Journal of Consumer Research (2025) — Experience, confidence, and consumer choice

Summary:

This paper examines how consumer confidence shifts with experience and how that shapes what people choose next.

It argues confidence often follows a U-shape: high at first, lower once people realize how much they don’t know, then higher again as expertise builds. It also tests whether lower confidence makes people switch products or brands.

Rocklage, Berger, and Boghrati (Journal of Consumer Research) analyze large-scale review data — about 1.0M wine notes, 1.45M BeerAdvocate reviews, ~985K RateBeer reviews, and ~229K Sephora reviews — tracking each user over time.

They use the Certainty Lexicon to score how confident reviews sound, model changes across experience, and run experiments (including a repeated photo-rating task with 801 MTurk participants) plus confidence manipulations.

Results consistently show the U-shaped confidence curve across categories and methods.

When confidence dips, people are more likely to change what they pick next: in wine/beer they choose less similar options; in beer they switch brands/styles and take longer to return; in cosmetics they switch brands within a subcategory. More analytical processing is linked to faster confidence recovery.

Key limitations: confidence is inferred from language, reviewers are self-selected, and review count is only a proxy for true experience. Experiments provide the clearest causal evidence; review analyses are strong observational support.