I ran A/B test on Instagram. It resulted in 10% usage up for Instagram Stories. 10% usage down for overall Instagram. Imagine I'm a Data Scientist. As a PM what metrics will you ask from me to make trade offs on the A/B Test?