Cinderella Glass-Slipper Effect: Users Stick With First-Fit AI Models

Analysis of model adoption shows 'glass slipper' lock-in: users who first find a model that fits a high-value pain point form stable, hard-to-displace cohorts; OpenRouter analysis of billions of calls identifies foundational cohorts whose loyalty isn’t directly tied to raw performance but to 'experience fit', implying product design should prioritize personality, memory and workflow matching for retention.