as potential entities to switch from first payment mode to second payment mode. The predicted entities do not include any entities from the adapted entities. Real-time customized notification is provided to predicted entities, to promote to switch from first payment mode to second payment mode.