Network-Aware Campaign Optimization in Telecommunications Under Communication Constraints

This study addresses the campaign optimization problem in telecommunications, focusing on determining how, when, and through which channels to engage customers most effectively. Recognizing that customer decisions are shaped by social interactions, we construct an enriched customer network by integrating Call Detail Record (CDR)–based social networks with tariff-change data, in which the temporal proximity of tariff changes among connected customers indicates potential influence relationships. A mathematical model is then formulated to optimize campaign message allocation under communication, channel, and regulatory constraints, while explicitly incorporating such influence over the network during the planning period. To handle large-scale instances, a simulated annealing–based heuristic is developed that embeds a linear programming component and a network-driven customer prioritization strategy. Computational experiments demonstrate that incorporating network effects into the decision-making process results, on average, in a nearly threefold (2.9x) improvement in the objective value over solutions that ignore them during decision-making. Furthermore, for large-scale instances, the proposed approach consistently achieves high-quality solutions, providing an average improvement of at least 17.2% in solution quality compared to the other three heuristics tested, leading to more effective and better-targeted campaign strategies.