Larry Dysert and Bruce G. Elliot write this pdf with the purpose of utilizing the fit-for-use methodology on estimate review and validation. The pdf explains how trusting the quality of the estimate is often the first concern and sticking point between the developers of a project and the customers waiting to recieve the final product (and paying for the development). Without verification that your estimate is valid – and just as importantly, that you checked the estimate yourselves before presenting it to your customer – stakeholders are far less likely to trust the estimate altogether. While Dysert and Elliot do cover some elements that are very basic (like making sure to check your math), they also delve into the basis of the estimate, how to determine the guidelines of the estimate and how to enforce them. The pdf also explains, of course, the importance of documenting the estimate and reviewing that documentation:
The first part of this review should be the examination of the estimate documentation by the project team and project manager. This includes the basis of estimate, as well as the estimate summary and estimate detail pages. The purpose is to ensure that the estimate is presented in an understandable manner. If standard estimating guidelines have been followed (as discussed above), all estimates should be presented in a consistent and understandable style. It is very important that the project manager fully understand how the estimate is prepared because he/she often becomes the person responsible for presenting, and sometimes defending, the estimate to upper management, and later to the eventual customer. The entire project team should also understand the entire estimate package, format and contents.
Predicting probably cost is more than just giving your stakeholders a chance to understand the total cost of the project. It builds the foundation for the work being performed, conversations between the stakeholders and the development team management, and a guidepost for project progress. The more accurate the estimate can be, the more predicitable success will become.