“Garry Tan finds that when using the G-Stack "Office Hours" feature for idea validation, he decides an idea is not worth pursuing in approximately one out of every three sessions.”