By applying predictive analytics to an organization’s historical data, project cost estimates and risk assessments can be more accurate, improving forecasts and execution on a multi-year investment plan. The challenge with using historical data is that it is often dispersed throughout various financial and project management software systems, requiring significant data analysis for the comparison to be viable. This paper describes the steps involved in developing a robust cost benchmarking tool that centralizes historical project data into a user-friendly platform and predicts the cost of a construction project with a high degree of confidence, using only a few input variables. This paper will detail the four development steps: data gathering, data classification, data analysis, implementation, and refinement. Using the results from the tool, organizations can negotiate project cost reductions and focus risk management processes to inform contingency decisions. Across a portfolio, the changes can result in significant savings and form the basis for setting achievable performance targets.