Space-born missions designed to search for exo-planets are providing us with high-precision photometric time series very well suited to measure the rotation period of solar- and lower-mass stars. Considering the large number of stars monitored during the mission life, the possibility to inspect each periodogram is out of question and, therefore, each method to search for rotation periods applied either individually or in combination is designed to automatically select the best rotation period estimate. In the case of unresolved binary stars it is possible to measure the rotation period of both components, when these are both variable and have comparable magnitude. On the other hand, the detection of two different rotation periods from the same photometric timeseries can represents a tool to unveil the binary nature of the star. In this paper, we present the case of the star TWA 23, a member of the young TWA association, as just one example of numerous cases. These may occur when the star is an unresolved binary and the variability of the secondary component is present in the time series. However, the frequency of the primary and dominant power peak in the periodogram, originating from the primary component, needs to be pre-whitened to allow the frequency of the secondary component to be detected. Considering the possible presence of undiscovered close binaries in the input catalogs of various ongoing or future missions aimed at exoplanet search, it becomes advisable to follow routinely this filtering approach to prevent to miss relevant information on the true binary nature of stars.