Promise Versus Performance: Sentiment and Appraisal Tensions in Online Comments on Metaverse Advertising

Metaverse advertising has been promoted as an immersive marketing frontier, yet evidence on critical and mixed responses remains limited. This study examines 5924 advertising-relevant comments collected primarily from YouTube, supplemented by Reddit, between March 2020 and June 2026. An abductive mixed-methods computational content analysis combines codebook-based thematic analysis of 250 comments with pretrained sentiment and emotion classifiers. Sentiment classifications were compared with human coding of the 250 comments, performed without access to the automated labels, yielding Cohen’s κ = 0.533 and 70.0% agreement; emotion classifications are treated as exploratory. Neutral sentiment was most common (40.4%), followed by negative (33.2%) and positive (26.4%). Qualitative analysis developed five appraisal tension categories, with promise versus performance being the most frequently coded. Thirty-one comments in the qualitative sample (12.4%) received a tension label. This proportion does not estimate psychological ambivalence, although selected comments expressed simultaneous favorable and unfavorable evaluations. Annual sentiment varied non-monotonically, including a rebound in positive sentiment in 2023. A second sentiment classifier produced highly correlated annual series for 2021–2025, despite differences in estimated proportions. Expectation–disconfirmation theory provides an interpretive lens for perceived promise–performance discrepancies rather than a tested causal mechanism. The findings highlight the credibility of advertising promises and suggest that advertisers distinguish current capabilities from future possibilities.