Job search and matching platforms are becoming increasingly common but there is limited evidence about the returns to using them. We randomize whether users of a platform in Pakistan receive information about potential job matches through only text messages or both text messages and phone calls. Calling users raises their job application rate by 1.5 percentage points or 750% of the placebo group mean, relative to only texting them. Returns to these marginal applications, measured by job interview invitations, are roughly equal to returns to inframarginal applications that would be sent regardless of the treatment status. Spillover effects of additional applications on other applicants’ probabilities of getting job interviews are negligible. The effects on job applications and interviews are substantial for all combinations of gender, baseline employment, and baseline search activity. This combination of findings suggests that search may be suboptimally low in this setting, for participants with a range of baseline levels of labor market engagement.

Written with Erica Field, Rob Garlick, and Nivedhitha Subramanian.