The effects of privacy policy presentation and length on trust in recommender systems: an online experiment

Recommender systems play a crucial role in e-commerce by simplifying consumer searches, improving decision making, and increasing user satisfaction, ultimately boosting web vendors’ revenues. Their effectiveness depends on the trust of the users and the availability of data to generate accurate recommendations. Using an online experiment, we examined the effect of privacy policies and the amount of requested information on trust in recommender systems and willingness to share data. The results showed that a long privacy policy reduced trust compared to a short or absent policy. The presentation of a privacy policy and the request for more data decreased participants’ willingness to share data with the system, but a long policy did not further decrease the sharing beyond the effect of a short policy. These findings suggest that to maintain trust, e-vendors should minimise data collection to what is strictly necessary, implement strong privacy protections by default, and provide concise policies that can be supplemented with more detailed versions on request.