Consumer responses to conflicting brand news in digital feeds

Abstract

Purpose This study aims to explain how negative brand information gains influence when presented alongside positive information in digital feeds. This study addresses a gap in brand-crisis research, which has not traced how competing cues capture attention, activate violation-specific emotions, alter corresponding legitimacy beliefs and persist over time. Design/methodology/approach Three studies test a process model grounded in negativity bias and cognitive appraisal theory. Study 1 uses within-participant eye tracking (n = 117). Study 2 combines eye tracking and survey measures in a randomised experiment (n = 285). Study 3 uses a three-wave longitudinal experiment (n = 279 at T3). Findings Negative information attracted more visual attention than concurrent positive information. Greater negative attention share (NAS) intensified contempt towards worker-treatment violations and disgust towards integrity violations, which, respectively, reduced perceived employee care and ethical integrity. Anger and sadness did not mediate these relationships, while social identification strengthened both pathways. Legitimacy threats also produced larger, more persistent changes in the affected legitimacy belief than in overall brand evaluation. Research limitations/implications Brand-crisis research should treat attention as an early selection mechanism and distinguish discrete moral emotions, legitimacy domains and evaluative layers. Practical implications Brand responses should address the specific obligation under threat. Positive communication may support overall brand evaluation but may not restore beliefs about employee care or ethical integrity once directly challenged. Originality/value The findings of this study extend negativity bias from a final weighting effect to a sequence involving attentional competition, moral appraisal and layered evaluation.

The authors' abstract, as published at the source. Journal of Product & Brand Management, 2026 · DOI ↗