Розмір шрифта: 

This paper addresses the development of a software tool to prioritise and select the next communication action for clients of banking institutions within a CRM environment. The topic is relevant because static mailing rules and fragmented integrations between CRM platforms, digital channels, messaging services, and operator workplaces do not provide the required degree of personalisation, timeliness, and communication consistency. A customer prioritisation and next-best-action model is proposed that takes into account RFM indicators, lead score, response probability, contact frequency constraints, consent to personal data processing, the optimal sending time window, and the relevance of a banking product. The approach to CRM-platform integration with banking communication channels is further developed by combining fuzzy-attribute matching for client-record deduplication, interaction-event logging, role-based access control, and idempotency control for API calls. To validate the approach, the paper describes a software prototype and a scenario-based experimental evaluation using a synthetic dataset that is structurally aligned with retail banking processes. The obtained results demonstrate a 19.4% increase in customer-base segmentation accuracy, a 16.8% increase in average response rate, a 24.3% reduction in duplicated client cards, and a 28.7% reduction in synchronisation errors between the CRM and external services. The proposed software tool may therefore serve as a foundation for modernising the communication layer of banking institutions.

banking institution, CRM, next-best-action, customer prioritisation, RFM analysis, lead scoring, omnichannel communication, record deduplication, event logging, API idempotency