This paper investigates whether the observed diversification "discount" is partly due to the benchmarking error driven by a failure to consider investor recognition, the driver for one of the benefits of corporate diversification. The R-square of regression on the traditional excess value increases more than 50% and the coefficient of the diversified dummy drops more than 20% when the investor recognition proxy is included. In diversifying acquisitions involving targets with low investor recognition, the target firms are traded at a discount, however, the market reaction for those deals are favorable. Investor recognition is positively related to the excess value for standalone firms, acquisition targets, and spunoff units. My findings suggest that the benchmarking error caused by investor recognition explains a significant part of the diversification "discount" and researchers should use the benchmarking procedure with care.