Estimate your software testing effort right

Sizing a software application, estimating the effort and cost for project activities, and developing the project schedule are very important for building a successful software project. But estimation is hard. Not only because no accurate estimates can be produced by any single set of techniques, but also there is uncertainty at every phase of the software development cycle. It is no exception for. testing. Test is a key activity for software quality. There is always pressure from customer and management to commit to shorter schedules and lower cost, especially for testing. In this paper, we present a parametric estimation model for estimating testing effort/cost, along with a test graph for test cases analysis to aid in producing a more accurate estimation. The model and the test graph have been used by multi-million-dollar software projects for more than three years. It is proven that they produced very accurate estimations, within 10% of deviation, even with very high requirement volatility.