The presence of an anomaly on the surface of a product can strongly influence how a customer perceives its quality. To resolve this problem, companies use visual inspection to check if the product contains any anomalies and check whether this anomaly can influence the perception of quality. However, given that inspectors' decisions are highly subjective, the results can be vary greatly. This paper presents a methodological approach that aims to reduce this variability. The approach is based on a distinction of three steps of visual inspection: detection, evaluation and decision steps. The methods associated with each of these steps are presented.