<![CDATA[Confident spoken English is a pivotal element for guest-facing roles in tourism and hospitality, yet many students report high levels of speaking anxiety that suppress participation and performance. This qualitative study maps the ecology of English-speaking anxiety among Food and Beverage Service learners and refines a set of reduction strategies grounded in learner, teacher, classroom, and peer dynamics. Building on in-depth semi-structured interviews with second-semester students at a public polytechnic in Indonesia (n = 20) and reflexive thematic analysis, we synthesize student accounts of what amplifies or alleviates apprehension during service-English tasks. Across participants, confidence emerged as a proximal antidote to anxiety when practice was frequent, predictable, and coupled with constructive, tactful feedback. Teacher stance such as approachability, empathy, humor, and private, improvement-oriented commentary functioned as an affective gatekeeper that calibrated perceived risk. Procedure design mattered: interactive formats (group discussion, role-play) and graded exposure (progressing from collaborative to more public performances) lowered anxiety more reliably than one-off, high-stakes presentations. In this context, short, repeated sessions of off-the-shelf virtual reality (VR) job simulation were integrated as a reduction strategy not as an additional causal factor providing a low-stakes, private rehearsal space for service routines (greeting, clarifying, confirming, apologizing). Students attributed perceived gains to embodied, repeatable exposure that increased communicative self-efficacy and perceived fluency, thereby dampening state anxiety and supporting transfer to peer-facing tasks. We propose a stacked pathway from private/VR rehearsal to small-group role-plays and then to public performance, and we argue that immersive tools should be modeled as climate-contingent affordances that operate through confidence-building mechanisms. Implications for theory, curriculum, teacher development, and pre-internship readiness are discussed, alongside limitations and avenues for mechanism-focused trials.]]>
