Source: Data on video capsule completion come from the log file of the mobile app. The dates of the business consulting meetings and the calendar availability of consultants come from the log books kept by the consultants.

Note: This table presents the impact of holding a business consulting meeting on the probability of watching a video capsule on the same date for the training subjects in the treatment group. Each observation is a trainee × capsule × date triplet. Panel A presents the ordinary least squares (OLS) regression estimates, while panel B shows the second-stage estimates from the instrumental variables (IV) strategy, calculated via two stage least squares (2SLS). The first-stage and reduced-form estimates from the IV strategy are reported in panels C and D, respectively. The dependent variable is an indicator of trainee i’s viewership of the video capsule c at date t in panels A, B, and D, while an indicator for the event that trainee i holds a consulting meeting at t is the dependent variable in panel C. The “text message dummy” control is an indicator for the event that the consulting company sent a text reminder to encourage video capsule completion to i at t. Standard errors within parentheses are clustered at the store-owner level and are robust to heteroskedasticity of unknown form. *** Significant at 1 percent. ** Significant at 5 percent. * Significant at 10 percent.