In the industrial field, most systems are composed of rotating elements that are subject to failure due to their frequency of use making their monitoring of utmost importance. For optimal monitoring, non-destructive techniques are used to assess the health of components without altering their quality or interrupting their operational cycle. Among these techniques, vibration signal analysis has gained popularity because of its ability to detect a wide range of existing and incipient defects. The analysis of vibration signals is heavily dependent on the processing of the information contained in the signal. Such processing is done through various signal processing methods across different domains, namely the time-scale and time-frequency domains. This study investigated the potential benefits of combining Variational Mode Decomposition (VMD) with a Wavelet Transform (WT) to improve the time-frequency representations of vibration signals. A factorial experimental design was implemented to systematically evaluate the effect of key decomposition parameters, namely the number of VMD modes, the penalty factor, as well as the type of wavelet transform and mother wavelet, on signal decomposition and time-frequency representation. The results revealed that the most effective configuration resulted from the decomposition of the initial signal into six VMD modes with a penalty factor of 500 coupled with the CWT with a morlet wavelet. This configuration remarkably enhanced the quality of the decomposition and the resulting scalograms.