This paper introduces a novel approach to dynamic multidimensional scaling (MDS) by incorporating the unfolding paradigm to analyze longitudinal preference data. The method addresses common limitations of traditional MDS in temporal settings, particularly when within-set dissimilarities are missing or ill-defined. By extending the unfolding framework and employing data augmentation techniques based on Kemeny-equivalent dissimilarities, the proposed methodology enables meaningful visualization of preference evolution over time. Applications to artificial data and applications to sensory evaluation and healthcare mobility data illustrate the flexibility and interpretability of the model, both in simulated and in real-world scenarios.