Firm characteristics like value, momentum, quality, or low-volatility help explain the cross-section of stock returns and have become core pillars in the practice of factor investing. When practically implementing factor strategies, however, transaction costs can significantly impact the corresponding factor portfolios’ performances. Using proprietary trading data from a large institutional asset manager, the authors construct a realistic transaction cost model to investigate how to optimally implement factor portfolios with transaction costs. The authors provide a framework to optimize factor performance net of transaction costs but do not overly sacrifice factor exposure at the expense of lower transaction costs. They show that their analysis can be readily extended to a multi-factor setting.