We conduct an RCT to test the effect of sending personalized communications to encourage usage of a free, on-demand tutoring resource. We randomly assigned 4,763 middle and high school students to a control group or one of three treatment arms, stratifying by school, grade, and home language (English or Spanish):
(1) Control (no additional communications),
(2) Student Only (i.e., students received personalized emails),
(3) Parent Only (i.e., parents received personalized text messages or emails), or
(4) Student and Parent (i.e., both students and parents received the personalized communications).