Currently existing mechanisms to optimize the quality of experience (QoE) in accordance with user data plans are rather based on user self-managed usage of his mobile user data plan with, for example, help from external applications which continuously evaluate the limits reached. There is thus a need for deployment of intelligence in the network, in order to avoid abrupt downgrades in the QoE for subscribers that go over the monthly quota of their data plan. To overcome these problems, or at least to mitigate them, an embodiment of the present invention provides for an optimization of the user's quota by gradually reducing the required bandwidth for the content that the user requests. This is achieved by a method which takes into account the consumed/remaining volumes in the user data plan as well as the user browsing habits, and which provides context enrichment to content providers to enable taking decisions.