
Early Experiments and Retention

How to scrappily hire for, measure, and unlock growth | Crystal Widjaja, Gojek and Kumu
Topic 5 of 11
Published Jul 31, 2022 · Clip 18:10–20:28 (2:18)
Crystal advises running experiments even with n=30 users, as trends persist but precision improves with more data; better than zero data. Focus on high conversion signals like 20/30 users acting. For retention, target pre-conversion steps like trust-building, e.g., Gojek's Facebook friends' purchases hack doubled new merchant trials.