The performance of industrial cleaning in place (CIP) procedures is critically important for food manufacture. CIP has yet to be optimised for many processes, in part since the mechanisms involved in cleaning are not fully understood. Laboratory tests have an important role in guiding industrial trials, and this paper introduces and compares two experimental techniques developed for studying CIP mechanisms: local phosphorescence detection (LPD), and scanning fluid dynamic gauging (sFDG). To illustrate the comparison, each technique is used to investigate the influence of soil topology on the cleaning of pre-gelatinised starch-based layers from 316 stainless steel substrates by aqueous NaOH solutions at ambient temperature. The roughness of the soil surface is varied by incorporating zinc sulphide particles with different particle size distributions (range 1 - 80 μm) into the starch suspensions. The soil roughness increased with the use of larger particles, increasing the 3D arithmetic mean roughness (Sa) of the dry layers (range 0.37 - 3.33 μm). Rough layers were cleaned more readily than those containing small inclusions, with a good correlation between the cleaning rates observed during LPD and FDG measurements. The LPD technique, which is an instrumented CIP test, gives a better indication of the cleaning time, while sFDG measurements provide further insight into the removal mechanisms.