Soft contact tribology has emerged as a useful technique for understanding the lubrication behaviour of foods and pharmaceutical systems. While this field developed by adapting techniques used to characterise lubrication of oils, such as using the PCS instruments Mini-Traction Machine (MTM) that consists of a ball rolling/sliding against a rotating plate, methodologies have also emerged using rheometers with tribocell attachments in pure sliding configurations. Our hypothesis is that soft-contact tribology is strongly dependent on the contact mechanics, and as such the measurement from ball-to-pin contact in rheo-tribocell should differ from the ball-to-disc contact in a Mini-Traction Machine (MTM) even at a similar relative speed and contact pressure. We find that when lubricated using Newtonian fluids, the rheo-tribocell results are comparable to the MTM. However, when the lubricant is a multiphase fluid (emulsion, polymer solution) with non-Newtonian rheology, tribological behaviour measured from two devices differs markedly. We interpreted such inconsistency according to the geometrical difference in soft tribological contact, and conclude that a universal relationship cannot be derived between the devices. We highlight that Hertzian contact (assumed in Stribeck-Hersey analysis) is not universally valid in soft tribological contacts, and that it is necessary to consider the mechanics and geometry of the tribological contact when analysing the results. This also makes it hard to compare measurements and conclusions obtained from the different devices.