In many cases, the use of product configurators provides a number of benefits for engineering-oriented companies, such as shorter lead times, the use of fewer man hours, and improved product quality. However, not all of these projects are successful. The task of defining what should be implemented into the configurator is much more complex than anticipated in many configurator projects. In the worst-case scenario, this implies that the configurator is never fully developed or it fails to support the users in a satisfactory manner. The difficulties in creating adequate specifications of what to implement in the knowledge base of a configurator have been explained in much of the configuration literature by the assumption that much of the relevant knowledge of the product experts is tacit. However, this paper shows that this kind of explanation can be misleading and may even be incorrect. This is accomplished by discussing the problems inherent in such explanations and through interviews with configuration project knowledge engineers. Therefore, according to this basis, this paper provides general recommendations concerning the use of the term tacit knowledge and outlines issues related to the acquisition of product knowledge in configuration projects.