Live-streaming commerce increasingly requires viewers to regulate attention under continuous information flow. However, little is known about how linguistic constructions of time interact with the communication context to shape sustained engagement. This research examines when and why temporal metaphors in streamer communication enhance or undermine viewers’ continuous watching intention. Drawing on Construal Level Theory and processing fluency, we conduct five scenario-based experiments. These involve 3645 participants across multiple product categories. The studies manipulate temporal metaphors (fast-time vs. slow-time), communication tone (mindful vs. neutral), and product information load (high vs. low). Results reveal a systematic matching effect. Fast-time metaphors are more effective in sustaining continuous watching under low information load. Slow-time metaphors become more effective when paired with a mindful communication tone under high information load. This congruence influences watching intention by enhancing processing fluency and reducing cognitive strain during real-time information processing. These effects depend on the joint configuration of communication tone and informational demands. Temporal acceleration facilitates engagement only when cognitive resources are not overtaxed. A behavioural continuation measure in the final study replicates the focal effects, providing evidence beyond self-reported intentions. The findings advance research on live-streaming commerce. They identify temporal metaphors as a critical but underexplored linguistic mechanism for regulating sustained attention. They further extend Construal Level Theory to dynamic consumption settings by clarifying processing fluency as a central mechanism linking temporal framing to engagement. From a managerial perspective, the results highlight the importance of aligning temporal pacing cues, communication tone, and product information load to optimise sustained viewer engagement in live-streaming commerce.
Yuan et al. (Fri,) studied this question.