Abstract | An increasing number of companies are choosing to outsource their digital marketing to achieve better marketing effectiveness at lower costs. Given the importance of digital marketing to a company’s brand image and product promotion, selecting the most appropriate and reliable digital marketing service (DMS) suppliers has become a critical decision. It is difficult for a single marketing channel to meet the promotion requirements and DMS suppliers are often limited in their business scope. It is therefore essential to optimise the DMS supplier portfolio from the vast number of potential portfolios in order to best achieve the company’s operational and marketing objectives. However, existing optimization approaches often fail to adequately capture the multi-channel promotion, heterogeneity, and intangibility characteristics inherent in DMS. Accordingly, a systematic three-stage supplier portfolio optimization approach for multi-channel DMS is proposed. First, a novel prospect theory-mixed aggregation by comprehensive normalization technique (PT-MACONT) sub-model is proposed to filter out unqualified suppliers, considering the psychological attitudes of decision-makers, while avoiding the singularity of the supplier qualification evaluation dimension of prospect theory. Second, Bayesian networks are innovatively utilized to forecast the qualified suppliers’ promotional capability in terms of heterogeneity and intangibility. Third, a targeted multi-objective optimization sub-model is constructed to obtain the ideal service supplier portfolio that effectively balances diverse marketing goals. Finally, data from a leading Chinese Internet company is used to verify the validity of the proposed approach. Sensitivity and comparative analyses are implemented to illustrate the advantages in practice of the proposed approach. |
---|