Chained advertisement involves breaking down a marketing campaign message into multiple banners that are shown to a user in a specific sequence in order to create a less intrusive and more effective campaign. The challenge is determining the most effective sequence of websites and banner order. This study aims to develop a recommendation system to assist with this issue. To address the vast size of the internet and the complexity of the problem, the research uses a data-driven computational approach to estimate the probability of different sequence events and apply this to real user data from a leading company. The proposed method is faster and more efficient than previous approaches.