Reconciling consumer intuition and machine: Algorithmic intuition conflict and the design of Consumer–AI collaboration

Consumers increasingly rely on AI recommenders, yet often resist their advice when it conflicts with “what feels right.†We theorize Algorithmic Intuition Conflict (AIC), the subjective misfit when algorithmic advice contradicts a user's gut choice, and show that AIC depresses satisfaction not because of perceived inaccuracy, but by eroding decision confidence and felt autonomy. Across six studies, including a field experiment with a music-streaming platform, we (1) establish AIC's causal effect on satisfaction above and beyond accuracy; (2) demonstrate dual mechanisms via reduced confidence and autonomy; (3) identify explanation quality as a system-side remedy that restores confidence; (4) reveal identity relevance as a boundary condition and show that a brief identity-anchoring prompt neutralizes AIC's confidence loss; and (5) translate mechanism to practice with a dual-track interface that presents a “logic-best†and an “intuition-friendly†option with commensurate reasons. The dual-track design improves satisfaction, reduces choice deferral, increases acceptance, and critically sustains adoption, repeat usage, and word-of-mouth in the field. We contribute a fit-based account of consumer–AI collaboration, clarify why contradictions hurt even when quality is held constant, and offer deployable interface guidance for platforms seeking to reconcile machine insight with human intuition.