WebML provides graphical, yet formal, specifications, embodied in a complete design process, which can be assisted by visual design tools. The main objectives of the WebML design process are:
(a) expressing the structure of a Web application with a high-level description, which can be used for querying, evolution, and maintenance;
(b) providing multiple views of the same content;
(c) separating the information content from its composition into pages, navigation, and presentation, which can be defined and evolved independently;
(d) storing the meta-information collected during the design process within a repository, which can be used during the lifetime of the application for dynamically generating Web pages;
(e) modelling users and communities explicitly in the repository, to permit the specification of personalization policies and one-to-one applications;
(f) enabling the specification of data manipulation operations for updating the site content or interacting with arbitrary external services.

Designing a data-intensive Web site amounts to specifying its characteristics in terms of various orthogonal abstractions, each captured by a distinct model.
The structure, composition, navigation, and presentation models enable the description of read-only web sites. They can be extended to cope with the specification of content management and integration with external services, through the addition of operations, which can be defined and added to the hypertext model. They are invoked as a side effect of navigation and permit one to specify commonly found interaction patterns as data entry, personal data management, and shopping carts.