Objective, reproducible and quantifiable measurements based on well-defined metrics are a widespread instrument for quality assurance in engineering disciplines and also in ontology engineering. Ontology metrics allow for the assessment of their quality and the comparison of different versions of the same ontology. We argue that such a comparison and especially the view on the evolutional evolvement bears valuable insights on the effect of explicit and implicit design decisions. This paper examines the use of quality metrics in the evolution of an ontology that is used in an image recognition context in the fashion domain. Overall, 51 incremental versions were analyzed using the OntoMetrics framework by Rostock University. Using 13 selected criteria, the evolution of the ontology is quantified and the effect of design decisions on the analyzed criteria is outlined. The critical assessment of ontology metrics is further used to uncover weak spots in the ontology. These weak spots enabled the deriving of improvement recommendations.