With end users demanding faster response time and management demanding lower costs and more flexibility, Service Oriented Architecture (SOA) projects are becoming more complex and brittle. Proper costing and identification of feasible benefits of SOA projects are quickly becoming a significant influence in the mainstream of all industries. SOA is intended to improve software interoperability by exposing dynamic applications as services. Current SOA quality metrics pay little attention to service complexity as an important key design feature that impacts other internal SOA quality attributes. Due to this complexity of SOA, cost and effort estimation for SOA-based software development is more difficult than that of traditional software development. Unfortunately, there is little or no effort about cost and effort estimation for SOA-based software. Traditional software cost estimation approaches are inadequate to address the complex service-oriented systems. Although numerous sources expound on the technical advantages of SOA as well as listing praises for their intuitive and qualitative benefits, until now no one has provided a reliable and quantifiable result from SOA implementations currently in production. This paper proposes a novel framework based on Work Breakdown Structure (WBS) approach for cost estimation of SOA-based software by dealing separately with service parts. The WBS framework can help organizations simplify and regulate SOA implementation cost estimation by explicit identification of SOA-specific tasks in the WBS. Furthermore, both cost estimation modelling and software sizing work can be satisfied respectively by switching the corresponding metrics within this framework. We provide an example case study to demonstrate proposed metrics and we also investigate the benefit of SOA to its adopters.