In a cloud environment, providers commit to offer the required services with specified quality metric value, as negotiated through service level agreement. But in real life scenario, often the services are not offered with committed QoS, even some quality metric may be dropped in run time. It may not be an acceptable solution for a consumer, especially in a safety critical system. We have proposed a solution to come out from this situation. The proposal is a new concept, called “Service Insurance” through which the desired quality level can be maintained against some additional payment i.e. “insurance premium”. The major contribution of the paper is to express the quality of a service in terms of underlying factors and the subsequent computation of dynamic premium amount. Another contribution is to guide the providers for re-gaining the deviated quality metric value, in case of quality metric deviation in run time. Key performance indicators are identified and the deviated quality metrics are regained through the tuning of KPIs. The entire concept is executed on a benchmark data set and the effectiveness of the mechanism is established through the validation of the hypothesis.