The second World Wide Web generation is increasingly oriented to services. This generation is rich in requests for dynamic and personalised content, which represents an important segment of current internet traffic. E-business applications are a particular case of the dynamic web, where dynamism and personalisation play an important role in business strategy. Understanding the characteristics of the users' workloads is very important when designing and providing web services. In e-business the importance is even greater, because this analysis is used to extract the knowledge needed to take business decisions. Workload characterisation techniques introduce some drawbacks when representing the dynamism of client behaviour. This drawback implies that most of the current workload generators model this characteristic in a simple and inaccurate way. This paper focuses on the dynamism of e-business applications and the new techniques to characterise user workloads. Our work is addressed to building a dynamic workload generator that considers the new behaviour of web clients.