Universities are continuously attempting to attract more customers via social media sites such as Facebook. However, the problem of inadequate marketing still remains. A prominent issue here is that the strategy of using different resources and contents to deliver information about the universities to their customers is inefficient because it does not solve the problem of identifying and delivering visitors' preferences. Consequently, we propose a conceptual model of a dynamic and self-Adaptive website design utilizing the concept of multi-Agent semantics web approach to include an intelligent strategic marketing for the website. The proposed model consists of five components which are, Agent Generator System; Preferences Model Base; Grouping Process; Ontology Library; Semantic Web; and Rendering Engine. The outcome of this paper is a Conceptual Multi-Agent Semantic Web Model that could be exploited to design a highly dynamic website that is expected to meet visitors' preferences.