Procedural knowledge is often transferred by spoken 'how-to' instructions delivered with simultaneous hands-on demonstration. As reflected by the popularity of YouTube instructional videos, such spoken texts have become a common way to learn how to learn a wide range of things. Surprisingly, there is a lack of research concerned with salient linguistic features and communicative functions associated with this type of discourse, referred to in the paper as procedural monologues. This study moves towards filling this gap by investigating discourse functions in procedural monologues associated with "wh"-clefts (e.g., What you want to do is use a screwdriver). This construction, considered as a highlighting device to mark relevant important points in spoken discourse, is frequently found in instructional videos. Extracted from a specialized corpus of 100 how-to-videos posted on YouTube, 130 "wh"-clefts were categorized by function to better understand how speakers shift between guiding listeners through the essential sequence of procedural steps and providing additional content to produce a coherent and cohesive text. With the underlying goal of supporting pedagogical approaches to prepare learners of English as a second or foreign language to produce communicatively dynamic procedural monologues, the study describes seven possible functions highlighted by "wh"-cleft utterances.