Feasibility Study of Applying Character-Based Recurrent Neural Networks in Training Marketing Personnel

Artificial intelligence (AI) has been helping companies propose business solutions based on relevant data. Because promotion is one of the critical marketing decisions, referred to as one of 4P (i.e., Product, Pricing, Promotion, and Place) in marketing, it has been an important application area of AI in business. Copywriting is an essential catalyst in bridging customers and products to deliver a compelling promotion message to customers. It is a challenge for marketing managers to obtain copywriting efficiently and effectively to seize a market opportunity because they often suffer iterative discussion problems and difficulty choosing among multiple copywriting contents. Therefore, an automatic copywriting generator implemented by AI may become a feasible approach for training marketing personnel and assisting managers in marketing decision-making. The purpose of this research is to study the effectiveness of applying AI in copywriting generation for the learning of marketing personnel. This research develops character-based recurrent neural networks to generate marketing copywriting automatically and examines its effectiveness compared with award-winning copywriting by surveying marketing managers. The results show that AI-based copywriting sometimes can outperform award-winning copywriting in terms of survey scores.