In this study, we explore problems that arise during the portfolio optimization process that ultimately result in disappointing performance. This is a common problem in quantitative investment management which leaves investors wondering why their strong alpha signals did not transform into superior investment results once they transitioned from research to portfolio construction. We assert that the use of strong alpha signals, statistically based risk models, and a robust optimization engine that allows for the correction of alignment problems alleviate the traditional pitfalls that cause optimized portfolios to be less than optimal. We run hundreds of scenarios utilizing CTEF, an earnings forecasting model developed by John Guerard of McKinley Capital Management, to derive expected return estimates. Further, we use Axioma's World Wide risk models and optimization engine which enables us to apply its proprietary Alpha Alignment Factor during the optimization process. Finally, we construct our portfolios, alpha signals, and analytics using FactSet's Portfolio Simulation and Portfolio Analysis applications as well as the FactSet Fundamentals and FactSet Estimates databases.