Your question is Experimentation in Low Traffic Areas. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You're evaluating a feature in a low-traffic product area where user volume is limited and experiments take a long time to mature. You need to decide how to run tests when standard A/B setups may be underpowered.
How do you think about experimentation in a low-traffic product area?