Considerable researches focus on the improvement of the algorithms of service composition while ignoring the flexibility of users to set QoS constraints. Most of them assume that users could specify QoS constraints easily. But in reality, there are more considerations in setting QoS constraints. One of them is that a user wants to know the QoS situation of service composition solutions which implement the user’s task based on candidate web services available in the market. Then he can consider both QoS requirements expectations and actual QoS implementation levels to set QoS constraints. So we propose a data structure QBand to provide more information on QoS implementation levels of current web service compositions implementing the user’s task. It calculates two aspects of information: 1) the QoS boundary of the service composition solutions implementing the task based on QoS data of candidate web services, 2) the numbers of the service compositions available for each value in QoS boundary. QBand provides users an overview of current QoS implementation levels of service composition solutions for the users’ task, to assist users during setting QoS constraints. The algorithm can also be used for quantitive attributes of other composite system.