Your question is Live Experiment Pitfalls. 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).
What pitfalls would you watch for when running an experiment in a live trading environment, such as peeking, novelty effects, or interference between strategies?