Are users of a bibliographic database interested in learning about replications? Can we motivate them to learn? To answer these questions, we performed an experiment on a RePEc (Research Papers in Economics) website: Using randomized stratification, we allocated 324 replications and their corresponding original studies to clusters; we then drew from those clusters to select treatment and control groups. We added brightly colored tabs to the relevant webpages to alert visitors to the existence of a replication study or to the original study of a replication. We monitored traffic over three phases lasting several months: a) no treatment, b) treatment on one group, c) treatment on both groups. We find a statistically significant increase in visits to replication pages, but the effect is small: Click-throughs to the replications occurred only 1% to 1.6% of the time.